You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: vtr_flow/benchmarks/verilog/koios/README.md
+32-11Lines changed: 32 additions & 11 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -13,7 +13,7 @@ Koios benchmarks use advanced DSP features that are available in only a few FPGA
13
13
14
14
Similarly, Koios benchmark instantiate hard memory blocks (single port and dual port BRAM). These blocks are available in most architectures provided with VTR, but may not be available in a user's architecture. These hard memory blocks can be controlled by using the macro ``hard_mem``. If `hard_mem` is defined in a benchmark file (using `` `define hard_mem``), then hard memory blocks will be used. If `hard_mem` is not defined, then equivalent functionality is obtained through behavioral Verilog that the synthesis tools can automatically infer as RAMs.
15
15
16
-
From a flow perspective, you can enable/disable a macro (like `complex_dsp`) without actually modifying the benchmark file(s). You can specify a separate Verilog header file while running a flow/task that contains these macros. For `run_vtr_flow` users, `-include <filename>` needs to be added. For `run_vtr_task` users, `includes_dir` and `include_list_add` need to be specified in the task file. An example task file can be seen [here](https://github.com/verilog-to-routing/vtr-verilog-to-routing/blob/master/vtr_flow/tasks/regression_tests/vtr_reg_basic/hdl_include/config/config.txt).
16
+
From a flow perspective, you can enable/disable a macro (like `complex_dsp`) without actually modifying the benchmark file(s). You can specify a separate Verilog header file while running a flow/task that contains these macros. For `run_vtr_flow` users, `-include <filename>` needs to be added. For `run_vtr_task` users, `includes_dir` and `include_list_add` need to be specified in the task file. An example task file can be seen [here](https://github.com/verilog-to-routing/vtr-verilog-to-routing/blob/master/vtr_flow/vtr_reg_basic/hdl_include/config/config.txt).
17
17
18
18
Using hard macros for DSPs and BRAMs to utilize advanced features (like chaining) is common in modern designs used with contemporary FPGAs. When using these benchmarks and enabling these advanced features, an FPGA architecture that supports these features must be provided. Supporting these features implies that the architecture XML file provided to VTR must describe such features (e.g. by defining a hard block macro DSP slice). We provide such architectures with Koios. The FPGA architectures with advanced DSP that work out-of-the-box with Koios benchmarks are available here:
19
19
@@ -30,16 +30,37 @@ If users want to use a different FPGA architecture file, they can replace the ma
30
30
In Koios 2.0, there are 8 synthetic/proxy benchmarks. These were generated using a framework that is present [here](https://github.com/UT-LCA/koios_proxy_benchmarks). To generate more benchmarks using this framework, use the generate_benchmark.py script.
31
31
32
32
## Regressions
33
-
Koios benchmarks are tested by the following regression tests in VTR:
34
-
| Suite |Test Description | Config file | Wall-clock time |
| Strong | A test circuit. Goal is to check the architecture files. | tasks/regression_tests/vtr_reg_strong/koios | 6 seconds |
37
-
| Strong | Same test circuit without enabling complex dsp features | tasks/regression_tests/vtr_reg_strong/koios_no_complex_dsp | 6 seconds|
38
-
| Nightly | Small-to-medium sized designs from Koios run with one arch file | tasks/regression_tests/vtr_reg_nightly_test6/koios | 2 hours with -j3 |
39
-
| Nightly | Small-to-medium sized designs from Koios run with an arch file without enabling complex dsp features | tasks/regression_tests/vtr_reg_nightly_test6/koios_no_complex_dsp | 2 hours with -j3 |
40
-
| Nightly | A small design from Koios run with various flavors of the arch file that enables complex dsp features | tasks/regression_tests/vtr_reg_nightly_test6/koios_multi_arch | 2 hours with -j3 |
41
-
| Weekly | Large designs from Koios run with one arch file | tasks/regression_tests/vtr_reg_weekly/koios | a little over 24 hours with -j4 |
42
-
| Weekly | Large designs from Koios run with an arch file without enabling complex dsp features | tasks/regression_tests/vtr_reg_weekly/koios_no_complex_dsp | a little over 24 hours with -j4 |
33
+
Koios benchmarks are tested by the following tasks in VTR:
34
+
| Suite |Test Description | Target | Complex DSP Features | Config file | Frontend | Parser |
| Strong | A test circuit to check the architecture file | k6FracN10LB_mem20K_complexDSP_customSB_22nm.xml |✓| vtr_reg_strong/koios_test | Parmys |
53
+
| Strong | Same test circuit without hard blocks | k6FracN10LB_mem20K_complexDSP_customSB_22nm.xml || vtr_reg_strong/koios_test_no_hb | Parmys |
54
+
| Strong | A test circuit to check the architecture file | k6FracN10LB_mem20K_complexDSP_customSB_22nm.xml |✓| vtr_reg_strong_odin/koios_test | Odin |
55
+
| Strong | Same test circuit without hard blocks | k6FracN10LB_mem20K_complexDSP_customSB_22nm.xml || vtr_reg_strong_odin/koios_test_no_hb | Odin |
For collecting QoR measurements on Koios benchmarks, follow the instructions [here](https://docs.verilogtorouting.org/en/latest/README.developers/#example-koios-benchmarks-qor-measurement).
0 commit comments