Merge pull request #298 from lnis-uofu/micro_benchmarks
Micro benchmarks addition and testing for FPGAs with DSP blocks
This commit is contained in:
commit
1baee10e61
|
@ -223,7 +223,9 @@ The ``circuit_model_name`` should match the given name of a ``circuit_model`` de
|
|||
.. note:: A ``<pb_type name="<string>">`` parent XML node is required for the interconnect-to-circuit bindings whose interconnects are defined under the ``pb_type`` in VPR architecture description.
|
||||
|
||||
.. option:: <port name="<string>" physical_mode_port="<string>"
|
||||
physical_mode_pin_initial_offset="<int>" physical_mode_pin_rotate_offset="<int>"/>
|
||||
physical_mode_pin_initial_offset="<int>"
|
||||
physical_mode_pin_rotate_offset="<int>"/>
|
||||
physical_mode_port_rotate_offset="<int>"/>
|
||||
|
||||
Link a port of an operating ``pb_type`` to a port of a physical ``pb_type``
|
||||
|
||||
|
@ -233,7 +235,6 @@ The ``circuit_model_name`` should match the given name of a ``circuit_model`` de
|
|||
|
||||
.. note:: Users can define multiple ports. For example: ``physical_mode_pin="a[0:1] b[2:2]"``. When multiple ports are used, the ``physical_mode_pin_initial_offset`` and ``physical_mode_pin_rotate_offset`` should also be adapt. For example: ``physical_mode_pin_rotate_offset="1 0"``)
|
||||
|
||||
|
||||
- ``physical_mode_pin_initial_offset="<int>"`` aims to align the pin indices for ``port`` of ``pb_type`` between operating and physical modes, especially when part of port of operating mode is mapped to a port in physical ``pb_type``. When ``physical_mode_pin_initial_offset`` is larger than zero, the pin index of ``pb_type`` (whose index is large than 1) will be shifted by the given offset.
|
||||
|
||||
.. note:: A quick example to understand the initial offset
|
||||
|
@ -249,7 +250,24 @@ The ``circuit_model_name`` should match the given name of a ``circuit_model`` de
|
|||
|
||||
.. note:: If not defined, the default value of ``physical_mode_pin_initial_offset`` is set to ``0``.
|
||||
|
||||
- ``physical_mode_pin_rotate_offset="<int>"`` aims to align the pin indices for ``port`` of ``pb_type`` between operating and physical modes, especially when an operating mode contains multiple ``pb_type`` (``num_pb``>1) that are linked to the same physical ``pb_type``. When ``physical_mode_pin_rotate_offset`` is larger than zero, the pin index of ``pb_type`` (whose index is large than 1) will be shifted by the given offset.
|
||||
- ``physical_mode_pin_rotate_offset="<int>"`` aims to align the pin indices for ``port`` of ``pb_type`` between operating and physical modes, especially when an operating mode contains multiple ``pb_type`` (``num_pb``>1) that are linked to the same physical ``pb_type``. When ``physical_mode_pin_rotate_offset`` is larger than zero, the pin index of ``pb_type`` (whose index is large than 1) will be shifted by the given offset, **each time a pin in the operating mode is binded to a pin in the physical mode**.
|
||||
|
||||
.. note:: A quick example to understand the rotate offset
|
||||
For example, a rotating offset of 9 is used to map
|
||||
|
||||
- operating pb_type ``mult_9x9[0].a[0]`` with a full path ``mult[frac].mult_9x9[0]``
|
||||
- operating pb_type ``mult_9x9[1].a[1]`` with a full path ``mult[frac].mult_9x9[1]``
|
||||
|
||||
to
|
||||
|
||||
- physical pb_type ``mult_36x36.a[0]`` with a full path ``mult[physical].mult_36x36[0]``
|
||||
- physical pb_type ``mult_36x36.a[9]`` with a full path ``mult[physical].mult_36x36[0]``
|
||||
|
||||
.. note:: If not defined, the default value of ``physical_mode_pin_rotate_offset`` is set to ``0``.
|
||||
|
||||
.. warning:: The result of using ``physical_mode_pin_rotate_offset`` is fundementally different than ``physical_mode_port_rotate_offset``!!! Please read the examples carefully and pick the one fitting your needs.
|
||||
|
||||
- ``physical_mode_port_rotate_offset="<int>"`` aims to align the port indices for ``port`` of ``pb_type`` between operating and physical modes, especially when an operating mode contains multiple ``pb_type`` (``num_pb``>1) that are linked to the same physical ``pb_type``. When ``physical_mode_port_rotate_offset`` is larger than zero, the pin index of ``pb_type`` (whose index is large than 1) will be shifted by the given offset, **only when all the pins of a port in the operating mode is binded to all the pins of a port in the physical mode**.
|
||||
|
||||
.. note:: A quick example to understand the rotate offset
|
||||
For example, a rotating offset of 9 is used to map
|
||||
|
@ -262,7 +280,8 @@ The ``circuit_model_name`` should match the given name of a ``circuit_model`` de
|
|||
- physical pb_type ``mult_36x36.a[0:8]`` with a full path ``mult[physical].mult_36x36[0]``
|
||||
- physical pb_type ``mult_36x36.a[9:17]`` with a full path ``mult[physical].mult_36x36[0]``
|
||||
|
||||
.. note:: If not defined, the default value of ``physical_mode_pin_rotate_offset`` is set to ``0``.
|
||||
.. note:: If not defined, the default value of ``physical_mode_port_rotate_offset`` is set to ``0``.
|
||||
|
||||
|
||||
.. note::
|
||||
It is highly recommended that only one physical mode is defined for a multi-mode configurable block. Try not to use nested physical mode definition. This will ease the debugging and lead to clean XML description.
|
||||
|
|
|
@ -86,11 +86,11 @@ std::vector<std::string> PbTypeAnnotation::port_names() const {
|
|||
return keys;
|
||||
}
|
||||
|
||||
std::map<BasicPort, std::array<int, 2>> PbTypeAnnotation::physical_pb_type_port(const std::string& port_name) const {
|
||||
std::map<std::string, std::map<BasicPort, std::array<int, 2>>>::const_iterator it = operating_pb_type_ports_.find(port_name);
|
||||
std::map<BasicPort, std::array<int, 3>> PbTypeAnnotation::physical_pb_type_port(const std::string& port_name) const {
|
||||
std::map<std::string, std::map<BasicPort, std::array<int, 3>>>::const_iterator it = operating_pb_type_ports_.find(port_name);
|
||||
if (it == operating_pb_type_ports_.end()) {
|
||||
/* Return an empty port */
|
||||
return std::map<BasicPort, std::array<int, 2>>();
|
||||
return std::map<BasicPort, std::array<int, 3>>();
|
||||
}
|
||||
return operating_pb_type_ports_.at(port_name);
|
||||
}
|
||||
|
@ -169,25 +169,25 @@ void PbTypeAnnotation::set_physical_pb_type_index_offset(const int& value) {
|
|||
void PbTypeAnnotation::add_pb_type_port_pair(const std::string& operating_pb_port_name,
|
||||
const BasicPort& physical_pb_port) {
|
||||
/* Give a warning if the operating_pb_port_name already exist */
|
||||
std::map<std::string, std::map<BasicPort, std::array<int, 2>>>::const_iterator it = operating_pb_type_ports_.find(operating_pb_port_name);
|
||||
std::map<std::string, std::map<BasicPort, std::array<int, 3>>>::const_iterator it = operating_pb_type_ports_.find(operating_pb_port_name);
|
||||
|
||||
/* If not exist, initialize and set a default value */
|
||||
if (it == operating_pb_type_ports_.end()) {
|
||||
operating_pb_type_ports_[operating_pb_port_name][physical_pb_port] = {0, 0};
|
||||
operating_pb_type_ports_[operating_pb_port_name][physical_pb_port] = {0, 0, 0};
|
||||
/* We can return early */
|
||||
return;
|
||||
}
|
||||
|
||||
/* If the physical port is not in the list, we create one and set a default value */
|
||||
if (0 == operating_pb_type_ports_[operating_pb_port_name].count(physical_pb_port)) {
|
||||
operating_pb_type_ports_[operating_pb_port_name][physical_pb_port] = {0, 0};
|
||||
operating_pb_type_ports_[operating_pb_port_name][physical_pb_port] = {0, 0, 0};
|
||||
}
|
||||
}
|
||||
|
||||
void PbTypeAnnotation::set_physical_pin_initial_offset(const std::string& operating_pb_port_name,
|
||||
const BasicPort& physical_pb_port,
|
||||
const int& physical_pin_initial_offset) {
|
||||
std::map<std::string, std::map<BasicPort, std::array<int, 2>>>::const_iterator it = operating_pb_type_ports_.find(operating_pb_port_name);
|
||||
std::map<std::string, std::map<BasicPort, std::array<int, 3>>>::const_iterator it = operating_pb_type_ports_.find(operating_pb_port_name);
|
||||
|
||||
if (it == operating_pb_type_ports_.end()) {
|
||||
VTR_LOG_ERROR("The operating pb_type port '%s' is not valid!\n",
|
||||
|
@ -210,7 +210,7 @@ void PbTypeAnnotation::set_physical_pin_initial_offset(const std::string& operat
|
|||
void PbTypeAnnotation::set_physical_pin_rotate_offset(const std::string& operating_pb_port_name,
|
||||
const BasicPort& physical_pb_port,
|
||||
const int& physical_pin_rotate_offset) {
|
||||
std::map<std::string, std::map<BasicPort, std::array<int, 2>>>::const_iterator it = operating_pb_type_ports_.find(operating_pb_port_name);
|
||||
std::map<std::string, std::map<BasicPort, std::array<int, 3>>>::const_iterator it = operating_pb_type_ports_.find(operating_pb_port_name);
|
||||
|
||||
if (it == operating_pb_type_ports_.end()) {
|
||||
VTR_LOG_ERROR("The operating pb_type port '%s' is not valid!\n",
|
||||
|
@ -230,6 +230,30 @@ void PbTypeAnnotation::set_physical_pin_rotate_offset(const std::string& operati
|
|||
operating_pb_type_ports_[operating_pb_port_name][physical_pb_port][1] = physical_pin_rotate_offset;
|
||||
}
|
||||
|
||||
void PbTypeAnnotation::set_physical_port_rotate_offset(const std::string& operating_pb_port_name,
|
||||
const BasicPort& physical_pb_port,
|
||||
const int& physical_port_rotate_offset) {
|
||||
std::map<std::string, std::map<BasicPort, std::array<int, 3>>>::const_iterator it = operating_pb_type_ports_.find(operating_pb_port_name);
|
||||
|
||||
if (it == operating_pb_type_ports_.end()) {
|
||||
VTR_LOG_ERROR("The operating pb_type port '%s' is not valid!\n",
|
||||
operating_pb_port_name.c_str());
|
||||
exit(1);
|
||||
}
|
||||
|
||||
if (operating_pb_type_ports_[operating_pb_port_name].end() == operating_pb_type_ports_[operating_pb_port_name].find(physical_pb_port)) {
|
||||
VTR_LOG_ERROR("The physical pb_type port '%s[%lu:%lu]' definition for operating pb_type port '%s' is not valid!\n",
|
||||
physical_pb_port.get_name().c_str(),
|
||||
physical_pb_port.get_lsb(),
|
||||
physical_pb_port.get_msb(),
|
||||
operating_pb_port_name.c_str());
|
||||
exit(1);
|
||||
}
|
||||
|
||||
operating_pb_type_ports_[operating_pb_port_name][physical_pb_port][2] = physical_port_rotate_offset;
|
||||
}
|
||||
|
||||
|
||||
void PbTypeAnnotation::add_interconnect_circuit_model_pair(const std::string& interc_name,
|
||||
const std::string& circuit_model_name) {
|
||||
std::map<std::string, std::string>::const_iterator it = interconnect_circuit_model_names_.find(interc_name);
|
||||
|
|
|
@ -49,7 +49,7 @@ class PbTypeAnnotation {
|
|||
float physical_pb_type_index_factor() const;
|
||||
int physical_pb_type_index_offset() const;
|
||||
std::vector<std::string> port_names() const;
|
||||
std::map<BasicPort, std::array<int, 2>> physical_pb_type_port(const std::string& port_name) const;
|
||||
std::map<BasicPort, std::array<int, 3>> physical_pb_type_port(const std::string& port_name) const;
|
||||
std::vector<std::string> interconnect_names() const;
|
||||
std::string interconnect_circuit_model_name(const std::string& interc_name) const;
|
||||
public: /* Public mutators */
|
||||
|
@ -73,6 +73,9 @@ class PbTypeAnnotation {
|
|||
void set_physical_pin_rotate_offset(const std::string& operating_pb_port_name,
|
||||
const BasicPort& physical_pb_port,
|
||||
const int& physical_pin_rotate_offset);
|
||||
void set_physical_port_rotate_offset(const std::string& operating_pb_port_name,
|
||||
const BasicPort& physical_pb_port,
|
||||
const int& physical_port_rotate_offset);
|
||||
void add_interconnect_circuit_model_pair(const std::string& interc_name,
|
||||
const std::string& circuit_model_name);
|
||||
private: /* Internal data */
|
||||
|
@ -138,10 +141,10 @@ class PbTypeAnnotation {
|
|||
int physical_pb_type_index_offset_;
|
||||
|
||||
/* Link from the pins under an operating pb_type to pairs of
|
||||
* its physical pb_type and its pin initial & rotating offset
|
||||
*
|
||||
* Note that initial offset is the first element of the std::array
|
||||
* Note that rotating offset is the second element of the std::array
|
||||
* its physical pb_type and
|
||||
* - its pin initial offset: the first element of the std::array
|
||||
* - pin-level rotating offset: the second element of the std::array
|
||||
* - port-level rotating offset: the third element of the std::array
|
||||
*
|
||||
* The offsets aim to align the pin indices for port of pb_type
|
||||
* between operating and physical modes, especially when an operating
|
||||
|
@ -158,14 +161,21 @@ class PbTypeAnnotation {
|
|||
* physical pb_type bram[0].dout_a[0] with a full path memory[physical].bram[0]
|
||||
* physical pb_type bram[0].dout_a[1] with a full path memory[physical].bram[0]
|
||||
*
|
||||
* For example, a rotating offset of 9 is used to map
|
||||
* For example, a pin-level rotating offset of 9 is used to map
|
||||
* operating pb_type mult_9x9[0].a[0] with a full path mult[frac].mult_9x9[0]
|
||||
* operating pb_type mult_9x9[0].a[1] with a full path mult[frac].mult_9x9[1]
|
||||
* to
|
||||
* physical pb_type mult_36x36.a[0] with a full path mult[physical].mult_36x36[0]
|
||||
* physical pb_type mult_36x36.a[9] with a full path mult[physical].mult_36x36[0]
|
||||
*
|
||||
* For example, a port-level rotating offset of 9 is used to map
|
||||
* operating pb_type mult_9x9[0].a[0:8] with a full path mult[frac].mult_9x9[0]
|
||||
* operating pb_type mult_9x9[1].a[0:8] with a full path mult[frac].mult_9x9[1]
|
||||
* to
|
||||
* physical pb_type mult_36x36.a[0:8] with a full path mult[physical].mult_36x36[0]
|
||||
* physical pb_type mult_36x36.a[9:17] with a full path mult[physical].mult_36x36[0]
|
||||
*/
|
||||
std::map<std::string, std::map<BasicPort, std::array<int, 2>>> operating_pb_type_ports_;
|
||||
std::map<std::string, std::map<BasicPort, std::array<int, 3>>> operating_pb_type_ports_;
|
||||
|
||||
/* Link between the interconnects under this pb_type and circuit model names */
