OpenMathLib
OpenBLAS
BlogDocsChangelog

Performance improvements of [SD]DOT with loop-unrolling on A64FX

#5358Merged
Comparing
iha-taisei:dot_unroll
(
f7ad906
) with
develop
(
36c2589
)
CodSpeed Performance Gauge
0%
Untouched
62

Benchmarks

Passed

test_syrk[100-d]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
340.8 µs339.2 µs
test_syrk[100-c]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
473.2 µs472 µs
test_dgbmv[1-100-c]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
39.9 µs39.8 µs
test_gesv[100-d]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
396 µs395.1 µs
test_dgbmv[1-1000-d]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
83.3 µs83.2 µs
test_gesv[100-c]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
695.8 µs695.3 µs
test_syev[50-d]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
1.4 ms1.4 ms
test_gesv[1000-d]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
93.3 ms93.3 ms
test_gemm[100-d]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
470.9 µs470.8 µs
test_dgemv[1000-s]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
7 ms7 ms
test_gemm[100-c]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
659.2 µs659.1 µs
test_gesv[100-s]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
256.5 µs256.5 µs
test_syev[200-d]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
58.6 ms58.6 ms
test_syev[50-s]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
1.3 ms1.3 ms
test_dgemv[100-z]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
230.6 µs230.6 µs
test_dgemv[1000-c]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
14.8 ms14.8 ms
test_syev[200-s]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
49.1 ms49.1 ms
test_syrk[1000-d]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
130.4 ms130.3 ms
test_gesdd[mn1-s]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
65.2 ms65.2 ms
test_gesdd[mn1-d]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
93.9 ms93.8 ms
test_dgemv[1000-z]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
26.3 ms26.3 ms
test_gemm[1000-s]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
117.4 ms117.4 ms
test_syrk[1000-c]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
227.5 ms227.5 ms
test_gesv[1000-c]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
188.6 ms188.6 ms
test_gesv[1000-z]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
353.6 ms353.6 ms
test_syrk[1000-z]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
476.4 ms476.4 ms
test_gemm[1000-d]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
239.4 ms239.4 ms
test_gemm[1000-c]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
426 ms426 ms
test_gesv[1000-s]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
52.6 ms52.6 ms
test_syrk[1000-s]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
65.4 ms65.4 ms
test_gemm[100-z]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
1.2 ms1.2 ms
test_dgemv[1000-d]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
13.9 ms13.9 ms
test_gemm[100-s]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
272.7 µs272.8 µs
test_syrk[100-s]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
212.9 µs213 µs
test_gemm[1000-z]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
875.2 ms875.6 ms
test_dgbmv[1-100-d]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
37.5 µs37.5 µs
test_gesv[100-z]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
936.8 µs937.4 µs
test_dgbmv[1-100-s]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
36.8 µs36.8 µs
test_dgbmv[1-1000-s]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
74.6 µs74.7 µs
test_nrm2[100-d]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
35.5 µs35.6 µs
test_syrk[100-z]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
855.3 µs856 µs
test_gesdd[mn0-s]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
108.7 µs108.8 µs
test_gesdd[mn0-d]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
119.5 µs119.7 µs
test_dgbmv[1-100-z]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
41.7 µs41.7 µs
test_dgbmv[1-1000-z]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
118.5 µs118.6 µs
test_dgbmv[1-1000-c]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
99 µs99.2 µs
test_daxpy[1000-d]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
31.9 µs31.9 µs
test_daxpy[1000-z]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
0%
40 µs40.1 µs
test_daxpy[100-d]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
-1%
23.7 µs23.8 µs
test_dot[1000]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
-1%
27.9 µs28 µs
test_dgemv[100-s]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
-1%
102.9 µs103.4 µs
test_daxpy[1000-s]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
-1%
26.9 µs27.1 µs
test_daxpy[100-z]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
-1%
25.2 µs25.4 µs
test_daxpy[100-c]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
-1%
24.5 µs24.7 µs
test_nrm2[100-dz]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
-1%
27.9 µs28.1 µs
test_daxpy[100-s]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
-1%
23.5 µs23.6 µs
test_dgemv[100-d]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
-1%
139.9 µs140.7 µs
test_dgemv[100-c]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
-1%
148.5 µs149.5 µs
test_nrm2[1000-d]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
-1%
30 µs30.2 µs
test_dot[100]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
-1%
21.9 µs22 µs
test_daxpy[1000-c]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
-1%
32.2 µs32.4 µs
test_nrm2[1000-dz]
benchmark/pybench/benchmarks/bench_blas.py
CodSpeed Performance Gauge
-1%
35 µs35.3 µs

Commits

Click on a commit to change the comparison range
Base
develop
36c2589
-0.14%
Performance improvements of [SD]DOT with loop-unrolling on A64FX
f7ad906
7 months ago
by iha-taisei
© 2026 CodSpeed Technology
Home Terms Privacy Docs