martin-frbg:issue5414 - Branch - OpenMathLib/OpenBLAS

Split VORTEXM4 from VORTEX target and fix SGEMM_DIRECT support for SME-capable targets

#5423

Comparing

martin-frbg:issue5414

(

9bfc361

) with

develop

(

b6d5057

)

Untouched: 62

Benchmarks

Passed

test_daxpy[100-d]

benchmark/pybench/benchmarks/bench_blas.py

+1%

24.3 µs24.1 µs

test_nrm2[100-dz]

benchmark/pybench/benchmarks/bench_blas.py

29 µs28.9 µs

test_daxpy[100-z]

benchmark/pybench/benchmarks/bench_blas.py

25.9 µs25.8 µs

test_daxpy[1000-s]

benchmark/pybench/benchmarks/bench_blas.py

27.6 µs27.5 µs

test_daxpy[100-s]

benchmark/pybench/benchmarks/bench_blas.py

24.1 µs24 µs

test_daxpy[1000-c]

benchmark/pybench/benchmarks/bench_blas.py

32.8 µs32.7 µs

test_dot[1000]

benchmark/pybench/benchmarks/bench_blas.py

28.5 µs28.4 µs

test_daxpy[100-c]

benchmark/pybench/benchmarks/bench_blas.py

25.2 µs25.1 µs

test_daxpy[1000-z]

benchmark/pybench/benchmarks/bench_blas.py

40.6 µs40.5 µs

test_dgbmv[1-100-d]

benchmark/pybench/benchmarks/bench_blas.py

38 µs37.9 µs

test_dgbmv[1-1000-d]

benchmark/pybench/benchmarks/bench_blas.py

83.7 µs83.5 µs

test_daxpy[1000-d]

benchmark/pybench/benchmarks/bench_blas.py

32.4 µs32.4 µs

test_nrm2[1000-dz]

benchmark/pybench/benchmarks/bench_blas.py

35.5 µs35.5 µs

test_dgemv[100-s]

benchmark/pybench/benchmarks/bench_blas.py

104 µs103.8 µs

test_dgemv[100-d]

benchmark/pybench/benchmarks/bench_blas.py

141.3 µs141.1 µs

test_dgemv[100-c]

benchmark/pybench/benchmarks/bench_blas.py

149.8 µs149.7 µs

test_dgbmv[1-100-z]

benchmark/pybench/benchmarks/bench_blas.py

42.1 µs42.1 µs

test_dgemv[100-z]

benchmark/pybench/benchmarks/bench_blas.py

231.2 µs231.1 µs

test_gesv[100-z]

benchmark/pybench/benchmarks/bench_blas.py

937.7 µs937.5 µs

test_gemm[100-d]

benchmark/pybench/benchmarks/bench_blas.py

471.3 µs471.2 µs

test_gemm[100-s]

benchmark/pybench/benchmarks/bench_blas.py

273.2 µs273.2 µs

test_dgbmv[1-1000-c]

benchmark/pybench/benchmarks/bench_blas.py

99.5 µs99.5 µs

test_gemm[100-z]

benchmark/pybench/benchmarks/bench_blas.py

1.2 ms1.2 ms

test_dgemv[1000-s]

benchmark/pybench/benchmarks/bench_blas.py

7 ms7 ms

test_gesv[100-d]

benchmark/pybench/benchmarks/bench_blas.py

395 µs395 µs

test_gemm[100-c]

benchmark/pybench/benchmarks/bench_blas.py

659.6 µs659.5 µs

test_syev[50-s]

benchmark/pybench/benchmarks/bench_blas.py

1.3 ms1.3 ms

test_gesv[100-c]

benchmark/pybench/benchmarks/bench_blas.py

696.1 µs696.1 µs

test_syrk[1000-c]

benchmark/pybench/benchmarks/bench_blas.py

227.5 ms227.5 ms

test_syrk[1000-s]

benchmark/pybench/benchmarks/bench_blas.py

65.4 ms65.4 ms

test_syrk[1000-d]

benchmark/pybench/benchmarks/bench_blas.py

130.4 ms130.4 ms

test_gemm[1000-c]

benchmark/pybench/benchmarks/bench_blas.py

426 ms426 ms

test_gemm[1000-d]

benchmark/pybench/benchmarks/bench_blas.py

239.4 ms239.4 ms

test_gemm[1000-z]

benchmark/pybench/benchmarks/bench_blas.py

875.6 ms875.6 ms

test_nrm2[1000-d]

benchmark/pybench/benchmarks/bench_blas.py

30.5 µs30.5 µs

test_dot[100]

benchmark/pybench/benchmarks/bench_blas.py

22.4 µs22.4 µs

test_gesv[1000-z]

