Skip to content
SIMD realization matrix

simd: mul_add rounds once on every backend (wasm emulation, exact scalar FMA, no libm) #196

Sign in to view logs

simd: mul_add rounds once on every backend (wasm emulation, exact scalar FMA, no libm)

simd: mul_add rounds once on every backend (wasm emulation, exact scalar FMA, no libm) #196

Run time

Learn about OS pricing on GitHub Actions
Job Run time
realization/avx2 × x86_64
1m 55s
cpu_guard × library host
1m 50s
realization/avx512 × x86_64
1m 44s
realization/neon × aarch64
2m 16s
realization/nightly × x86_64
2m 2s
realization/avx × x86_64
3m 41s
realization/host-native × x86_64 (informational)
1m 14s
realization/wasm + scalar × wasm32
1m 48s
realization/v2 × x86_64
3m 16s
19m 46s