|
||||
std::map<std::string, std::string> interconnect_circuit_model_names_;
|
||||
|
|
|
@ -114,6 +114,31 @@ void read_xml_pb_port_annotation(pugi::xml_node& xml_port,
|
|||
std::stoi(rotate_offsets[iport]));
|
||||
}
|
||||
}
|
||||
|
||||
/* We have an optional attribute: physical_mode_port_rotate_offset
|
||||
* Split based on the number of physical pb_type ports that have been defined
|
||||
*/
|
||||
const std::string& physical_port_rotate_offset_attr = get_attribute(xml_port, "physical_mode_port_rotate_offset", loc_data, pugiutil::ReqOpt::OPTIONAL).as_string();
|
||||
|
||||
if (false == physical_port_rotate_offset_attr.empty()) {
|
||||
/* Split the physical mode port attributes with space */
|
||||
openfpga::StringToken offset_tokenizer(physical_port_rotate_offset_attr);
|
||||
const std::vector<std::string> rotate_offsets = offset_tokenizer.split();
|
||||
|
||||
/* Error out if the offset does not match the port definition */
|
||||
if (physical_mode_ports.size() != rotate_offsets.size()) {
|
||||
archfpga_throw(loc_data.filename_c_str(), loc_data.line(xml_port),
|
||||
"Defined %lu physical mode ports but only %lu physical port rotate offset are defined! Expect size matching.\n",
|
||||
physical_mode_ports.size(), rotate_offsets.size());
|
||||
}
|
||||
|
||||
for (size_t iport = 0; iport < physical_mode_ports.size(); ++iport) {
|
||||
openfpga::PortParser port_parser(physical_mode_ports[iport]);
|
||||
pb_type_annotation.set_physical_port_rotate_offset(name_attr,
|
||||
port_parser.port(),
|
||||
std::stoi(rotate_offsets[iport]));
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
/********************************************************************
|
||||
|
|
|
@ -144,7 +144,14 @@ void write_xml_pb_port_annotation(std::fstream& fp,
|
|||
physical_mode_pin_rotate_offset_attr += std::to_string(physical_pb_port_pair.second[1]);
|
||||
}
|
||||
write_xml_attribute(fp, "physical_mode_pin_rotate_offset", physical_mode_pin_rotate_offset_attr.c_str());
|
||||
|
||||
std::string physical_mode_port_rotate_offset_attr;
|
||||
for (const auto& physical_pb_port_pair : pb_type_annotation.physical_pb_type_port(port_name)) {
|
||||
if (false == physical_mode_port_rotate_offset_attr.empty()) {
|
||||
physical_mode_port_rotate_offset_attr += " ";
|
||||
}
|
||||
physical_mode_port_rotate_offset_attr += std::to_string(physical_pb_port_pair.second[2]);
|
||||
}
|
||||
write_xml_attribute(fp, "physical_mode_port_rotate_offset", physical_mode_port_rotate_offset_attr.c_str());
|
||||
fp << "/>" << "\n";
|
||||
}
|
||||
|
||||
|
|
|
@ -333,7 +333,7 @@ bool try_match_pb_graph_pin(t_pb_graph_pin* operating_pb_graph_pin,
|
|||
* by the pin rotate offset value
|
||||
* The accumulated offset will be reset to 0 when it exceeds the msb() of the physical port
|
||||
*/
|
||||
int acc_offset = vpr_device_annotation.physical_pb_pin_offset(operating_pb_graph_pin->port, candidate_port);
|
||||
int acc_offset = vpr_device_annotation.physical_pb_pin_offset(operating_pb_graph_pin->port, candidate_port) + vpr_device_annotation.physical_pb_port_offset(operating_pb_graph_pin->port, candidate_port);
|
||||
int init_offset = vpr_device_annotation.physical_pb_pin_initial_offset(operating_pb_graph_pin->port, candidate_port);
|
||||
const BasicPort& physical_port_range = vpr_device_annotation.physical_pb_port_range(operating_pb_graph_pin->port, candidate_port);
|
||||
if (physical_pb_graph_pin->pin_number != operating_pb_graph_pin->pin_number
|
||||
|
@ -463,6 +463,14 @@ void annotate_physical_pb_graph_node_pins(t_pb_graph_node* operating_pb_graph_no
|
|||
physical_pb_graph_node, vpr_device_annotation,
|
||||
verbose_output);
|
||||
}
|
||||
/* Finish a port, accumulate the port-level offset affiliated to the port */
|
||||
if (0 == operating_pb_graph_node->num_input_pins[iport]) {
|
||||
continue;
|
||||
}
|
||||
t_pb_graph_pin* operating_pb_graph_pin = &(operating_pb_graph_node->input_pins[iport][0]);
|
||||
for (t_port* candidate_port : vpr_device_annotation.physical_pb_port(operating_pb_graph_pin->port)) {
|
||||
vpr_device_annotation.accumulate_physical_pb_port_rotate_offset(operating_pb_graph_pin->port, candidate_port);
|
||||
}
|
||||
}
|
||||
|
||||
for (int iport = 0; iport < operating_pb_graph_node->num_output_ports; ++iport) {
|
||||
|
@ -471,6 +479,14 @@ void annotate_physical_pb_graph_node_pins(t_pb_graph_node* operating_pb_graph_no
|
|||
physical_pb_graph_node, vpr_device_annotation,
|
||||
verbose_output);
|
||||
}
|
||||
/* Finish a port, accumulate the port-level offset affiliated to the port */
|
||||
if (0 == operating_pb_graph_node->num_output_pins[iport]) {
|
||||
continue;
|
||||
}
|
||||
t_pb_graph_pin* operating_pb_graph_pin = &(operating_pb_graph_node->output_pins[iport][0]);
|
||||
for (t_port* candidate_port : vpr_device_annotation.physical_pb_port(operating_pb_graph_pin->port)) {
|
||||
vpr_device_annotation.accumulate_physical_pb_port_rotate_offset(operating_pb_graph_pin->port, candidate_port);
|
||||
}
|
||||
}
|
||||
|
||||
for (int iport = 0; iport < operating_pb_graph_node->num_clock_ports; ++iport) {
|
||||
|
@ -479,6 +495,14 @@ void annotate_physical_pb_graph_node_pins(t_pb_graph_node* operating_pb_graph_no
|
|||
physical_pb_graph_node, vpr_device_annotation,
|
||||
verbose_output);
|
||||
}
|
||||
/* Finish a port, accumulate the port-level offset affiliated to the port */
|
||||
if (0 == operating_pb_graph_node->num_clock_pins[iport]) {
|
||||
continue;
|
||||
}
|
||||
t_pb_graph_pin* operating_pb_graph_pin = &(operating_pb_graph_node->clock_pins[iport][0]);
|
||||
for (t_port* candidate_port : vpr_device_annotation.physical_pb_port(operating_pb_graph_pin->port)) {
|
||||
vpr_device_annotation.accumulate_physical_pb_port_rotate_offset(operating_pb_graph_pin->port, candidate_port);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
|
|
|
@ -211,7 +211,7 @@ bool pair_operating_and_physical_pb_types(t_pb_type* operating_pb_type,
|
|||
* if not found, we assume that the physical port is the same as the operating pb_port
|
||||
*/
|
||||
for (t_port* operating_pb_port : pb_type_ports(operating_pb_type)) {
|
||||
std::map<BasicPort, std::array<int, 2>> expected_physical_pb_ports = pb_type_annotation.physical_pb_type_port(std::string(operating_pb_port->name));
|
||||
std::map<BasicPort, std::array<int, 3>> expected_physical_pb_ports = pb_type_annotation.physical_pb_type_port(std::string(operating_pb_port->name));
|
||||
|
||||
/* If not defined in the annotation, set the default pair:
|
||||
* rotate_offset is 0 by default!
|
||||
|
@ -243,6 +243,7 @@ bool pair_operating_and_physical_pb_types(t_pb_type* operating_pb_type,
|
|||
vpr_device_annotation.add_physical_pb_port_range(operating_pb_port, physical_pb_port, expected_physical_pb_port.first);
|
||||
vpr_device_annotation.add_physical_pb_pin_initial_offset(operating_pb_port, physical_pb_port, expected_physical_pb_port.second[0]);
|
||||
vpr_device_annotation.add_physical_pb_pin_rotate_offset(operating_pb_port, physical_pb_port, expected_physical_pb_port.second[1]);
|
||||
vpr_device_annotation.add_physical_pb_port_rotate_offset(operating_pb_port, physical_pb_port, expected_physical_pb_port.second[2]);
|
||||
}
|
||||
}
|
||||
|
||||
|
|
|
@ -220,6 +220,21 @@ int VprDeviceAnnotation::physical_pb_pin_rotate_offset(t_port* operating_pb_port
|
|||
return physical_pb_pin_rotate_offsets_.at(operating_pb_port).at(physical_pb_port);
|
||||
}
|
||||
|
||||
int VprDeviceAnnotation::physical_pb_port_rotate_offset(t_port* operating_pb_port,
|
||||
t_port* physical_pb_port) const {
|
||||
/* Ensure that the pb_type is in the list */
|
||||
std::map<t_port*, std::map<t_port*, int>>::const_iterator it = physical_pb_port_rotate_offsets_.find(operating_pb_port);
|
||||
if (it == physical_pb_port_rotate_offsets_.end()) {
|
||||
/* Default value is 0 */
|
||||
return 0;
|
||||
}
|
||||
if (0 == physical_pb_port_rotate_offsets_.at(operating_pb_port).count(physical_pb_port)) {
|
||||
/* Default value is 0 */
|
||||
return 0;
|
||||
}
|
||||
return physical_pb_port_rotate_offsets_.at(operating_pb_port).at(physical_pb_port);
|
||||
}
|
||||
|
||||
int VprDeviceAnnotation::physical_pb_pin_offset(t_port* operating_pb_port,
|
||||
t_port* physical_pb_port) const {
|
||||
/* Ensure that the pb_type is in the list */
|
||||
|
@ -235,6 +250,21 @@ int VprDeviceAnnotation::physical_pb_pin_offset(t_port* operating_pb_port,
|
|||
return physical_pb_pin_offsets_.at(operating_pb_port).at(physical_pb_port);
|
||||
}
|
||||
|
||||
int VprDeviceAnnotation::physical_pb_port_offset(t_port* operating_pb_port,
|
||||
t_port* physical_pb_port) const {
|
||||
/* Ensure that the pb_type is in the list */
|
||||
std::map<t_port*, std::map<t_port*, int>>::const_iterator it = physical_pb_port_offsets_.find(operating_pb_port);
|
||||
if (it == physical_pb_port_offsets_.end()) {
|
||||
/* Default value is 0 */
|
||||
return 0;
|
||||
}
|
||||
if (0 == physical_pb_port_offsets_.at(operating_pb_port).count(physical_pb_port)) {
|
||||
/* Default value is 0 */
|
||||
return 0;
|
||||
}
|
||||
return physical_pb_port_offsets_.at(operating_pb_port).at(physical_pb_port);
|
||||
}
|
||||
|
||||
t_pb_graph_pin* VprDeviceAnnotation::physical_pb_graph_pin(const t_pb_graph_pin* pb_graph_pin) const {
|
||||
/* Ensure that the pb_type is in the list */
|
||||
std::map<const t_pb_graph_pin*, t_pb_graph_pin*>::const_iterator it = physical_pb_graph_pins_.find(pb_graph_pin);
|
||||
|
@ -478,6 +508,28 @@ void VprDeviceAnnotation::add_physical_pb_pin_initial_offset(t_port* operating_p
|
|||
physical_pb_pin_initial_offsets_[operating_pb_port][physical_pb_port] = offset;
|
||||
}
|
||||
|
||||
void VprDeviceAnnotation::add_physical_pb_port_rotate_offset(t_port* operating_pb_port,
|
||||
t_port* physical_pb_port,
|
||||
const int& offset) {
|
||||
/* Warn any override attempt */
|
||||
std::map<t_port*, std::map<t_port*, int>>::const_iterator it = physical_pb_port_rotate_offsets_.find(operating_pb_port);
|
||||
if ( (it != physical_pb_port_rotate_offsets_.end())
|
||||
&& (0 < physical_pb_port_rotate_offsets_[operating_pb_port].count(physical_pb_port)) ) {
|
||||
VTR_LOG_WARN("Override the annotation between operating pb_port '%s' and it physical pb_port '%s' port rotate offset '%d'!\n",
|
||||
operating_pb_port->name, offset);
|
||||
}
|
||||
|
||||
physical_pb_port_rotate_offsets_[operating_pb_port][physical_pb_port] = offset;
|
||||
/* We initialize the accumulated offset to 0 */
|
||||
physical_pb_port_offsets_[operating_pb_port][physical_pb_port] = 0;
|
||||
}
|
||||
|
||||
|
||||
void VprDeviceAnnotation::accumulate_physical_pb_port_rotate_offset(t_port* operating_pb_port,
|
||||
t_port* physical_pb_port) {
|
||||
physical_pb_port_offsets_[operating_pb_port][physical_pb_port] += physical_pb_port_rotate_offsets_[operating_pb_port][physical_pb_port];
|
||||
}
|
||||
|
||||
void VprDeviceAnnotation::add_physical_pb_pin_rotate_offset(t_port* operating_pb_port,
|
||||
t_port* physical_pb_port,
|
||||
const int& offset) {
|
||||
|
|
|
@ -67,6 +67,9 @@ class VprDeviceAnnotation {
|
|||
int physical_pb_pin_rotate_offset(t_port* operating_pb_port,
|
||||
t_port* physical_pb_port) const;
|
||||
|
||||
int physical_pb_port_rotate_offset(t_port* operating_pb_port,
|
||||
t_port* physical_pb_port) const;
|
||||
|
||||
/**This function returns an accumulated offset. Note that the
|
||||
* accumulated offset is NOT the pin rotate offset specified by users
|
||||
* It is an aggregation of the offset during pin pairing
|
||||
|
@ -76,6 +79,8 @@ class VprDeviceAnnotation {
|
|||
*/
|
||||
int physical_pb_pin_offset(t_port* operating_pb_port,
|
||||
t_port* physical_pb_port) const;
|
||||
int physical_pb_port_offset(t_port* operating_pb_port,
|
||||
t_port* physical_pb_port) const;
|
||||
t_pb_graph_pin* physical_pb_graph_pin(const t_pb_graph_pin* pb_graph_pin) const;
|
||||
CircuitModelId rr_switch_circuit_model(const RRSwitchId& rr_switch) const;
|
||||
CircuitModelId rr_segment_circuit_model(const RRSegmentId& rr_segment) const;
|
||||
|
@ -106,6 +111,11 @@ class VprDeviceAnnotation {
|
|||
void add_physical_pb_pin_initial_offset(t_port* operating_pb_port,
|
||||
t_port* physical_pb_port,
|
||||
const int& offset);
|
||||
void add_physical_pb_port_rotate_offset(t_port* operating_pb_port,
|
||||
t_port* physical_pb_port,
|
||||
const int& offset);
|
||||
void accumulate_physical_pb_port_rotate_offset(t_port* operating_pb_port,
|
||||
t_port* physical_pb_port);
|
||||
void add_physical_pb_pin_rotate_offset(t_port* operating_pb_port,
|
||||
t_port* physical_pb_port,
|
||||
const int& offset);
|
||||
|
@ -166,8 +176,11 @@ class VprDeviceAnnotation {
|
|||
std::map<t_port*, std::vector<t_port*>> physical_pb_ports_;
|
||||
std::map<t_port*, std::map<t_port*, int>> physical_pb_pin_initial_offsets_;
|
||||
std::map<t_port*, std::map<t_port*, int>> physical_pb_pin_rotate_offsets_;
|
||||
std::map<t_port*, std::map<t_port*, int>> physical_pb_port_rotate_offsets_;
|
||||
|
||||
/* Accumulated offsets for a physical pb_type port, just for internal usage */
|
||||
/* Accumulated offsets for a physical pb port, just for internal usage */
|
||||
std::map<t_port*, std::map<t_port*, int>> physical_pb_port_offsets_;
|
||||
/* Accumulated offsets for a physical pb_graph_pin, just for internal usage */
|
||||
std::map<t_port*, std::map<t_port*, int>> physical_pb_pin_offsets_;
|
||||
|
||||
/* Pair a pb_port to its LSB and MSB of a physical pb_port
|
||||
|
|
|
@ -28,6 +28,8 @@ namespace openfpga {
|
|||
* - sink is an input of a primitive pb_type
|
||||
*
|
||||
* Note:
|
||||
* - This function is applicable ONLY to single-mode pb_types!!! Because their routing traces
|
||||
* are deterministic: there is only 1 valid path from a source pin to a sink pin!!!
|
||||
* - If there is a fan-out of the current source pb graph pin is not a direct interconnection
|
||||
* the direct search should stop.