benchmark/pybench/benchmarks/bench_blas.py

353.6 ms353.6 ms

test_gesv[1000-d]

benchmark/pybench/benchmarks/bench_blas.py

93.3 ms93.3 ms

test_syrk[1000-z]

benchmark/pybench/benchmarks/bench_blas.py

476.4 ms476.4 ms

test_dgemv[1000-z]

benchmark/pybench/benchmarks/bench_blas.py

26.3 ms26.3 ms

test_syev[200-s]

benchmark/pybench/benchmarks/bench_blas.py

49.1 ms49.1 ms

test_syrk[100-z]

benchmark/pybench/benchmarks/bench_blas.py

856.5 µs856.5 µs

test_gesv[1000-c]

benchmark/pybench/benchmarks/bench_blas.py

188.6 ms188.6 ms

test_dgbmv[1-1000-s]

benchmark/pybench/benchmarks/bench_blas.py

75 µs75 µs

test_gemm[1000-s]

benchmark/pybench/benchmarks/bench_blas.py

117.4 ms117.4 ms

test_syrk[100-c]

benchmark/pybench/benchmarks/bench_blas.py

472.4 µs472.4 µs

test_dgemv[1000-c]

benchmark/pybench/benchmarks/bench_blas.py

14.8 ms14.8 ms

test_dgemv[1000-d]

benchmark/pybench/benchmarks/bench_blas.py

13.9 ms13.9 ms

test_gesv[1000-s]

benchmark/pybench/benchmarks/bench_blas.py

52.6 ms52.6 ms

test_syev[200-d]

benchmark/pybench/benchmarks/bench_blas.py

58.6 ms58.6 ms

test_syrk[100-s]

benchmark/pybench/benchmarks/bench_blas.py

213.4 µs213.4 µs

test_gesdd[mn1-s]

benchmark/pybench/benchmarks/bench_blas.py

65.2 ms65.2 ms

test_gesdd[mn1-d]

benchmark/pybench/benchmarks/bench_blas.py

93.9 ms93.9 ms

test_syrk[100-d]

benchmark/pybench/benchmarks/bench_blas.py

339.6 µs339.6 µs

test_gesv[100-s]

benchmark/pybench/benchmarks/bench_blas.py

257.1 µs257.1 µs

test_syev[50-d]

benchmark/pybench/benchmarks/bench_blas.py

1.4 ms1.4 ms

test_gesdd[mn0-d]

benchmark/pybench/benchmarks/bench_blas.py

120 µs120.2 µs

test_dgbmv[1-1000-z]

benchmark/pybench/benchmarks/bench_blas.py

118.8 µs119 µs

test_gesdd[mn0-s]

benchmark/pybench/benchmarks/bench_blas.py

109.1 µs109.2 µs

test_dgbmv[1-100-s]

benchmark/pybench/benchmarks/bench_blas.py

37.2 µs37.2 µs

test_dgbmv[1-100-c]

benchmark/pybench/benchmarks/bench_blas.py

40.2 µs40.3 µs

test_nrm2[100-d]

benchmark/pybench/benchmarks/bench_blas.py

-1%

38 µs38.5 µs

Commits

Click on a commit to change the comparison range

Base

develop

b6d5057

-0.15%

Add VORTEXM4

18f9582

2 months ago

by martin-frbg

Add VORTEXM4

ca542f3

2 months ago

by martin-frbg

Add VORTEXM4 to DYNAMIC_ARCH list

bf98e44

2 months ago

by martin-frbg

Relax version number requirement for AppleClang

4609732

2 months ago

by martin-frbg

Update SME-related kernels

107c883

2 months ago

by martin-frbg

+0.07%

Hide the local 2VLx2VL symbol as static is insufficient for this with gcc

edaa73f

2 months ago

by martin-frbg

-0.1%

Add VORTEXM4

1ee8879

2 months ago

by martin-frbg

+0.13%

Add registers d8 to d15 to clobber lists as the code does not expressly save them

b4fc09e

2 months ago

by martin-frbg

+0.03%

remove debugging printout

2b5d8c7

2 months ago

by martin-frbg

+0.02%

Merge branch 'develop' into issue5414

fc516af

12 days ago

by martin-frbg

-0.01%

Update Makefile.L3

b3d0bc4

11 days ago

by martin-frbg

+0.01%

Rework for DYNAMIC_ARCH use and use of SGEMM functions by SSYMM

c889558

11 days ago

by martin-frbg

+0.07%

Update limits based on benchmarking the SME code on Apple M4

47a66ae

5 days ago

by martin-frbg

-0.03%

Merge branch 'OpenMathLib:develop' into issue5414

9bfc361

20 hours ago

by martin-frbg

Home Terms Privacy Docs