|
||||
* - This function is designed for pb graph without local routing
|
||||
|
@ -58,6 +60,11 @@ bool rec_direct_search_sink_pb_graph_pins(const t_pb_graph_pin* source_pb_pin,
|
|||
|
||||
std::vector<t_pb_graph_pin*> sink_pb_pins_to_search;
|
||||
|
||||
/* Only support single-mode pb_type!!! */
|
||||
//if (1 != source_pb_pin->parent_node->pb_type->num_modes) {
|
||||
// return false;
|
||||
//}
|
||||
|
||||
for (int iedge = 0; iedge < source_pb_pin->num_output_edges; ++iedge) {
|
||||
if (DIRECT_INTERC != source_pb_pin->output_edges[iedge]->interconnect->type) {
|
||||
return false;
|
||||
|
|
|
@ -0,0 +1,22 @@
|
|||
//-------------------------------------------------------
|
||||
// Functionality: A 12-bit multiply-acculumate circuit
|
||||
// Author: Xifan Tang
|
||||
//-------------------------------------------------------
|
||||
|
||||
module mac_12(a, b, c, out);
|
||||
parameter DATA_WIDTH = 12; /* declare a parameter. default required */
|
||||
input [DATA_WIDTH - 1 : 0] a, b, c;
|
||||
output [DATA_WIDTH - 1 : 0] out;
|
||||
|
||||
assign out = a * b + c;
|
||||
|
||||
endmodule
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
|
@ -0,0 +1,22 @@
|
|||
//-------------------------------------------------------
|
||||
// Functionality: A 16-bit multiply-acculumate circuit
|
||||
// Author: Xifan Tang
|
||||
//-------------------------------------------------------
|
||||
|
||||
module mac_16(a, b, c, out);
|
||||
parameter DATA_WIDTH = 16; /* declare a parameter. default required */
|
||||
input [DATA_WIDTH - 1 : 0] a, b, c;
|
||||
output [DATA_WIDTH - 1 : 0] out;
|
||||
|
||||
assign out = a * b + c;
|
||||
|
||||
endmodule
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
|
@ -0,0 +1,22 @@
|
|||
//-------------------------------------------------------
|
||||
// Functionality: A 2-bit multiply-acculumate circuit
|
||||
// Author: Xifan Tang
|
||||
//-------------------------------------------------------
|
||||
|
||||
module mac_2(a, b, c, out);
|
||||
parameter DATA_WIDTH = 2; /* declare a parameter. default required */
|
||||
input [DATA_WIDTH - 1 : 0] a, b, c;
|
||||
output [DATA_WIDTH - 1 : 0] out;
|
||||
|
||||
assign out = a * b + c;
|
||||
|
||||
endmodule
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
|
@ -0,0 +1,22 @@
|
|||
//-------------------------------------------------------
|
||||
// Functionality: A 32-bit multiply-acculumate circuit
|
||||
// Author: Xifan Tang
|
||||
//-------------------------------------------------------
|
||||
|
||||
module mac_32(a, b, c, out);
|
||||
parameter DATA_WIDTH = 32; /* declare a parameter. default required */
|
||||
input [DATA_WIDTH - 1 : 0] a, b, c;
|
||||
output [DATA_WIDTH - 1 : 0] out;
|
||||
|
||||
assign out = a * b + c;
|
||||
|
||||
endmodule
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
|
@ -0,0 +1,22 @@
|
|||
//-------------------------------------------------------
|
||||
// Functionality: A 4-bit multiply-acculumate circuit
|
||||
// Author: Xifan Tang
|
||||
//-------------------------------------------------------
|
||||
|
||||
module mac_4(a, b, c, out);
|
||||
parameter DATA_WIDTH = 4; /* declare a parameter. default required */
|
||||
input [DATA_WIDTH - 1 : 0] a, b, c;
|
||||
output [DATA_WIDTH - 1 : 0] out;
|
||||
|
||||
assign out = a * b + c;
|
||||
|
||||
endmodule
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
|
@ -0,0 +1,22 @@
|
|||
//-------------------------------------------------------
|
||||
// Functionality: A 6-bit multiply-acculumate circuit
|
||||
// Author: Xifan Tang
|
||||
//-------------------------------------------------------
|
||||
|
||||
module mac_6(a, b, c, out);
|
||||
parameter DATA_WIDTH = 6; /* declare a parameter. default required */
|
||||
input [DATA_WIDTH - 1 : 0] a, b, c;
|
||||
output [DATA_WIDTH - 1 : 0] out;
|
||||
|
||||
assign out = a * b + c;
|
||||
|
||||
endmodule
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
|
@ -262,7 +262,6 @@
|
|||
<!-- Binding operating pb_type to physical pb_type -->
|
||||
<!-- Binding operating pb_types in mode 'arithmetic' -->
|
||||
<pb_type name="clb.fle[arithmetic].soft_adder.adder_lut4" physical_pb_type_name="clb.fle[physical].fabric.frac_logic.frac_lut4" mode_bits="1"/>
|
||||
<pb_type name="clb.fle[arithmetic].soft_adder.ff" physical_pb_type_name="clb.fle[physical].fabric.ff"/>
|
||||
<pb_type name="clb.fle[arithmetic].soft_adder.carry_follower" physical_pb_type_name="clb.fle[physical].fabric.frac_logic.carry_follower"/>
|
||||
<!-- Binding operating pb_types in mode 'n2_lut3' -->
|
||||
<pb_type name="clb.fle[n2_lut3].lut3inter.ble3.lut3" physical_pb_type_name="clb.fle[physical].fabric.frac_logic.frac_lut4" mode_bits="1" physical_pb_type_index_factor="0.5">
|
||||
|
|
|
@ -0,0 +1,295 @@
|
|||
<!-- Architecture annotation for OpenFPGA framework
|
||||
This annotation supports the k4_frac_cc_sky130nm.xml
|
||||
- General purpose logic block
|
||||
- K = 6, N = 10, I = 40
|
||||
- Single mode
|
||||
- Routing architecture
|
||||
- L = 4, fc_in = 0.15, fc_out = 0.1
|
||||
- Skywater 130nm PDK
|
||||
- circuit models are binded to the opensource skywater
|
||||
foundry middle-speed (ms) standard cell library
|
||||
-->
|
||||
<openfpga_architecture>
|
||||
<technology_library>
|
||||
<device_library>
|
||||
<device_model name="logic" type="transistor">
|
||||
<lib type="industry" corner="TOP_TT" ref="M" path="${OPENFPGA_PATH}/openfpga_flow/tech/PTM_45nm/45nm.pm"/>
|
||||
<design vdd="0.9" pn_ratio="2"/>
|
||||
<pmos name="pch" chan_length="40e-9" min_width="140e-9" variation="logic_transistor_var"/>
|
||||
<nmos name="nch" chan_length="40e-9" min_width="140e-9" variation="logic_transistor_var"/>
|
||||
</device_model>
|
||||
<device_model name="io" type="transistor">
|
||||
<lib type="academia" ref="M" path="${OPENFPGA_PATH}/openfpga_flow/tech/PTM_45nm/45nm.pm"/>
|
||||
<design vdd="2.5" pn_ratio="3"/>
|
||||
<pmos name="pch_25" chan_length="270e-9" min_width="320e-9" variation="io_transistor_var"/>
|
||||
<nmos name="nch_25" chan_length="270e-9" min_width="320e-9" variation="io_transistor_var"/>
|
||||
</device_model>
|
||||
</device_library>
|
||||
<variation_library>
|
||||
<variation name="logic_transistor_var" abs_deviation="0.1" num_sigma="3"/>
|
||||
<variation name="io_transistor_var" abs_deviation="0.1" num_sigma="3"/>
|
||||
</variation_library>
|
||||
</technology_library>
|
||||
<circuit_library>
|
||||
<circuit_model type="inv_buf" name="sky130_fd_sc_hd__inv_1" prefix="sky130_fd_sc_hd__inv_1" is_default="true">
|
||||
<design_technology type="cmos" topology="inverter" size="1"/>
|
||||
<device_technology device_model_name="logic"/>
|
||||
<port type="input" prefix="in" lib_name="A" size="1"/>
|
||||
<port type="output" prefix="out" lib_name="Y" size="1"/>
|
||||
<delay_matrix type="rise" in_port="in" out_port="out">
|
||||
10e-12
|
||||
</delay_matrix>
|
||||
<delay_matrix type="fall" in_port="in" out_port="out">
|
||||
10e-12
|
||||
</delay_matrix>
|
||||
</circuit_model>
|
||||
<circuit_model type="inv_buf" name="sky130_fd_sc_hd__buf_2" prefix="sky130_fd_sc_hd__buf_2" is_default="false">
|
||||
<design_technology type="cmos" topology="buffer" size="1" num_level="2" f_per_stage="2"/>
|
||||
<device_technology device_model_name="logic"/>
|
||||
<port type="input" prefix="in" lib_name="A" size="1"/>
|
||||
<port type="output" prefix="out" lib_name="X" size="1"/>
|
||||
<delay_matrix type="rise" in_port="in" out_port="out">
|
||||
10e-12
|
||||
</delay_matrix>
|
||||
<delay_matrix type="fall" in_port="in" out_port="out">
|
||||
10e-12
|
||||
</delay_matrix>
|
||||
</circuit_model>
|
||||
<circuit_model type="inv_buf" name="sky130_fd_sc_hd__buf_4" prefix="sky130_fd_sc_hd__buf_4" is_default="false">
|
||||
<design_technology type="cmos" topology="buffer" size="1" num_level="2" f_per_stage="4"/>
|
||||
<device_technology device_model_name="logic"/>
|
||||
<port type="input" prefix="in" lib_name="A" size="1"/>
|
||||
<port type="output" prefix="out" lib_name="X" size="1"/>
|
||||
<delay_matrix type="rise" in_port="in" out_port="out">
|
||||
10e-12
|
||||
</delay_matrix>
|
||||
<delay_matrix type="fall" in_port="in" out_port="out">
|
||||
10e-12
|
||||
</delay_matrix>
|
||||
</circuit_model>
|
||||
<circuit_model type="inv_buf" name="sky130_fd_sc_hd__inv_2" prefix="sky130_fd_sc_hd__inv_2" is_default="false">
|
||||
<design_technology type="cmos" topology="buffer" size="1"/>
|
||||
<device_technology device_model_name="logic"/>
|
||||
<port type="input" prefix="in" lib_name="A" size="1"/>
|
||||
<port type="output" prefix="out" lib_name="Y" size="1"/>
|
||||
<delay_matrix type="rise" in_port="in" out_port="out">
|
||||
10e-12
|
||||
</delay_matrix>
|
||||
<delay_matrix type="fall" in_port="in" out_port="out">
|
||||
10e-12
|
||||
</delay_matrix>
|
||||
</circuit_model>
|
||||
<circuit_model type="gate" name="sky130_fd_sc_hd__or2_1" prefix="sky130_fd_sc_hd__or2_1" is_default="true">
|
||||
<design_technology type="cmos" topology="OR"/>
|
||||
<device_technology device_model_name="logic"/>
|
||||
<input_buffer exist="false"/>
|
||||
<output_buffer exist="false"/>
|
||||
<port type="input" prefix="a" lib_name="A" size="1"/>
|
||||
<port type="input" prefix="b" lib_name="B" size="1"/>
|
||||
<port type="output" prefix="out" lib_name="X" size="1"/>
|
||||
<delay_matrix type="rise" in_port="a b" out_port="out">
|
||||
10e-12 5e-12
|
||||
</delay_matrix>
|
||||
<delay_matrix type="fall" in_port="a b" out_port="out">
|
||||
10e-12 5e-12
|
||||
</delay_matrix>
|
||||
</circuit_model>
|
||||
<!-- Define a circuit model for the standard cell MUX2
|
||||
OpenFPGA requires the following truth table for the MUX2
|
||||
When the select signal sel is enabled, the first input, i.e., in0
|
||||
will be propagated to the output, i.e., out
|
||||
If your standard cell provider does not offer the exact truth table,
|
||||
you can simply swap the inputs as shown in the example below
|
||||
-->
|
||||
<circuit_model type="gate" name="sky130_fd_sc_hd__mux2_1" prefix="sky130_fd_sc_hd__mux2_1">
|
||||
<design_technology type="cmos" topology="MUX2"/>
|
||||
<device_technology device_model_name="logic"/>
|
||||
<input_buffer exist="false"/>
|
||||
<output_buffer exist="false"/>
|
||||
<port type="input" prefix="in0" lib_name="A1" size="1"/>
|
||||
<port type="input" prefix="in1" lib_name="A0" size="1"/>
|
||||
<port type="input" prefix="sel" lib_name="S" size="1"/>
|
||||
<port type="output" prefix="out" lib_name="X" size="1"/>
|
||||
</circuit_model>
|
||||
<circuit_model type="chan_wire" name="chan_segment" prefix="track_seg" is_default="true">
|
||||
<design_technology type="cmos"/>
|
||||
<input_buffer exist="false"/>
|
||||
<output_buffer exist="false"/>
|
||||
<port type="input" prefix="in" size="1"/>
|
||||
<port type="output" prefix="out" size="1"/>
|
||||
<wire_param model_type="pi" R="101" C="22.5e-15" num_level="1"/>
|
||||
<!-- model_type could be T, res_val and cap_val DON'T CARE -->
|
||||
</circuit_model>
|
||||
<circuit_model type="wire" name="direct_interc" prefix="direct_interc" is_default="true">
|
||||
<design_technology type="cmos"/>
|
||||
<input_buffer exist="false"/>
|
||||
<output_buffer exist="false"/>
|
||||
<port type="input" prefix="in" size="1"/>
|
||||
<port type="output" prefix="out" size="1"/>
|
||||
<wire_param model_type="pi" R="0" C="0" num_level="1"/>
|
||||
<!-- model_type could be T, res_val cap_val should be defined -->
|
||||
</circuit_model>
|
||||
<circuit_model type="mux" name="mux_tree" prefix="mux_tree" is_default="true" dump_structural_verilog="true">
|
||||
<design_technology type="cmos" structure="tree" add_const_input="true" const_input_val="1"/>
|
||||
<input_buffer exist="false"/>
|
||||
<output_buffer exist="false"/>
|
||||
<pass_gate_logic circuit_model_name="sky130_fd_sc_hd__mux2_1"/>
|
||||
<port type="input" prefix="in" size="1"/>
|
||||
<port type="output" prefix="out" size="1"/>
|
||||
<port type="sram" prefix="sram" size="1"/>
|
||||
</circuit_model>
|
||||
<circuit_model type="mux" name="mux_tree_tapbuf" prefix="mux_tree_tapbuf" dump_structural_verilog="true">
|
||||
<design_technology type="cmos" structure="tree" add_const_input="true" const_input_val="1"/>
|
||||
<input_buffer exist="false"/>
|
||||
<output_buffer exist="true" circuit_model_name="sky130_fd_sc_hd__buf_4"/>
|
||||
<pass_gate_logic circuit_model_name="sky130_fd_sc_hd__mux2_1"/>
|
||||
<port type="input" prefix="in" size="1"/>
|
||||
<port type="output" prefix="out" size="1"/>
|
||||
<port type="sram" prefix="sram" size="1"/>
|
||||
</circuit_model>
|
||||
<!--DFF subckt ports should be defined as <D> <Q> <CLK> <RESET> <SET> -->
|
||||
<circuit_model type="ff" name="SDFFRQ" prefix="SDFFRQ" verilog_netlist="${OPENFPGA_PATH}/openfpga_flow/openfpga_cell_library/verilog/dff.v">
|
||||
<design_technology type="cmos"/>
|
||||
<input_buffer exist="true" circuit_model_name="sky130_fd_sc_hd__inv_1"/>
|
||||
<output_buffer exist="true" circuit_model_name="sky130_fd_sc_hd__inv_1"/>
|
||||
<port type="input" prefix="D" size="1"/>
|
||||
<port type="input" prefix="DI" lib_name="SI" size="1"/>
|
||||
<port type="input" prefix="Test_en" lib_name="SE" size="1" is_global="true" default_val="0"/>
|
||||
<port type="input" prefix="reset" lib_name="RST" size="1" default_val="0"/>
|
||||
<port type="output" prefix="Q" size="1"/>
|
||||
<port type="clock" prefix="clk" lib_name="CK" size="1" default_val="0" />
|
||||
</circuit_model>
|
||||
<circuit_model type="lut" name="frac_lut4" prefix="frac_lut4" dump_structural_verilog="true">
|
||||
<design_technology type="cmos" fracturable_lut="true"/>
|
||||
<input_buffer exist="false"/>
|
||||
<output_buffer exist="true" circuit_model_name="sky130_fd_sc_hd__buf_2"/>
|
||||
<lut_input_inverter exist="true" circuit_model_name="sky130_fd_sc_hd__inv_1"/>
|
||||
<lut_input_buffer exist="true" circuit_model_name="sky130_fd_sc_hd__buf_2"/>
|
||||
<lut_intermediate_buffer exist="true" circuit_model_name="sky130_fd_sc_hd__buf_2" location_map="-1-"/>
|
||||
<pass_gate_logic circuit_model_name="sky130_fd_sc_hd__mux2_1"/>
|
||||
<port type="input" prefix="in" size="4" tri_state_map="---1" circuit_model_name="sky130_fd_sc_hd__or2_1"/>
|
||||
<port type="output" prefix="lut2_out" size="2" lut_frac_level="2" lut_output_mask="2,3"/>
|
||||
<port type="output" prefix="lut3_out" size="2" lut_frac_level="3" lut_output_mask="0,1"/>
|
||||
<port type="output" prefix="lut4_out" size="1" lut_output_mask="0"/>
|
||||
<port type="sram" prefix="sram" size="16"/>
|
||||
<port type="sram" prefix="mode" size="1" mode_select="true" circuit_model_name="DFFRQ" default_val="1"/>
|
||||
</circuit_model>
|
||||
<!--Scan-chain DFF subckt ports should be defined as <D> <Q> <Qb> <CLK> <RESET> <SET> -->
|
||||
<circuit_model type="ccff" name="DFFRQ" prefix="DFFRQ" verilog_netlist="${OPENFPGA_PATH}/openfpga_flow/openfpga_cell_library/verilog/dff.v">
|
||||
<design_technology type="cmos"/>
|
||||
<input_buffer exist="true" circuit_model_name="sky130_fd_sc_hd__inv_1"/>
|
||||
<output_buffer exist="true" circuit_model_name="sky130_fd_sc_hd__inv_1"/>
|
||||
<port type="input" prefix="D" size="1"/>
|
||||
<port type="output" prefix="Q" size="1"/>
|
||||
<port type="clock" prefix="prog_clk" lib_name="CK" size="1" is_global="true" default_val="0" is_prog="true"/>
|
||||
<port type="input" prefix="pReset" lib_name="RST" size="1" is_global="true" default_val="0" is_prog="true" is_reset="true"/>
|
||||
</circuit_model>
|
||||
<circuit_model type="iopad" name="EMBEDDED_IO_ISOLN" prefix="EMBEDDED_IO_ISOLN" is_default="true" verilog_netlist="${OPENFPGA_PATH}/openfpga_flow/openfpga_cell_library/verilog/gpio.v">
|
||||
<design_technology type="cmos"/>
|
||||
<input_buffer exist="true" circuit_model_name="sky130_fd_sc_hd__inv_1"/>
|
||||
<output_buffer exist="true" circuit_model_name="sky130_fd_sc_hd__inv_1"/>
|
||||
<port type="input" prefix="SOC_IN" lib_name="SOC_IN" size="1" is_global="true" is_io="true" is_data_io="true"/>
|
||||
<port type="output" prefix="SOC_OUT" lib_name="SOC_OUT" size="1" is_global="true" is_io="true" is_data_io="true"/>
|
||||
<port type="output" prefix="SOC_DIR" lib_name="SOC_DIR" size="1" is_global="true" is_io="true"/>
|
||||
<port type="input" prefix="IO_ISOL_N" lib_name="IO_ISOL_N" size="1" is_global="true" default_val="1"/>
|
||||
<port type="output" prefix="inpad" lib_name="FPGA_IN" size="1"/>
|
||||
<port type="input" prefix="outpad" lib_name="FPGA_OUT" size="1"/>
|
||||
<port type="sram" prefix="en" lib_name="FPGA_DIR" size="1" mode_select="true" circuit_model_name="DFFRQ" default_val="1"/>
|
||||
</circuit_model>
|
||||
<circuit_model type="hard_logic" name="CARRY_MUX2" prefix="CARRY_MUX2" verilog_netlist="${OPENFPGA_PATH}/openfpga_flow/openfpga_cell_library/verilog/mux2.v">
|
||||
<design_technology type="cmos"/>
|
||||
<device_technology device_model_name="logic"/>
|
||||
<input_buffer exist="false"/>
|
||||
<output_buffer exist="false"/>
|
||||
<port type="input" prefix="a" lib_name="A0" size="1"/>
|
||||
<port type="input" prefix="b" lib_name="A1" size="1"/>
|
||||
<port type="input" prefix="cin" lib_name="S" size="1"/>
|
||||
<port type="output" prefix="cout" lib_name="Y" size="1"/>
|
||||
</circuit_model>
|
||||
<circuit_model type="hard_logic" name="frac_mult_16x16" prefix="frac_mult_16x16" is_default="true" spice_netlist="${OPENFPGA_PATH}/openfpga_flow/openfpga_cell_library/spice/frac_mult_16x16.sp" verilog_netlist="${OPENFPGA_PATH}/openfpga_flow/openfpga_cell_library/verilog/frac_mult_16x16.v">
|
||||
<design_technology type="cmos"/>
|
||||
<input_buffer exist="true" circuit_model_name="INVTX1"/>
|
||||
<output_buffer exist="true" circuit_model_name="INVTX1"/>
|
||||
<port type="input" prefix="A" lib_name="a" size="16"/>
|
||||
<port type="input" prefix="B" lib_name="b" size="16"/>
|
||||
<port type="output" prefix="Y" lib_name="out" size="32"/>
|
||||
<port type="sram" prefix="mode" lib_name="mode" size="1" mode_select="true" circuit_model_name="DFFRQ" default_val="0"/>
|
||||
</circuit_model>
|
||||
</circuit_library>
|
||||
<configuration_protocol>
|
||||
<organization type="scan_chain" circuit_model_name="DFFRQ" num_regions="1"/>
|
||||
</configuration_protocol>
|
||||
<connection_block>
|
||||
<switch name="ipin_cblock" circuit_model_name="mux_tree_tapbuf"/>
|
||||
</connection_block>
|
||||
<switch_block>
|
||||
<switch name="L1_mux" circuit_model_name="mux_tree_tapbuf"/>
|
||||
<switch name="L2_mux" circuit_model_name="mux_tree_tapbuf"/>
|
||||
<switch name="L4_mux" circuit_model_name="mux_tree_tapbuf"/>
|
||||
</switch_block>
|
||||
<routing_segment>
|
||||
<segment name="L1" circuit_model_name="chan_segment"/>
|
||||
<segment name="L2" circuit_model_name="chan_segment"/>
|
||||
<segment name="L4" circuit_model_name="chan_segment"/>
|
||||
</routing_segment>
|
||||
<direct_connection>
|
||||
<direct name="carry_chain" circuit_model_name="direct_interc"/>
|
||||
<direct name="shift_register" circuit_model_name="direct_interc"/>
|
||||
<direct name="scan_chain" circuit_model_name="direct_interc" type="column" x_dir="positive" y_dir="positive"/>
|
||||
</direct_connection>
|
||||
<tile_annotations>
|
||||
<global_port name="clk" is_clock="true" default_val="0">
|
||||
<tile name="clb" port="clk" x="-1" y="-1"/>
|
||||
</global_port>
|
||||
<global_port name="Reset" is_reset="true" default_val="1">
|
||||
<tile name="clb" port="reset" x="-1" y="-1"/>
|
||||
</global_port>
|
||||
</tile_annotations>
|
||||
<pb_type_annotations>
|
||||
<!-- physical pb_type binding in complex block IO -->
|
||||
<pb_type name="io" physical_mode_name="physical" idle_mode_name="inpad"/>
|
||||
<!-- IMPORTANT: must set unused I/Os to operating in INPUT mode !!! -->
|
||||
<pb_type name="io[physical].iopad" circuit_model_name="EMBEDDED_IO_ISOLN" mode_bits="1"/>
|
||||
<pb_type name="io[inpad].inpad" physical_pb_type_name="io[physical].iopad" mode_bits="1"/>
|
||||
<pb_type name="io[outpad].outpad" physical_pb_type_name="io[physical].iopad" mode_bits="0"/>
|
||||
<!-- End physical pb_type binding in complex block IO -->
|
||||
|
||||
<!-- physical pb_type binding in complex block CLB -->
|
||||
<!-- physical mode will be the default mode if not specified -->
|
||||
<pb_type name="clb.fle" physical_mode_name="physical"/>
|
||||
<pb_type name="clb.fle[physical].fabric.frac_logic.frac_lut4" circuit_model_name="frac_lut4" mode_bits="0"/>
|
||||
<pb_type name="clb.fle[physical].fabric.frac_logic.carry_follower" circuit_model_name="CARRY_MUX2"/>
|
||||
<pb_type name="clb.fle[physical].fabric.ff" circuit_model_name="SDFFRQ"/>
|
||||
<!-- Binding operating pb_type to physical pb_type -->
|
||||
<!-- Binding operating pb_types in mode 'arithmetic' -->
|
||||
<pb_type name="clb.fle[arithmetic].soft_adder.adder_lut4" physical_pb_type_name="clb.fle[physical].fabric.frac_logic.frac_lut4" mode_bits="1"/>
|
||||
<pb_type name="clb.fle[arithmetic].soft_adder.carry_follower" physical_pb_type_name="clb.fle[physical].fabric.frac_logic.carry_follower"/>
|
||||
<!-- Binding operating pb_types in mode 'n2_lut3' -->
|
||||
<pb_type name="clb.fle[n2_lut3].lut3inter.ble3.lut3" physical_pb_type_name="clb.fle[physical].fabric.frac_logic.frac_lut4" mode_bits="1" physical_pb_type_index_factor="0.5">
|
||||
<!-- Binding the lut3 to the first 3 inputs of fracturable lut4 -->
|
||||
<port name="in" physical_mode_port="in[0:2]"/>
|
||||
<port name="out" physical_mode_port="lut3_out[0:0]" physical_mode_pin_rotate_offset="1"/>
|
||||
</pb_type>
|
||||
<pb_type name="clb.fle[n2_lut3].lut3inter.ble3.ff" physical_pb_type_name="clb.fle[physical].fabric.ff"/>
|
||||
<!-- Binding operating pb_types in mode 'ble4' -->
|
||||
<pb_type name="clb.fle[n1_lut4].ble4.lut4" physical_pb_type_name="clb.fle[physical].fabric.frac_logic.frac_lut4" mode_bits="0">
|
||||
<!-- Binding the lut4 to the first 4 inputs of fracturable lut4 -->
|
||||
<port name="in" physical_mode_port="in[0:3]"/>
|
||||
<port name="out" physical_mode_port="lut4_out"/>
|
||||
</pb_type>
|
||||
<pb_type name="clb.fle[n1_lut4].ble4.ff" physical_pb_type_name="clb.fle[physical].fabric.ff" physical_pb_type_index_factor="2" physical_pb_type_index_offset="0"/>
|
||||
<!-- Binding operating pb_types in mode 'shift_register' -->
|
||||
<pb_type name="clb.fle[shift_register].shift_reg.ff" physical_pb_type_name="clb.fle[physical].fabric.ff"/>
|
||||
<!-- End physical pb_type binding in complex block IO -->
|
||||
|
||||
<!-- physical pb_type binding in complex block dsp -->
|
||||
<pb_type name="mult_16" physical_mode_name="mult_16x16"/>
|
||||
<!-- Bind the primitive pb_type in the physical mode to a circuit model -->
|
||||
<pb_type name="mult_16[mult_16x16].mult_16x16_slice.mult_16x16" circuit_model_name="frac_mult_16x16" mode_bits="0"/>
|
||||
<pb_type name="mult_16[mult_8x8].mult_8x8_slice.mult_8x8" physical_pb_type_name="mult_16[mult_16x16].mult_16x16_slice.mult_16x16" mode_bits="1" physical_pb_type_index_factor="0">
|
||||
<port name="A" physical_mode_port="A[0:7]" physical_mode_port_rotate_offset="8"/>
|
||||
<port name="B" physical_mode_port="B[0:7]" physical_mode_port_rotate_offset="8"/>
|
||||
<port name="Y" physical_mode_port="Y[0:15]" physical_mode_port_rotate_offset="16"/>
|
||||
</pb_type>
|
||||
</pb_type_annotations>
|
||||
</openfpga_architecture>
|
|
@ -0,0 +1,29 @@
|
|||
//-----------------------------------------------------
|
||||
// Design Name : frac_mult_16x16
|
||||
// File Name : frac_mult_16x16.v
|
||||
// Function : A 16-bit multiplier which can operate in fracturable modes:
|
||||
// 1. two 8-bit multipliers
|
||||
// 2. one 16-bit multipliers
|
||||
// Coder : Xifan Tang
|
||||
//-----------------------------------------------------
|
||||
|
||||
module frac_mult_16x16 (
|
||||
input [0:15] a,
|
||||
input [0:15] b,
|
||||
output [0:31] out,
|
||||
input [0:0] mode);
|
||||
|
||||
reg [0:31] out_reg;
|
||||
|
||||
always @(mode, a, b) begin
|
||||
if (1'b1 == mode) begin
|
||||
out_reg[0:15] <= a[0:7] * b[0:7];
|
||||
out_reg[16:31] <= a[8:15] * b[8:15];
|
||||
end else begin
|
||||
out_reg <= a * b;
|
||||
end
|
||||
end
|
||||
|
||||
assign out = out_reg;
|
||||
|
||||
endmodule
|
|
@ -1,6 +1,6 @@
|
|||
# Run VPR for the 'and' design
|
||||
#--write_rr_graph example_rr_graph.xml
|
||||
vpr ${VPR_ARCH_FILE} ${VPR_TESTBENCH_BLIF} --clock_modeling route --device ${OPENFPGA_VPR_DEVICE_LAYOUT}
|
||||
vpr ${VPR_ARCH_FILE} ${VPR_TESTBENCH_BLIF} --device ${OPENFPGA_VPR_DEVICE_LAYOUT} --constant_net_method route
|
||||
|
||||
# Read OpenFPGA architecture definition
|
||||
read_openfpga_arch -f ${OPENFPGA_ARCH_FILE}
|
||||
|
|
|
@ -0,0 +1,25 @@
|
|||
//-----------------------------
|
||||
// 8-bit multiplier
|
||||
//-----------------------------
|
||||
module mult_8(
|
||||
input [0:7] A,
|
||||
input [0:7] B,
|
||||
output [0:15] Y
|
||||
);
|
||||
|
||||
assign Y = A * B;
|
||||
|
||||
endmodule
|
||||
|
||||
//-----------------------------
|
||||
// 16-bit multiplier
|
||||
//-----------------------------
|
||||
module mult_16(
|
||||
input [0:15] A,
|
||||
input [0:15] B,
|
||||
output [0:31] Y
|
||||
);
|
||||
|
||||
assign Y = A * B;
|
||||
|
||||
endmodule
|
|
@ -0,0 +1,41 @@
|
|||
//-----------------------------
|
||||
// 8-bit multiplier
|
||||
//-----------------------------
|
||||
module mult_8x8 (
|
||||
input [0:7] A,
|
||||
input [0:7] B,
|
||||
output [0:15] Y
|
||||
);
|
||||
parameter A_SIGNED = 0;
|
||||
parameter B_SIGNED = 0;
|
||||
parameter A_WIDTH = 0;
|
||||
parameter B_WIDTH = 0;
|
||||
parameter Y_WIDTH = 0;
|
||||
|
||||
mult_8 #() _TECHMAP_REPLACE_ (
|
||||
.A (A),
|
||||
.B (B),
|
||||
.Y (Y) );
|
||||
|
||||
endmodule
|
||||
|
||||
//-----------------------------
|
||||
// 16-bit multiplier
|
||||
//-----------------------------
|
||||
module mult_16x16 (
|
||||
input [0:15] A,
|
||||
input [0:15] B,
|
||||
output [0:31] Y
|
||||
);
|
||||
parameter A_SIGNED = 0;
|
||||
parameter B_SIGNED = 0;
|
||||
parameter A_WIDTH = 0;
|
||||
parameter B_WIDTH = 0;
|
||||
parameter Y_WIDTH = 0;
|
||||
|
||||
mult_16 #() _TECHMAP_REPLACE_ (
|
||||
.A (A),
|
||||
.B (B),
|
||||
.Y (Y) );
|
||||
|
||||
endmodule
|
|
@ -47,6 +47,9 @@ run-task fpga_verilog/bram/wide_dpram16k --debug --show_thread_logs
|
|||
echo -e "Testing Verilog generation with heterogeneous fabric using 8-bit single-mode multipliers ";
|
||||
run-task fpga_verilog/dsp/single_mode_mult_8x8 --debug --show_thread_logs
|
||||
|
||||
echo -e "Testing Verilog generation with heterogeneous fabric using 16-bit multi-mode multipliers ";
|
||||
run-task fpga_verilog/dsp/multi_mode_mult_16x16 --debug --show_thread_logs
|
||||
|
||||
echo -e "Testing Verilog generation with different I/O capacities on each side of an FPGA ";
|
||||
run-task fpga_verilog/io/multi_io_capacity --debug --show_thread_logs
|
||||
|
||||
|
|
|
@ -0,0 +1,49 @@
|
|||
# = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
|
||||
# Configuration file for running experiments
|
||||
# = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
|
||||
# timeout_each_job : FPGA Task script splits fpga flow into multiple jobs
|
||||
# Each job execute fpga_flow script on combination of architecture & benchmark
|
||||
# timeout_each_job is timeout for each job
|
||||
# = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
|
||||
|
||||
[GENERAL]
|
||||
run_engine=openfpga_shell
|
||||
power_tech_file = ${PATH:OPENFPGA_PATH}/openfpga_flow/tech/PTM_45nm/45nm.xml
|
||||
power_analysis = false
|
||||
spice_output=false
|
||||
verilog_output=true
|
||||
timeout_each_job = 20*60
|
||||
fpga_flow=yosys_vpr
|
||||
|
||||
[OpenFPGA_SHELL]
|
||||
openfpga_shell_template=${PATH:OPENFPGA_PATH}/openfpga_flow/openfpga_shell_scripts/fix_heterogeneous_device_example_script.openfpga
|
||||
openfpga_arch_file=${PATH:OPENFPGA_PATH}/openfpga_flow/openfpga_arch/k4_frac_N8_reset_softadder_register_scan_chain_frac_dsp16_caravel_io_skywater130nm_fdhd_cc_openfpga.xml
|
||||
openfpga_sim_setting_file=${PATH:OPENFPGA_PATH}/openfpga_flow/openfpga_simulation_settings/fixed_sim_openfpga.xml
|
||||
# Yosys script parameters
|
||||
yosys_cell_sim_verilog=${PATH:OPENFPGA_PATH}/openfpga_flow/openfpga_yosys_techlib/k4_frac_N8_tileable_reset_softadder_register_scan_chain_frac_dsp16_nonLR_caravel_io_skywater130nm_cell_sim.v
|
||||
yosys_dsp_map_verilog=${PATH:OPENFPGA_PATH}/openfpga_flow/openfpga_yosys_techlib/k4_frac_N8_tileable_reset_softadder_register_scan_chain_frac_dsp16_nonLR_caravel_io_skywater130nm_dsp_map.v
|
||||
yosys_dsp_map_parameters=-D DSP_A_MAXWIDTH=8 -D DSP_B_MAXWIDTH=8 -D DSP_A_MINWIDTH=2 -D DSP_B_MINWIDTH=2 -D DSP_NAME=mult_8x8
|
||||
# VPR parameter
|
||||
openfpga_vpr_device_layout=3x4
|
||||
|
||||
[ARCHITECTURES]
|
||||
arch0=${PATH:OPENFPGA_PATH}/openfpga_flow/vpr_arch/k4_frac_N8_tileable_reset_softadder_register_scan_chain_frac_dsp16_nonLR_caravel_io_skywater130nm.xml
|
||||
|
||||
[BENCHMARKS]
|
||||
bench0=${PATH:OPENFPGA_PATH}/openfpga_flow/benchmarks/micro_benchmark/mac/mac_4/mac_4.v
|
||||
bench1=${PATH:OPENFPGA_PATH}/openfpga_flow/benchmarks/micro_benchmark/mac/mac_8/mac_8.v
|
||||
bench2=${PATH:OPENFPGA_PATH}/openfpga_flow/benchmarks/micro_benchmark/mac/mac_12/mac_12.v
|
||||
bench3=${PATH:OPENFPGA_PATH}/openfpga_flow/benchmarks/micro_benchmark/mac/mac_16/mac_16.v
|
||||
|
||||
[SYNTHESIS_PARAM]
|
||||
bench_yosys_common=${PATH:OPENFPGA_PATH}/openfpga_flow/misc/ys_tmpl_yosys_vpr_dsp_flow.ys
|
||||
bench_yosys_rewrite_common=${PATH:OPENFPGA_PATH}/openfpga_flow/misc/ys_tmpl_yosys_vpr_flow_with_rewrite.ys;${PATH:OPENFPGA_PATH}/openfpga_flow/misc/ys_tmpl_rewrite_flow.ys
|
||||
|
||||
bench0_top = mac_4
|
||||
bench1_top = mac_8
|
||||
bench2_top = mac_12
|
||||
bench3_top = mac_16
|
||||
|
||||
[SCRIPT_PARAM_MIN_ROUTE_CHAN_WIDTH]
|
||||
end_flow_with_test=
|
||||
vpr_fpga_verilog_formal_verification_top_netlist=
|
|
@ -30,13 +30,19 @@ openfpga_vpr_device_layout=3x2
|
|||
arch0=${PATH:OPENFPGA_PATH}/openfpga_flow/vpr_arch/k4_frac_N8_tileable_reset_softadder_register_scan_chain_dsp8_nonLR_caravel_io_skywater130nm.xml
|
||||
|
||||
[BENCHMARKS]
|
||||
bench0=${PATH:OPENFPGA_PATH}/openfpga_flow/benchmarks/micro_benchmark/mac_8/mac_8.v
|
||||
bench0=${PATH:OPENFPGA_PATH}/openfpga_flow/benchmarks/micro_benchmark/mac/mac_2/mac_2.v
|
||||
bench1=${PATH:OPENFPGA_PATH}/openfpga_flow/benchmarks/micro_benchmark/mac/mac_4/mac_4.v
|
||||
bench2=${PATH:OPENFPGA_PATH}/openfpga_flow/benchmarks/micro_benchmark/mac/mac_6/mac_6.v
|
||||
bench3=${PATH:OPENFPGA_PATH}/openfpga_flow/benchmarks/micro_benchmark/mac/mac_8/mac_8.v
|
||||
|
||||
[SYNTHESIS_PARAM]
|
||||
bench_yosys_common=${PATH:OPENFPGA_PATH}/openfpga_flow/misc/ys_tmpl_yosys_vpr_dsp_flow.ys
|
||||
bench_yosys_rewrite_common=${PATH:OPENFPGA_PATH}/openfpga_flow/misc/ys_tmpl_yosys_vpr_flow_with_rewrite.ys;${PATH:OPENFPGA_PATH}/openfpga_flow/misc/ys_tmpl_rewrite_flow.ys
|
||||
|
||||
bench0_top = mac_8
|
||||
bench0_top = mac_2
|
||||
bench1_top = mac_4
|
||||
bench2_top = mac_6
|
||||
bench3_top = mac_8
|
||||
|
||||
[SCRIPT_PARAM_MIN_ROUTE_CHAN_WIDTH]
|
||||
end_flow_with_test=
|
||||
|
|
|
@ -0,0 +1,964 @@
|
|||
<!--
|
||||
Low-cost homogeneous FPGA Architecture.
|
||||
|
||||
- Skywater 130 nm technology
|
||||
- General purpose logic block:
|
||||
K = 4, N = 8, fracturable 4 LUTs (can operate as one 4-LUT or two 3-LUTs with all 3 inputs shared)
|
||||
with optionally registered outputs
|
||||
- Heterogeneous block
|
||||
8-bit multiplier
|
||||
- Routing architecture:
|
||||
- 10% L = 1, fc_in = 0.15, Fc_out = 0.10
|
||||
- 10% L = 2, fc_in = 0.15, Fc_out = 0.10
|
||||
- 80% L = 4, fc_in = 0.15, Fc_out = 0.10
|
||||
- 100 routing tracks per channel
|
||||
|
||||
Authors: Xifan Tang
|
||||
-->
|
||||
<architecture>
|
||||
<!--
|
||||
ODIN II specific config begins
|
||||
Describes the types of user-specified netlist blocks (in blif, this corresponds to
|
||||
".model [type_of_block]") that this architecture supports.
|
||||
|
||||
Note: Basic LUTs, I/Os, and flip-flops are not included here as there are
|
||||
already special structures in blif (.names, .input, .output, and .latch)
|
||||
that describe them.
|
||||
-->
|
||||
<models>
|
||||
<model name="mult_8">
|
||||
<input_ports>
|
||||
<port name="A" combinational_sink_ports="Y"/>
|
||||
<port name="B" combinational_sink_ports="Y"/>
|
||||
</input_ports>
|
||||
<output_ports>
|
||||
<port name="Y"/>
|
||||
</output_ports>
|
||||
</model>
|
||||
<model name="mult_16">
|
||||
<input_ports>
|
||||
<port name="A" combinational_sink_ports="Y"/>
|
||||
<port name="B" combinational_sink_ports="Y"/>
|
||||
</input_ports>
|
||||
<output_ports>
|
||||
<port name="Y"/>
|
||||
</output_ports>
|
||||
</model>
|
||||
<!-- A virtual model for I/O to be used in the physical mode of io block -->
|
||||
<model name="io">
|
||||
<input_ports>
|
||||
<port name="outpad"/>
|
||||
</input_ports>
|
||||
<output_ports>
|
||||
<port name="inpad"/>
|
||||
</output_ports>
|
||||
</model>
|
||||
<model name="adder_lut4">
|
||||
<input_ports>
|
||||
<port name="in" combinational_sink_ports="lut2_out lut4_out"/>
|
||||
</input_ports>
|
||||
<output_ports>
|
||||
<port name="lut2_out"/>
|
||||
<port name="lut4_out"/>
|
||||
</output_ports>
|
||||
</model>
|
||||
<model name="carry_follower">
|
||||
<input_ports>
|
||||
<port name="a" combinational_sink_ports="cout"/>
|
||||
<port name="b" combinational_sink_ports="cout"/>
|
||||
<port name="cin" combinational_sink_ports="cout"/>
|
||||
</input_ports>
|
||||
<output_ports>
|
||||
<port name="cout"/>
|
||||
</output_ports>
|
||||
</model>
|
||||
<model name="frac_lut4">
|
||||
<input_ports>
|
||||
<port name="in"/>
|
||||
</input_ports>
|
||||
<output_ports>
|
||||
<port name="lut2_out"/>
|
||||
<port name="lut3_out"/>
|
||||
<port name="lut4_out"/>
|
||||
</output_ports>
|
||||
</model>
|
||||
<model name="carry_follower_physical">
|
||||
<input_ports>
|
||||
<port name="a" combinational_sink_ports="cout"/>
|
||||
<port name="b" combinational_sink_ports="cout"/>
|
||||
<port name="cin" combinational_sink_ports="cout"/>
|
||||
</input_ports>
|
||||
<output_ports>
|
||||
<port name="cout"/>
|
||||
</output_ports>
|
||||
</model>
|
||||
<!-- A virtual model for scan-chain flip-flop to be used in the physical mode of FF -->
|
||||
<model name="scff">
|
||||
<input_ports>
|
||||
<port name="D" clock="clk"/>
|
||||
<port name="DI" clock="clk"/>
|
||||
<port name="reset" clock="clk"/>
|
||||
<port name="clk" is_clock="1"/>
|
||||
</input_ports>
|
||||
<output_ports>
|
||||
<port name="Q" clock="clk"/>
|
||||
</output_ports>
|
||||
</model>
|
||||
</models>
|
||||
<tiles>
|
||||
<!-- Do NOT add clock pins to I/O here!!! VPR does not build clock network in the way that OpenFPGA can support
|
||||
If you need to register the I/O, define clocks in the circuit models
|
||||
These clocks can be handled in back-end
|
||||
-->
|
||||
<!-- Top-side has 1 I/O per tile -->
|
||||
<tile name="io_top" capacity="9" area="0">
|
||||
<equivalent_sites>
|
||||
<site pb_type="io"/>
|
||||
</equivalent_sites>
|
||||
<input name="outpad" num_pins="1"/>
|
||||
<output name="inpad" num_pins="1"/>
|
||||
<fc in_type="frac" in_val="0.15" out_type="frac" out_val="0.10"/>
|
||||
<pinlocations pattern="custom">
|
||||
<loc side="bottom">io_top.outpad io_top.inpad</loc>
|
||||
</pinlocations>
|
||||
</tile>
|
||||
<!-- Right-side has 1 I/O per tile -->
|
||||
<tile name="io_right" capacity="9" area="0">
|
||||
<equivalent_sites>
|
||||
<site pb_type="io"/>
|
||||
</equivalent_sites>
|
||||
<input name="outpad" num_pins="1"/>
|
||||
<output name="inpad" num_pins="1"/>
|
||||
<fc in_type="frac" in_val="0.15" out_type="frac" out_val="0.10"/>
|
||||
<pinlocations pattern="custom">
|
||||
<loc side="left">io_right.outpad io_right.inpad</loc>
|
||||
</pinlocations>
|
||||
</tile>
|
||||
<!-- Bottom-side has 9 I/O per tile -->
|
||||
<tile name="io_bottom" capacity="9" area="0">
|
||||
<equivalent_sites>
|
||||
<site pb_type="io"/>
|
||||
</equivalent_sites>
|
||||
<input name="outpad" num_pins="1"/>
|
||||
<output name="inpad" num_pins="1"/>
|
||||
<fc in_type="frac" in_val="0.15" out_type="frac" out_val="0.10"/>
|
||||
<pinlocations pattern="custom">
|
||||
<loc side="top">io_bottom.outpad io_bottom.inpad</loc>
|
||||
</pinlocations>
|
||||
</tile>
|
||||
<!-- Left-side has 1 I/O per tile -->
|
||||
<tile name="io_left" capacity="9" area="0">
|
||||
<equivalent_sites>
|
||||
<site pb_type="io"/>
|
||||
</equivalent_sites>
|
||||
<input name="outpad" num_pins="1"/>
|
||||
<output name="inpad" num_pins="1"/>
|
||||
<fc in_type="frac" in_val="0.15" out_type="frac" out_val="0.10"/>
|
||||
<pinlocations pattern="custom">
|
||||
<loc side="right">io_left.outpad io_left.inpad</loc>
|
||||
</pinlocations>
|
||||
</tile>
|
||||
<!-- CLB has most pins on the top and right sides -->
|
||||
<tile name="clb" area="53894">
|
||||
<equivalent_sites>
|
||||
<site pb_type="clb"/>
|
||||
</equivalent_sites>
|
||||
<input name="I0" num_pins="2" equivalent="full"/>
|
||||
<input name="I0i" num_pins="2" equivalent="none"/>
|
||||
<input name="I1" num_pins="2" equivalent="full"/>
|
||||
<input name="I1i" num_pins="2" equivalent="none"/>
|
||||
<input name="I2" num_pins="2" equivalent="full"/>
|
||||
<input name="I2i" num_pins="2" equivalent="none"/>
|
||||
<input name="I3" num_pins="2" equivalent="full"/>
|
||||
<input name="I3i" num_pins="2" equivalent="none"/>
|
||||
<input name="I4" num_pins="2" equivalent="full"/>
|
||||
<input name="I4i" num_pins="2" equivalent="none"/>
|
||||
<input name="I5" num_pins="2" equivalent="full"/>
|
||||
<input name="I5i" num_pins="2" equivalent="none"/>
|
||||
<input name="I6" num_pins="2" equivalent="full"/>
|
||||
<input name="I6i" num_pins="2" equivalent="none"/>
|
||||
<input name="I7" num_pins="2" equivalent="full"/>
|
||||
<input name="I7i" num_pins="2" equivalent="none"/>
|
||||
<input name="reg_in" num_pins="1"/>
|
||||
<input name="sc_in" num_pins="1"/>
|
||||
<input name="cin" num_pins="1"/>
|
||||
<input name="reset" num_pins="1" is_non_clock_global="true"/>
|
||||
<output name="O" num_pins="16" equivalent="none"/>
|
||||
<output name="reg_out" num_pins="1"/>
|
||||
<output name="sc_out" num_pins="1"/>
|
||||
<output name="cout" num_pins="1"/>
|
||||
<clock name="clk" num_pins="1"/>
|
||||
<fc in_type="frac" in_val="0.15" out_type="frac" out_val="0.10">
|
||||
<fc_override port_name="reg_in" fc_type="frac" fc_val="0"/>
|
||||
<fc_override port_name="reg_out" fc_type="frac" fc_val="0"/>
|
||||
<fc_override port_name="sc_in" fc_type="frac" fc_val="0"/>
|
||||
<fc_override port_name="sc_out" fc_type="frac" fc_val="0"/>
|
||||
<fc_override port_name="cin" fc_type="frac" fc_val="0"/>
|
||||
<fc_override port_name="cout" fc_type="frac" fc_val="0"/>
|
||||
<fc_override port_name="clk" fc_type="frac" fc_val="0"/>
|
||||
<fc_override port_name="reset" fc_type="frac" fc_val="0"/>
|
||||
</fc>
|
||||
<!--pinlocations pattern="spread"/-->
|
||||
<pinlocations pattern="custom">
|
||||
<loc side="left">clb.clk clb.reset</loc>
|
||||
<loc side="top">clb.reg_in clb.sc_in clb.cin clb.O[7:0] clb.I0 clb.I0i clb.I1 clb.I1i clb.I2 clb.I2i clb.I3 clb.I3i</loc>
|
||||
<loc side="right">clb.O[15:8] clb.I4 clb.I4i clb.I5 clb.I5i clb.I6 clb.I6i clb.I7 clb.I7i</loc>
|
||||
<loc side="bottom">clb.reg_out clb.sc_out clb.cout</loc>
|
||||
</pinlocations>
|
||||
</tile>
|
||||
<tile name="mult_16" height="2" area="396000">
|
||||
<equivalent_sites>
|
||||
<site pb_type="mult_16" pin_mapping="direct"/>
|
||||
</equivalent_sites>
|
||||
<input name="a" num_pins="16"/>
|
||||
<input name="b" num_pins="16"/>
|
||||
<output name="out" num_pins="32"/>
|
||||
<fc in_type="frac" in_val="0.15" out_type="frac" out_val="0.10"/>
|
||||
<!-- Highly recommand to customize pin location when direct connection is used!!! -->
|
||||
<!--pinlocations pattern="spread"/-->
|
||||
<pinlocations pattern="custom">
|
||||
<loc side="left" yoffset="0">mult_16.a[0:2] mult_16.b[0:2] mult_16.out[0:5]</loc>
|
||||
<loc side="left" yoffset="1">mult_16.a[3:5] mult_16.b[3:5] mult_16.out[6:10]</loc>
|
||||
<loc side="top"></loc>
|
||||
<loc side="right" yoffset="0">mult_16.a[8:10] mult_16.b[8:10] mult_16.out[16:21]</loc>
|
||||
<loc side="right" yoffset="1">mult_16.a[11:13] mult_16.b[11:13] mult_16.out[22:26]</loc>
|
||||
<loc side="bottom">mult_16.a[6:7] mult_16.b[6:7] mult_16.out[11:15] mult_16.a[14:15] mult_16.b[14:15] mult_16.out[27:31]</loc>
|
||||
</pinlocations>
|
||||
</tile>
|
||||
</tiles>
|
||||
<!-- ODIN II specific config ends -->
|
||||
<!-- Physical descriptions begin -->
|
||||
<layout tileable="true">
|
||||
<auto_layout aspect_ratio="1.0">
|
||||
<!--Perimeter of 'io' blocks with 'EMPTY' blocks at corners-->
|
||||
<row type="io_top" starty="H-1" priority="100"/>
|
||||
<row type="io_bottom" starty="0" priority="100"/>
|
||||
<col type="io_left" startx="0" priority="100"/>
|
||||
<col type="io_right" startx="W-1" priority="100"/>
|
||||
<corners type="EMPTY" priority="101"/>
|
||||
<!--Fill with 'clb'-->
|
||||
<fill type="clb" priority="10"/>
|
||||
<!--Column of 'mult_8' with 'EMPTY' blocks wherever a 'mult_8' does not fit. Vertical offset by 1 for perimeter.-->
|
||||
<col type="mult_16" startx="2" starty="1" repeatx="8" priority="20"/>
|
||||
</auto_layout>
|
||||
<fixed_layout name="3x4" width="5" height="6">
|
||||
<!--Perimeter of 'io' blocks with 'EMPTY' blocks at corners-->
|
||||
<row type="io_top" starty="H-1" priority="100"/>
|
||||
<row type="io_bottom" starty="0" priority="100"/>
|
||||
<col type="io_left" startx="0" priority="100"/>
|
||||
<col type="io_right" startx="W-1" priority="100"/>
|
||||
<corners type="EMPTY" priority="101"/>
|
||||
<!--Fill with 'clb'-->
|
||||
<fill type="clb" priority="10"/>
|
||||
<!--Column of 'mult_8' with 'EMPTY' blocks wherever a 'mult_8' does not fit. Vertical offset by 1 for perimeter.-->
|
||||
<col type="mult_16" startx="2" starty="1" repeatx="8" priority="20"/>
|
||||
</fixed_layout>
|
||||
<fixed_layout name="12x12" width="14" height="14">
|
||||
<!--Perimeter of 'io' blocks with 'EMPTY' blocks at corners-->
|
||||
<row type="io_top" starty="H-1" priority="100"/>
|
||||
<row type="io_bottom" starty="0" priority="100"/>
|
||||
<col type="io_left" startx="0" priority="100"/>
|
||||
<col type="io_right" startx="W-1" priority="100"/>
|
||||
<corners type="EMPTY" priority="101"/>
|
||||
<!--Fill with 'clb'-->
|
||||
<fill type="clb" priority="10"/>
|
||||
<!--Column of 'mult_8' with 'EMPTY' blocks wherever a 'mult_8' does not fit. Vertical offset by 1 for perimeter.-->
|
||||
<col type="mult_16" startx="2" starty="1" repeatx="8" priority="20"/>
|
||||
</fixed_layout>
|
||||
</layout>
|
||||
<device>
|
||||
<!-- VB & JL: Using Ian Kuon's transistor sizing and drive strength data for routing, at 40 nm. Ian used BPTM
|
||||
models. We are modifying the delay values however, to include metal C and R, which allows more architecture
|
||||
experimentation. We are also modifying the relative resistance of PMOS to be 1.8x that of NMOS
|
||||
(vs. Ian's 3x) as 1.8x lines up with Jeff G's data from a 45 nm process (and is more typical of
|
||||
45 nm in general). I'm upping the Rmin_nmos from Ian's just over 6k to nearly 9k, and dropping
|
||||
RminW_pmos from 18k to 16k to hit this 1.8x ratio, while keeping the delays of buffers approximately
|
||||
lined up with Stratix IV.
|
||||
We are using Jeff G.'s capacitance data for 45 nm (in tech/ptm_45nm).
|
||||
Jeff's tables list C in for transistors with widths in multiples of the minimum feature size (45 nm).
|
||||
The minimum contactable transistor is 2.5 * 45 nm, so I need to multiply drive strength sizes in this file
|
||||
by 2.5x when looking up in Jeff's tables.
|
||||
The delay values are lined up with Stratix IV, which has an architecture similar to this
|
||||
proposed FPGA, and which is also 40 nm
|
||||
C_ipin_cblock: input capacitance of a track buffer, which VPR assumes is a single-stage
|
||||
4x minimum drive strength buffer. -->
|
||||
<sizing R_minW_nmos="8926" R_minW_pmos="16067"/>
|
||||
<!-- The grid_logic_tile_area below will be used for all blocks that do not explicitly set their own (non-routing)
|
||||
area; set to 0 since we explicitly set the area of all blocks currently in this architecture file.
|
||||
-->
|
||||
<area grid_logic_tile_area="0"/>
|
||||
<chan_width_distr>
|
||||
<x distr="uniform" peak="1.000000"/>
|
||||
<y distr="uniform" peak="1.000000"/>
|
||||
</chan_width_distr>
|
||||
<switch_block type="wilton" fs="3" sub_type="subset" sub_fs="3"/>
|
||||
<connection_block input_switch_name="ipin_cblock"/>
|
||||
</device>
|
||||
<switchlist>
|
||||
<!-- VB: the mux_trans_size and buf_size data below is in minimum width transistor *areas*, assuming the purple
|
||||
book area formula. This means the mux transistors are about 5x minimum drive strength.
|
||||
We assume the first stage of the buffer is 3x min drive strength to be reasonable given the large
|
||||
mux transistors, and this gives a reasonable stage ratio of a bit over 5x to the second stage. We assume
|
||||
the n and p transistors in the first stage are equal-sized to lower the buffer trip point, since it's fed
|
||||
by a pass transistor mux. We can then reverse engineer the buffer second stage to hit the specified
|
||||
buf_size (really buffer area) - 16.2x minimum drive nmos and 1.8*16.2 = 29.2x minimum drive.
|
||||
I then took the data from Jeff G.'s PTM modeling of 45 nm to get the Cin (gate of first stage) and Cout
|
||||
(diff of second stage) listed below. Jeff's models are in tech/ptm_45nm, and are in min feature multiples.
|
||||
The minimum contactable transistor is 2.5 * 45 nm, so I need to multiply the drive strength sizes above by
|
||||
2.5x when looking up in Jeff's tables.
|
||||
Finally, we choose a switch delay (58 ps) that leads to length 4 wires having a delay equal to that of SIV of 126 ps.
|
||||
This also leads to the switch being 46% of the total wire delay, which is reasonable. -->
|
||||
<switch type="mux" name="L1_mux" R="551" Cin=".77e-15" Cout="4e-15" Tdel="58e-12" mux_trans_size="2.630740" buf_size="27.645901"/>
|
||||
<switch type="mux" name="L2_mux" R="551" Cin=".77e-15" Cout="4e-15" Tdel="58e-12" mux_trans_size="2.630740" buf_size="27.645901"/>
|
||||
<switch type="mux" name="L4_mux" R="551" Cin=".77e-15" Cout="4e-15" Tdel="58e-12" mux_trans_size="2.630740" buf_size="27.645901"/>
|
||||
<!--switch ipin_cblock resistance set to yeild for 4x minimum drive strength buffer-->
|
||||
<switch type="mux" name="ipin_cblock" R="2231.5" Cout="0." Cin="1.47e-15" Tdel="7.247000e-11" mux_trans_size="1.222260" buf_size="auto"/>
|
||||
</switchlist>
|
||||
<segmentlist>
|
||||
<!--- VB & JL: using ITRS metal stack data, 96 nm half pitch wires, which are intermediate metal width/space.
|
||||
With the 96 nm half pitch, such wires would take 60 um of height, vs. a 90 nm high (approximated as square) Stratix IV tile so this seems
|
||||
reasonable. Using a tile length of 90 nm, corresponding to the length of a Stratix IV tile if it were square. -->
|
||||
<!-- GIVE a specific name for the segment! OpenFPGA appreciate that! -->
|
||||
<segment name="L1" freq="0.10" length="1" type="unidir" Rmetal="101" Cmetal="22.5e-15">
|
||||
<mux name="L1_mux"/>
|
||||
<sb type="pattern">1 1</sb>
|
||||
<cb type="pattern">1</cb>
|
||||
</segment>
|
||||
<segment name="L2" freq="0.10" length="2" type="unidir" Rmetal="101" Cmetal="22.5e-15">
|
||||
<mux name="L2_mux"/>
|
||||
<sb type="pattern">1 1 1</sb>
|
||||
<cb type="pattern">1 1</cb>
|
||||
</segment>
|
||||
<segment name="L4" freq="0.80" length="4" type="unidir" Rmetal="101" Cmetal="22.5e-15">
|
||||
<mux name="L4_mux"/>
|
||||
<sb type="pattern">1 1 1 1 1</sb>
|
||||
<cb type="pattern">1 1 1 1</cb>
|
||||
</segment>
|
||||
</segmentlist>
|
||||
<directlist>
|
||||
<direct name="carry_chain" from_pin="clb.cout" to_pin="clb.cin" x_offset="0" y_offset="-1" z_offset="0"/>
|
||||
<direct name="shift_register" from_pin="clb.reg_out" to_pin="clb.reg_in" x_offset="0" y_offset="-1" z_offset="0"/>
|
||||
<direct name="scan_chain" from_pin="clb.sc_out" to_pin="clb.sc_in" x_offset="0" y_offset="-1" z_offset="0"/>
|
||||
</directlist>
|
||||
<complexblocklist>
|
||||
<!-- Define input pads begin -->
|
||||
<pb_type name="io">
|
||||
<input name="outpad" num_pins="1"/>
|
||||
<output name="inpad" num_pins="1"/>
|
||||
<!-- Do NOT add clock pins to I/O here!!! VPR does not build clock network in the way that OpenFPGA can support
|
||||
If you need to register the I/O, define clocks in the circuit models
|
||||
These clocks can be handled in back-end
|
||||
-->
|
||||
<!-- A mode denotes the physical implementation of an I/O
|
||||
This mode will be not packable but is mainly used for fabric verilog generation
|
||||
-->
|
||||
<mode name="physical" disable_packing="true">
|
||||
<pb_type name="iopad" blif_model=".subckt io" num_pb="1">
|
||||
<input name="outpad" num_pins="1"/>
|
||||
<output name="inpad" num_pins="1"/>
|
||||
</pb_type>
|
||||
<interconnect>
|
||||
<direct name="outpad" input="io.outpad" output="iopad.outpad">
|
||||
<delay_constant max="1.394e-11" in_port="io.outpad" out_port="iopad.outpad"/>
|
||||
</direct>
|
||||
<direct name="inpad" input="iopad.inpad" output="io.inpad">
|
||||
<delay_constant max="4.243e-11" in_port="iopad.inpad" out_port="io.inpad"/>
|
||||
</direct>
|
||||
</interconnect>
|
||||
</mode>
|
||||
|
||||
<!-- IOs can operate as either inputs or outputs.
|
||||
Delays below come from Ian Kuon. They are small, so they should be interpreted as
|
||||
the delays to and from registers in the I/O (and generally I/Os are registered
|
||||
today and that is when you timing analyze them.
|
||||
-->
|
||||
<mode name="inpad">
|
||||
<pb_type name="inpad" blif_model=".input" num_pb="1">
|
||||
<output name="inpad" num_pins="1"/>
|
||||
</pb_type>
|
||||
<interconnect>
|
||||
<direct name="inpad" input="inpad.inpad" output="io.inpad">
|
||||
<delay_constant max="4.243e-11" in_port="inpad.inpad" out_port="io.inpad"/>
|
||||
</direct>
|
||||
</interconnect>
|
||||
</mode>
|
||||
<mode name="outpad">
|
||||
<pb_type name="outpad" blif_model=".output" num_pb="1">
|
||||
<input name="outpad" num_pins="1"/>
|
||||
</pb_type>
|
||||
<interconnect>
|
||||
<direct name="outpad" input="io.outpad" output="outpad.outpad">
|
||||
<delay_constant max="1.394e-11" in_port="io.outpad" out_port="outpad.outpad"/>
|
||||
</direct>
|
||||
</interconnect>
|
||||
</mode>
|
||||
<power method="ignore"/>
|
||||
</pb_type>
|
||||
<!-- Define I/O pads ends -->
|
||||
<!-- Define general purpose logic block (CLB) begin -->
|
||||
<!-- -Due to the absence of local routing,
|
||||
the 4 inputs of fracturable LUT4 are no longer equivalent,
|
||||
because the 4th input can not be switched when the dual-LUT3 modes are used.
|
||||
So pin equivalence should be applied to the first 3 inputs only
|
||||
-->
|
||||
<pb_type name="clb">
|
||||
<input name="I0" num_pins="2" equivalent="full"/>
|
||||
<input name="I0i" num_pins="2" equivalent="none"/>
|
||||
<input name="I1" num_pins="2" equivalent="full"/>
|
||||
<input name="I1i" num_pins="2" equivalent="none"/>
|
||||
<input name="I2" num_pins="2" equivalent="full"/>
|
||||
<input name="I2i" num_pins="2" equivalent="none"/>
|
||||
<input name="I3" num_pins="2" equivalent="full"/>
|
||||
<input name="I3i" num_pins="2" equivalent="none"/>
|
||||
<input name="I4" num_pins="2" equivalent="full"/>
|
||||
<input name="I4i" num_pins="2" equivalent="none"/>
|
||||
<input name="I5" num_pins="2" equivalent="full"/>
|
||||
<input name="I5i" num_pins="2" equivalent="none"/>
|
||||
<input name="I6" num_pins="2" equivalent="full"/>
|
||||
<input name="I6i" num_pins="2" equivalent="none"/>
|
||||
<input name="I7" num_pins="2" equivalent="full"/>
|
||||
<input name="I7i" num_pins="2" equivalent="none"/>
|
||||
<input name="reg_in" num_pins="1"/>
|
||||
<input name="sc_in" num_pins="1"/>
|
||||
<input name="cin" num_pins="1"/>
|
||||
<input name="reset" num_pins="1" is_non_clock_global="true"/>
|
||||
<output name="O" num_pins="16" equivalent="none"/>
|
||||
<output name="reg_out" num_pins="1"/>
|
||||
<output name="sc_out" num_pins="1"/>
|
||||
<output name="cout" num_pins="1"/>
|
||||
<clock name="clk" num_pins="1"/>
|
||||
<!-- Describe fracturable logic element.
|
||||
Each fracturable logic element has a 6-LUT that can alternatively operate as two 5-LUTs with shared inputs.
|
||||
The outputs of the fracturable logic element can be optionally registered
|
||||
-->
|
||||
<pb_type name="fle" num_pb="8">
|
||||
<input name="in" num_pins="4"/>
|
||||
<input name="reg_in" num_pins="1"/>
|
||||
<input name="sc_in" num_pins="1"/>
|
||||
<input name="cin" num_pins="1"/>
|
||||
<input name="reset" num_pins="1"/>
|
||||
<output name="out" num_pins="2"/>
|
||||
<output name="reg_out" num_pins="1"/>
|
||||
<output name="sc_out" num_pins="1"/>
|
||||
<output name="cout" num_pins="1"/>
|
||||
<clock name="clk" num_pins="1"/>
|
||||
<!-- Physical mode definition begin (physical implementation of the fle) -->
|
||||
<mode name="physical" disable_packing="true">
|
||||
<pb_type name="fabric" num_pb="1">
|
||||
<input name="in" num_pins="4"/>
|
||||
<input name="reg_in" num_pins="1"/>
|
||||
<input name="sc_in" num_pins="1"/>
|
||||
<input name="cin" num_pins="1"/>
|
||||
<input name="reset" num_pins="1"/>
|
||||
<output name="out" num_pins="2"/>
|
||||
<output name="reg_out" num_pins="1"/>
|
||||
<output name="sc_out" num_pins="1"/>
|
||||
<output name="cout" num_pins="1"/>
|
||||
<clock name="clk" num_pins="1"/>
|
||||
<pb_type name="frac_logic" num_pb="1">
|
||||
<input name="in" num_pins="4"/>
|
||||
<input name="cin" num_pins="1"/>
|
||||
<output name="out" num_pins="2"/>
|
||||
<output name="cout" num_pins="1"/>
|
||||
<!-- Define LUT -->
|
||||
<pb_type name="frac_lut4" blif_model=".subckt frac_lut4" num_pb="1">
|
||||
<input name="in" num_pins="4"/>
|
||||
<output name="lut2_out" num_pins="2"/>
|
||||
<output name="lut3_out" num_pins="2"/>
|
||||
<output name="lut4_out" num_pins="1"/>
|
||||
</pb_type>
|
||||
<pb_type name="carry_follower" blif_model=".subckt carry_follower_physical" num_pb="1">
|
||||
<input name="a" num_pins="1"/>
|
||||
<input name="b" num_pins="1"/>
|
||||
<input name="cin" num_pins="1"/>
|
||||
<output name="cout" num_pins="1"/>
|
||||
<delay_constant max="0.3e-9" in_port="carry_follower.a" out_port="carry_follower.cout"/>
|
||||
<delay_constant max="0.3e-9" in_port="carry_follower.b" out_port="carry_follower.cout"/>
|
||||
<delay_constant max="0.3e-9" in_port="carry_follower.cin" out_port="carry_follower.cout"/>
|
||||
</pb_type>
|
||||
<interconnect>
|
||||
<direct name="direct1" input="frac_logic.in[0:1]" output="frac_lut4.in[0:1]"/>
|
||||
<direct name="direct2" input="frac_logic.in[3:3]" output="frac_lut4.in[3:3]"/>
|
||||
<direct name="direct3" input="frac_logic.cin" output="carry_follower.b"/>
|
||||
<direct name="direct4" input="frac_lut4.lut2_out[1:1]" output="carry_follower.a"/>
|
||||
<direct name="direct5" input="frac_lut4.lut2_out[0:0]" output="carry_follower.cin"/>
|
||||
<direct name="direct6" input="carry_follower.cout" output="frac_logic.cout"/>
|
||||
<direct name="direct7" input="frac_lut4.lut3_out[1]" output="frac_logic.out[1]"/>
|
||||
<!-- Xifan Tang: I use out[0] because the output of lut6 in lut6 mode is wired to the out[0] -->
|
||||
<mux name="mux1" input="frac_lut4.lut4_out frac_lut4.lut3_out[0]" output="frac_logic.out[0]"/>
|
||||
<mux name="mux2" input="frac_logic.cin frac_logic.in[2:2]" output="frac_lut4.in[2:2]"/>
|
||||
</interconnect>
|
||||
</pb_type>
|
||||
<!-- Define flip-flop with scan-chain capability, DI is the scan-chain data input -->
|
||||
<pb_type name="ff" blif_model=".subckt scff" num_pb="2">
|
||||
<input name="D" num_pins="1"/>
|
||||
<input name="DI" num_pins="1"/>
|
||||
<input name="reset" num_pins="1"/>
|
||||
<output name="Q" num_pins="1"/>
|
||||
<clock name="clk" num_pins="1"/>
|
||||
<T_setup value="66e-12" port="ff.D" clock="clk"/>
|
||||
<T_setup value="66e-12" port="ff.DI" clock="clk"/>
|
||||
<T_setup value="66e-12" port="ff.reset" clock="clk"/>
|
||||
<T_clock_to_Q max="124e-12" port="ff.Q" clock="clk"/>
|
||||
</pb_type>
|
||||
<interconnect>
|
||||
<direct name="direct1" input="fabric.in" output="frac_logic.in"/>
|
||||
<direct name="direct2" input="fabric.cin" output="frac_logic.cin"/>
|
||||
<direct name="direct3" input="fabric.sc_in" output="ff[0].DI"/>
|
||||
<direct name="direct4" input="ff[0].Q" output="ff[1].DI"/>
|
||||
<direct name="direct5" input="ff[1].Q" output="fabric.sc_out"/>
|
||||
<direct name="direct6" input="ff[1].Q" output="fabric.reg_out"/>
|
||||
<direct name="direct7" input="frac_logic.cout" output="fabric.cout"/>
|
||||
<complete name="complete1" input="fabric.clk" output="ff[1:0].clk"/>
|
||||
<complete name="complete2" input="fabric.reset" output="ff[1:0].reset"/>
|
||||
<mux name="mux1" input="frac_logic.out[0:0] fabric.reg_in" output="ff[0:0].D">
|
||||
<delay_constant max="25e-12" in_port="frac_logic.out[0:0]" out_port="ff[0:0].D"/>
|
||||
<delay_constant max="45e-12" in_port="fabric.reg_in" out_port="ff[0:0].D"/>
|
||||
</mux>
|
||||
<mux name="mux2" input="frac_logic.out[1:1] ff[0:0].Q" output="ff[1:1].D">
|
||||
<delay_constant max="25e-12" in_port="frac_logic.out[1:1]" out_port="ff[1:1].D"/>
|
||||
<delay_constant max="45e-12" in_port="ff[0:0].Q" out_port="ff[1:1].D"/>
|
||||
</mux>
|
||||
<mux name="mux3" input="ff[0].Q frac_logic.out[0]" output="fabric.out[0]">
|
||||
<!-- LUT to output is faster than FF to output on a Stratix IV -->
|
||||
<delay_constant max="25e-12" in_port="frac_logic.out[0]" out_port="fabric.out[0]"/>
|
||||
<delay_constant max="45e-12" in_port="ff[0].Q" out_port="fabric.out[0]"/>
|
||||
</mux>
|
||||
<mux name="mux4" input="ff[1].Q frac_logic.out[1]" output="fabric.out[1]">
|
||||
<!-- LUT to output is faster than FF to output on a Stratix IV -->
|
||||
<delay_constant max="25e-12" in_port="frac_logic.out[1]" out_port="fabric.out[1]"/>
|
||||
<delay_constant max="45e-12" in_port="ff[1].Q" out_port="fabric.out[1]"/>
|
||||
</mux>
|
||||
</interconnect>
|
||||
</pb_type>
|
||||
<interconnect>
|
||||
<direct name="direct1" input="fle.in" output="fabric.in"/>
|
||||
<direct name="direct2" input="fle.reg_in" output="fabric.reg_in"/>
|
||||
<direct name="direct3" input="fle.sc_in" output="fabric.sc_in"/>
|
||||
<direct name="direct4" input="fle.cin" output="fabric.cin"/>
|
||||
<direct name="direct5" input="fabric.out" output="fle.out"/>
|
||||
<direct name="direct6" input="fabric.reg_out" output="fle.reg_out"/>
|
||||
<direct name="direct7" input="fabric.sc_out" output="fle.sc_out"/>
|
||||
<direct name="direct8" input="fabric.cout" output="fle.cout"/>
|
||||
<direct name="direct9" input="fle.clk" output="fabric.clk"/>
|
||||
<direct name="direct10" input="fle.reset" output="fabric.reset"/>
|
||||
</interconnect>
|
||||
</mode>
|
||||
<!-- Physical mode definition end (physical implementation of the fle) -->
|
||||
<!-- Arithmetic mode definition begin -->
|
||||
<mode name="arithmetic">
|
||||
<pb_type name="soft_adder" num_pb="1">
|
||||
<input name="in" num_pins="4"/>
|
||||
<input name="cin" num_pins="1"/>
|
||||
<output name="sumout" num_pins="1"/>
|
||||
<output name="cout" num_pins="1"/>
|
||||
<!-- Define special LUT marco to be used as adder -->
|
||||
<pb_type name="adder_lut4" blif_model=".subckt adder_lut4" num_pb="1">
|
||||
<input name="in" num_pins="4"/>
|
||||
<output name="lut2_out" num_pins="2"/>
|
||||
<output name="lut4_out" num_pins="1"/>
|
||||
<delay_constant max="0.3e-9" in_port="adder_lut4.in" out_port="adder_lut4.lut2_out"/>
|
||||
<delay_constant max="0.3e-9" in_port="adder_lut4.in" out_port="adder_lut4.lut4_out"/>
|
||||
</pb_type>
|
||||
<pb_type name="carry_follower" blif_model=".subckt carry_follower" num_pb="1">
|
||||
<input name="a" num_pins="1"/>
|
||||
<input name="b" num_pins="1"/>
|
||||
<input name="cin" num_pins="1"/>
|
||||
<output name="cout" num_pins="1"/>
|
||||
<delay_constant max="0.3e-9" in_port="carry_follower.a" out_port="carry_follower.cout"/>
|
||||
<delay_constant max="0.3e-9" in_port="carry_follower.b" out_port="carry_follower.cout"/>
|
||||
<delay_constant max="0.3e-9" in_port="carry_follower.cin" out_port="carry_follower.cout"/>
|
||||
</pb_type>
|
||||
<interconnect>
|
||||
<direct name="direct1" input="soft_adder.in[0:1]" output="adder_lut4.in[0:1]"/>
|
||||
<direct name="direct2" input="soft_adder.in[3:3]" output="adder_lut4.in[3:3]"/>
|
||||
<direct name="direct3" input="soft_adder.cin" output="carry_follower.b">
|
||||
<!-- Pack pattern to build an adder chain connection considered by packer -->
|
||||
<pack_pattern name="chain" in_port="soft_adder.cin" out_port="carry_follower.b"/>
|
||||
</direct>
|
||||
<direct name="direct4" input="adder_lut4.lut2_out[1:1]" output="carry_follower.a">
|
||||
<!-- Pack pattern to pair adder_lut4 and carry_follower into a molecule
|
||||
considered by packer -->
|
||||
<pack_pattern name="lut_follower" in_port="adder_lut4.lut2_out[1:1]" out_port="carry_follower.a"/>
|
||||
</direct>
|
||||
<direct name="direct5" input="adder_lut4.lut2_out[0:0]" output="carry_follower.cin">
|
||||
</direct>
|
||||
<direct name="direct6" input="carry_follower.cout" output="soft_adder.cout">
|
||||
<!-- Pack pattern to build an adder chain connection considered by packer -->
|
||||
<pack_pattern name="chain" in_port="carry_follower.cout" out_port="soft_adder.cout"/>
|
||||
</direct>
|
||||
<direct name="direct7" input="adder_lut4.lut4_out" output="soft_adder.sumout[0:0]">
|
||||
</direct>
|
||||
<mux name="mux1" input="soft_adder.cin soft_adder.in[2:2]" output="adder_lut4.in[2:2]">
|
||||
</mux>
|
||||
</interconnect>
|
||||
</pb_type>
|
||||
<interconnect>
|
||||
<direct name="direct1" input="fle.in" output="soft_adder.in"/>
|
||||
<direct name="direct2" input="fle.cin" output="soft_adder.cin">
|
||||
<!-- Pack pattern to build an adder chain connection considered by packer -->
|
||||
<pack_pattern name="chain" in_port="fle.cin" out_port="soft_adder.cin"/>
|
||||
</direct>
|
||||
<direct name="direct3" input="soft_adder.sumout" output="fle.out[0:0]"/>
|
||||
<direct name="direct4" input="soft_adder.cout" output="fle.cout">
|
||||
<!-- Pack pattern to build an adder chain connection considered by packer -->
|
||||
<pack_pattern name="chain" in_port="soft_adder.cout" out_port="fle.cout"/>
|
||||
</direct>
|
||||
</interconnect>
|
||||
</mode>
|
||||
<!-- Arithmetic mode definition end -->
|
||||
<!-- Dual 3-LUT mode definition begin -->
|
||||
<mode name="n2_lut3">
|
||||
<pb_type name="lut3inter" num_pb="1">
|
||||
<input name="in" num_pins="3"/>
|
||||
<output name="out" num_pins="2"/>
|
||||
<clock name="clk" num_pins="1"/>
|
||||
<pb_type name="ble3" num_pb="2">
|
||||
<input name="in" num_pins="3"/>
|
||||
<output name="out" num_pins="1"/>
|
||||
<clock name="clk" num_pins="1"/>
|
||||
<!-- Define the LUT -->
|
||||
<pb_type name="lut3" blif_model=".names" num_pb="1" class="lut">
|
||||
<input name="in" num_pins="3" port_class="lut_in"/>
|
||||
<output name="out" num_pins="1" port_class="lut_out"/>
|
||||
<!-- LUT timing using delay matrix -->
|
||||
<!-- These are the physical delay inputs on a Stratix IV LUT but because VPR cannot do LUT rebalancing,
|
||||
we instead take the average of these numbers to get more stable results
|
||||
82e-12
|
||||
173e-12
|
||||
261e-12
|
||||
263e-12
|
||||
398e-12
|
||||
-->
|
||||
<delay_matrix type="max" in_port="lut3.in" out_port="lut3.out">
|
||||
235e-12
|
||||
235e-12
|
||||
235e-12
|
||||
</delay_matrix>
|
||||
</pb_type>
|
||||
<!-- Define the flip-flop -->
|
||||
<pb_type name="ff" blif_model=".latch" num_pb="1" class="flipflop">
|
||||
<input name="D" num_pins="1" port_class="D"/>
|
||||
<output name="Q" num_pins="1" port_class="Q"/>
|
||||
<clock name="clk" num_pins="1" port_class="clock"/>
|
||||
<T_setup value="66e-12" port="ff.D" clock="clk"/>
|
||||
<T_clock_to_Q max="124e-12" port="ff.Q" clock="clk"/>
|
||||
</pb_type>
|
||||
<interconnect>
|
||||
<direct name="direct1" input="ble3.in[2:0]" output="lut3[0:0].in[2:0]"/>
|
||||
<direct name="direct2" input="lut3[0:0].out" output="ff[0:0].D">
|
||||
<!-- Advanced user option that tells CAD tool to find LUT+FF pairs in netlist -->
|
||||
<pack_pattern name="ble3" in_port="lut3[0:0].out" out_port="ff[0:0].D"/>
|
||||
</direct>
|
||||
<direct name="direct3" input="ble3.clk" output="ff[0:0].clk"/>
|
||||
<mux name="mux1" input="ff[0:0].Q lut3.out[0:0]" output="ble3.out[0:0]">
|
||||
<!-- LUT to output is faster than FF to output on a Stratix IV -->
|
||||
<delay_constant max="25e-12" in_port="lut3.out[0:0]" out_port="ble3.out[0:0]"/>
|
||||
<delay_constant max="45e-12" in_port="ff[0:0].Q" out_port="ble3.out[0:0]"/>
|
||||
</mux>
|
||||
</interconnect>
|
||||
</pb_type>
|
||||
<interconnect>
|
||||
<direct name="direct1" input="lut3inter.in" output="ble3[0:0].in"/>
|
||||
<direct name="direct2" input="lut3inter.in" output="ble3[1:1].in"/>
|
||||
<direct name="direct3" input="ble3[1:0].out" output="lut3inter.out"/>
|
||||
<complete name="complete1" input="lut3inter.clk" output="ble3[1:0].clk"/>
|
||||
</interconnect>
|
||||
</pb_type>
|
||||
<interconnect>
|
||||
<direct name="direct1" input="fle.in[2:0]" output="lut3inter.in"/>
|
||||
<direct name="direct2" input="lut3inter.out" output="fle.out"/>
|
||||
<direct name="direct3" input="fle.clk" output="lut3inter.clk"/>
|
||||
</interconnect>
|
||||
</mode>
|
||||
<!-- Dual 3-LUT mode definition end -->
|
||||
<!-- 4-LUT mode definition begin -->
|
||||
<mode name="n1_lut4">
|
||||
<!-- Define 4-LUT mode -->
|
||||
<pb_type name="ble4" num_pb="1">
|
||||
<input name="in" num_pins="4"/>
|
||||
<output name="out" num_pins="1"/>
|
||||
<clock name="clk" num_pins="1"/>
|
||||
<!-- Define LUT -->
|
||||
<pb_type name="lut4" blif_model=".names" num_pb="1" class="lut">
|
||||
<input name="in" num_pins="4" port_class="lut_in"/>
|
||||
<output name="out" num_pins="1" port_class="lut_out"/>
|
||||
<!-- LUT timing using delay matrix -->
|
||||
<!-- These are the physical delay inputs on a Stratix IV LUT but because VPR cannot do LUT rebalancing,
|
||||
we instead take the average of these numbers to get more stable results
|
||||
82e-12
|
||||
173e-12
|
||||
261e-12
|
||||
263e-12
|
||||
398e-12
|
||||
397e-12
|
||||
-->
|
||||
<delay_matrix type="max" in_port="lut4.in" out_port="lut4.out">
|
||||
261e-12
|
||||
261e-12
|
||||
261e-12
|
||||
261e-12
|
||||
</delay_matrix>
|
||||
</pb_type>
|
||||
<!-- Define flip-flop -->
|
||||
<pb_type name="ff" blif_model=".latch" num_pb="1" class="flipflop">
|
||||
<input name="D" num_pins="1" port_class="D"/>
|
||||
<output name="Q" num_pins="1" port_class="Q"/>
|
||||
<clock name="clk" num_pins="1" port_class="clock"/>
|
||||
<T_setup value="66e-12" port="ff.D" clock="clk"/>
|
||||
<T_clock_to_Q max="124e-12" port="ff.Q" clock="clk"/>
|
||||
</pb_type>
|
||||
<interconnect>
|
||||
<direct name="direct1" input="ble4.in" output="lut4[0:0].in"/>
|
||||
<direct name="direct2" input="lut4.out" output="ff.D">
|
||||
<!-- Advanced user option that tells CAD tool to find LUT+FF pairs in netlist -->
|
||||
<pack_pattern name="ble4" in_port="lut4.out" out_port="ff.D"/>
|
||||
</direct>
|
||||
<direct name="direct3" input="ble4.clk" output="ff.clk"/>
|
||||
<mux name="mux1" input="ff.Q lut4.out" output="ble4.out">
|
||||
<!-- LUT to output is faster than FF to output on a Stratix IV -->
|
||||
<delay_constant max="25e-12" in_port="lut4.out" out_port="ble4.out"/>
|
||||
<delay_constant max="45e-12" in_port="ff.Q" out_port="ble4.out"/>
|
||||
</mux>
|
||||
</interconnect>
|
||||
</pb_type>
|
||||
<interconnect>
|
||||
<direct name="direct1" input="fle.in" output="ble4.in"/>
|
||||
<direct name="direct2" input="ble4.out" output="fle.out[0:0]"/>
|
||||
<direct name="direct3" input="fle.clk" output="ble4.clk"/>
|
||||
</interconnect>
|
||||
</mode>
|
||||
<!-- 4-LUT mode definition end -->
|
||||
<!-- Define shift register begin -->
|
||||
<mode name="shift_register">
|
||||
<pb_type name="shift_reg" num_pb="1">
|
||||
<input name="reg_in" num_pins="1"/>
|
||||
<output name="ff_out" num_pins="2"/>
|
||||
<output name="reg_out" num_pins="1"/>
|
||||
<clock name="clk" num_pins="1"/>
|
||||
<pb_type name="ff" blif_model=".latch" num_pb="2" class="flipflop">
|
||||
<input name="D" num_pins="1" port_class="D"/>
|
||||
<output name="Q" num_pins="1" port_class="Q"/>
|
||||
<clock name="clk" num_pins="1" port_class="clock"/>
|
||||
<T_setup value="66e-12" port="ff.D" clock="clk"/>
|
||||
<T_clock_to_Q max="124e-12" port="ff.Q" clock="clk"/>
|
||||
</pb_type>
|
||||
<interconnect>
|
||||
<direct name="direct1" input="shift_reg.reg_in" output="ff[0].D"/>
|
||||
<direct name="direct2" input="ff[0].Q" output="ff[1].D"/>
|
||||
<direct name="direct3" input="ff[1].Q" output="shift_reg.reg_out"/>
|
||||
<direct name="direct4" input="ff[0].Q" output="shift_reg.ff_out[0:0]"/>
|
||||
<direct name="direct5" input="ff[1].Q" output="shift_reg.ff_out[1:1]"/>
|
||||
<complete name="complete1" input="shift_reg.clk" output="ff.clk"/>
|
||||
</interconnect>
|
||||
</pb_type>
|
||||
<interconnect>
|
||||
<direct name="direct1" input="fle.reg_in" output="shift_reg.reg_in"/>
|
||||
<direct name="direct2" input="shift_reg.reg_out" output="fle.reg_out"/>
|
||||
<direct name="direct3" input="shift_reg.ff_out" output="fle.out"/>
|
||||
<direct name="direct4" input="fle.clk" output="shift_reg.clk"/>
|
||||
</interconnect>
|
||||
</mode>
|
||||
<!-- Define shift register end -->
|
||||
</pb_type>
|
||||
<interconnect>
|
||||
<!-- We use direct connections to reduce the area to the most
|
||||
The global local routing is going to compensate the loss in routability
|
||||
-->
|
||||
<!-- FIXME: The implicit port definition results in I0[0] connected to
|
||||
in[2]. Such twisted connection is not expected.
|
||||
I[0] should be connected to in[0]
|
||||
-->
|
||||
<direct name="direct_fle0" input="clb.I0[0:1]" output="fle[0:0].in[0:1]">
|
||||
<!-- TODO: Timing should be backannotated from post-PnR results -->
|
||||
</direct>
|
||||
<direct name="direct_fle0i" input="clb.I0i[0:1]" output="fle[0:0].in[2:3]">
|
||||
<!-- TODO: Timing should be backannotated from post-PnR results -->
|
||||
</direct>
|
||||
<direct name="direct_fle1" input="clb.I1[0:1]" output="fle[1:1].in[0:1]">
|
||||
<!-- TODO: Timing should be backannotated from post-PnR results -->
|
||||
</direct>
|
||||
<direct name="direct_fle1i" input="clb.I1i[0:1]" output="fle[1:1].in[2:3]">
|
||||
<!-- TODO: Timing should be backannotated from post-PnR results -->
|
||||
</direct>
|
||||
<direct name="direct_fle2" input="clb.I2[0:1]" output="fle[2:2].in[0:1]">
|
||||
<!-- TODO: Timing should be backannotated from post-PnR results -->
|
||||
</direct>
|
||||
<direct name="direct_fle2i" input="clb.I2i[0:1]" output="fle[2:2].in[2:3]">
|
||||
<!-- TODO: Timing should be backannotated from post-PnR results -->
|
||||
</direct>
|
||||
<direct name="direct_fle3" input="clb.I3[0:1]" output="fle[3:3].in[0:1]">
|
||||
<!-- TODO: Timing should be backannotated from post-PnR results -->
|
||||
</direct>
|
||||
<direct name="direct_fle3i" input="clb.I3i[0:1]" output="fle[3:3].in[2:3]">
|
||||
<!-- TODO: Timing should be backannotated from post-PnR results -->
|
||||
</direct>
|
||||
<direct name="direct_fle4" input="clb.I4[0:1]" output="fle[4:4].in[0:1]">
|
||||
<!-- TODO: Timing should be backannotated from post-PnR results -->
|
||||
</direct>
|
||||
<direct name="direct_fle4i" input="clb.I4i[0:1]" output="fle[4:4].in[2:3]">
|
||||
<!-- TODO: Timing should be backannotated from post-PnR results -->
|
||||
</direct>
|
||||
<direct name="direct_fle5" input="clb.I5[0:1]" output="fle[5:5].in[0:1]">
|
||||
<!-- TODO: Timing should be backannotated from post-PnR results -->
|
||||
</direct>
|
||||
<direct name="direct_fle5i" input="clb.I5i[0:1]" output="fle[5:5].in[2:3]">
|
||||
<!-- TODO: Timing should be backannotated from post-PnR results -->
|
||||
</direct>
|
||||
<direct name="direct_fle6" input="clb.I6[0:1]" output="fle[6:6].in[0:1]">
|
||||
<!-- TODO: Timing should be backannotated from post-PnR results -->
|
||||
</direct>
|
||||
<direct name="direct_fle6i" input="clb.I6i[0:1]" output="fle[6:6].in[2:3]">
|
||||
<!-- TODO: Timing should be backannotated from post-PnR results -->
|
||||
</direct>
|
||||
<direct name="direct_fle7" input="clb.I7[0:1]" output="fle[7:7].in[0:1]">
|
||||
<!-- TODO: Timing should be backannotated from post-PnR results -->
|
||||
</direct>
|
||||
<direct name="direct_fle7i" input="clb.I7i[0:1]" output="fle[7:7].in[2:3]">
|
||||
<!-- TODO: Timing should be backannotated from post-PnR results -->
|
||||
</direct>
|
||||
<complete name="clks" input="clb.clk" output="fle[7:0].clk">
|
||||
</complete>
|
||||
<complete name="resets" input="clb.reset" output="fle[7:0].reset">
|
||||
</complete>
|
||||
<!-- This way of specifying direct connection to clb outputs is important because this architecture uses automatic spreading of opins.
|
||||
By grouping to output pins in this fashion, if a logic block is completely filled by 6-LUTs,
|
||||
then the outputs those 6-LUTs take get evenly distributed across all four sides of the CLB instead of clumped on two sides (which is what happens with a more
|
||||
naive specification).
|
||||
-->
|
||||
<direct name="clbouts1" input="fle[3:0].out[0:1]" output="clb.O[7:0]"/>
|
||||
<direct name="clbouts2" input="fle[7:4].out[0:1]" output="clb.O[15:8]"/>
|
||||
<!-- Shift register chain links -->
|
||||
<direct name="shift_register_in" input="clb.reg_in" output="fle[0:0].reg_in">
|
||||
<!-- Put all inter-block carry chain delay on this one edge -->
|
||||
<delay_constant max="0.16e-9" in_port="clb.reg_in" out_port="fle[0:0].reg_in"/>
|
||||
<!--pack_pattern name="chain" in_port="clb.reg_in" out_port="fle[0:0].reg_in"/-->
|
||||
</direct>
|
||||
<direct name="shift_register_out" input="fle[7:7].reg_out" output="clb.reg_out">
|
||||
<!--pack_pattern name="chain" in_port="fle[7:7].reg_out" out_port="clb.reg_out"/-->
|
||||
</direct>
|
||||
<direct name="shift_register_link" input="fle[6:0].reg_out" output="fle[7:1].reg_in">
|
||||
<!--pack_pattern name="chain" in_port="fle[6:0].reg_out" out_port="fle[7:1].reg_in"/-->
|
||||
</direct>
|
||||
<!-- Scan chain links -->
|
||||
<direct name="scan_chain_in" input="clb.sc_in" output="fle[0:0].sc_in">
|
||||
<!-- Put all inter-block carry chain delay on this one edge -->
|
||||
<delay_constant max="0.16e-9" in_port="clb.sc_in" out_port="fle[0:0].sc_in"/>
|
||||
</direct>
|
||||
<direct name="scan_chain_out" input="fle[7:7].sc_out" output="clb.sc_out">
|
||||
</direct>
|
||||
<direct name="scan_chain_link" input="fle[6:0].sc_out" output="fle[7:1].sc_in">
|
||||
</direct>
|
||||
<!-- Carry chain links -->
|
||||
<direct name="carry_chain_in" input="clb.cin" output="fle[0:0].cin">
|
||||
<!-- Put all inter-block carry chain delay on this one edge -->
|
||||
<pack_pattern name="chain" in_port="clb.cin" out_port="fle[0:0].cin"/>
|
||||
<delay_constant max="0.16e-9" in_port="clb.cin" out_port="fle[0:0].cin"/>
|
||||
</direct>
|
||||
<direct name="carry_chain_out" input="fle[7:7].cout" output="clb.cout">
|
||||
<pack_pattern name="chain" in_port="fle[7:7].cout" out_port="clb.cout"/>
|
||||
</direct>
|
||||
<direct name="carry_chain_link" input="fle[6:0].cout" output="fle[7:1].cin">
|
||||
<pack_pattern name="chain" in_port="fle[6:0].cout" out_port="fle[7:1].cin"/>
|
||||
</direct>
|
||||
</interconnect>
|
||||
<!-- Every input pin is driven by 15% of the tracks in a channel, every output pin is driven by 10% of the tracks in a channel -->
|
||||
<!-- Place this general purpose logic block in any unspecified column -->
|
||||
</pb_type>
|
||||
<!-- Define general purpose logic block (CLB) ends -->
|
||||
<!-- Define fracturable multiplier begin -->
|
||||
<pb_type name="mult_16">
|
||||
<input name="a" num_pins="16"/>
|
||||
<input name="b" num_pins="16"/>
|
||||
<output name="out" num_pins="32"/>
|
||||
<mode name="mult_8x8">
|
||||
<pb_type name="mult_8x8_slice" num_pb="2">
|
||||
<input name="A_cfg" num_pins="8"/>
|
||||
<input name="B_cfg" num_pins="8"/>
|
||||
<output name="OUT_cfg" num_pins="16"/>
|
||||
<pb_type name="mult_8x8" blif_model=".subckt mult_8" num_pb="1">
|
||||
<input name="A" num_pins="8"/>
|
||||
<input name="B" num_pins="8"/>
|
||||
<output name="Y" num_pins="16"/>
|
||||
<delay_constant max="1.523e-9" min="0.776e-9" in_port="mult_8x8.A" out_port="mult_8x8.Y"/>
|
||||
<delay_constant max="1.523e-9" min="0.776e-9" in_port="mult_8x8.B" out_port="mult_8x8.Y"/>
|
||||
</pb_type>
|
||||
<interconnect>
|
||||
<direct name="a2a" input="mult_8x8_slice.A_cfg" output="mult_8x8.A">
|
||||
</direct>
|
||||
<direct name="b2b" input="mult_8x8_slice.B_cfg" output="mult_8x8.B">
|
||||
</direct>
|
||||
<direct name="out2out" input="mult_8x8.Y" output="mult_8x8_slice.OUT_cfg">
|
||||
</direct>
|
||||
</interconnect>
|
||||
<power method="pin-toggle">
|
||||
<port name="A_cfg" energy_per_toggle="2.13e-12"/>
|
||||
<port name="B_cfg" energy_per_toggle="2.13e-12"/>
|
||||
<static_power power_per_instance="0.0"/>
|
||||
</power>
|
||||
</pb_type>
|
||||
<interconnect>
|
||||
<direct name="a2a_0" input="mult_16.a[0:7]" output="mult_8x8_slice[0].A_cfg[0:7]">
|
||||
<delay_constant max="134e-12" min="74e-12" in_port="mult_16.a[0:7]" out_port="mult_8x8_slice[0].A_cfg[0:7]"/>
|
||||
</direct>
|
||||
<direct name="a2a_1" input="mult_16.a[8:15]" output="mult_8x8_slice[1].A_cfg[0:7]">
|
||||
<delay_constant max="134e-12" min="74e-12" in_port="mult_16.a[8:15]" out_port="mult_8x8_slice[1].A_cfg[0:7]"/>
|
||||
</direct>
|
||||
<direct name="b2b_0" input="mult_16.b[0:7]" output="mult_8x8_slice[0].B_cfg[0:7]">
|
||||
<delay_constant max="134e-12" min="74e-12" in_port="mult_16.b[0:7]" out_port="mult_8x8_slice[0].B_cfg[0:7]"/>
|
||||
</direct>
|
||||
<direct name="b2b_1" input="mult_16.b[8:15]" output="mult_8x8_slice[1].B_cfg[0:7]">
|
||||
<delay_constant max="134e-12" min="74e-12" in_port="mult_16.b[8:15]" out_port="mult_8x8_slice[1].B_cfg[0:7]"/>
|
||||
</direct>
|
||||
<direct name="out2out_0" input="mult_8x8_slice[0].OUT_cfg[0:15]" output="mult_16.out[0:15]">
|
||||
<delay_constant max="1.93e-9" min="74e-12" in_port="mult_8x8_slice[0].OUT_cfg[0:15]" out_port="mult_16.out[0:15]"/>
|
||||
</direct>
|
||||
<direct name="out2out_1" input="mult_8x8_slice[1].OUT_cfg[0:15]" output="mult_16.out[16:31]">
|
||||
<delay_constant max="1.93e-9" min="74e-12" in_port="mult_8x8_slice[1].OUT_cfg[0:15]" out_port="mult_16.out[16:31]"/>
|
||||
</direct>
|
||||
</interconnect>
|
||||
</mode>
|
||||
<mode name="mult_16x16">
|
||||
<pb_type name="mult_16x16_slice" num_pb="1">
|
||||
<input name="A_cfg" num_pins="16"/>
|
||||
<input name="B_cfg" num_pins="16"/>
|
||||
<output name="OUT_cfg" num_pins="32"/>
|
||||
<pb_type name="mult_16x16" blif_model=".subckt mult_16" num_pb="1">
|
||||
<input name="A" num_pins="16"/>
|
||||
<input name="B" num_pins="16"/>
|
||||
<output name="Y" num_pins="32"/>
|
||||
<delay_constant max="1.523e-9" min="0.776e-9" in_port="mult_16x16.A" out_port="mult_16x16.Y"/>
|
||||
<delay_constant max="1.523e-9" min="0.776e-9" in_port="mult_16x16.B" out_port="mult_16x16.Y"/>
|
||||
</pb_type>
|
||||
<interconnect>
|
||||
<direct name="a2a" input="mult_16x16_slice.A_cfg" output="mult_16x16.A">
|
||||
</direct>
|
||||
<direct name="b2b" input="mult_16x16_slice.B_cfg" output="mult_16x16.B">
|
||||
</direct>
|
||||
<direct name="out2out" input="mult_16x16.Y" output="mult_16x16_slice.OUT_cfg">
|
||||
</direct>
|
||||
</interconnect>
|
||||
<power method="pin-toggle">
|
||||
<port name="A_cfg" energy_per_toggle="2.13e-12"/>
|
||||
<port name="B_cfg" energy_per_toggle="2.13e-12"/>
|
||||
<static_power power_per_instance="0.0"/>
|
||||
</power>
|
||||
</pb_type>
|
||||
<interconnect>
|
||||
<direct name="a2a" input="mult_16.a" output="mult_16x16_slice.A_cfg">
|
||||
<delay_constant max="134e-12" min="74e-12" in_port="mult_16.a" out_port="mult_16x16_slice.A_cfg"/>
|
||||
</direct>
|
||||
<direct name="b2b" input="mult_16.b" output="mult_16x16_slice.B_cfg">
|
||||
<delay_constant max="134e-12" min="74e-12" in_port="mult_16.b" out_port="mult_16x16_slice.B_cfg"/>
|
||||
</direct>
|
||||
<direct name="out2out" input="mult_16x16_slice.OUT_cfg" output="mult_16.out">
|
||||
<delay_constant max="1.93e-9" min="74e-12" in_port="mult_16x16_slice.OUT_cfg" out_port="mult_16.out"/>
|
||||
</direct>
|
||||
</interconnect>
|
||||
</mode>
|
||||
<!-- Place this multiplier block every 8 columns from (and including) the sixth column -->
|
||||
<power method="sum-of-children"/>
|
||||
</pb_type>
|
||||
<!-- Define fracturable multiplier end -->
|
||||
</complexblocklist>
|
||||
</architecture>
|
Loading…
Reference in New Issue