fp32 pow
Eight independent scalar f32 pow-add chains per thread, unrolled 4x. pow(x, e) for a non-integer e is usually exp2(e * log2(x)) under the hood — several instructions — so expect this well below sqrt/div throughput.
Measured in FLOP/s. Higher values mean better performance. Benchmark your device.
Performance distribution
233 MFLOP/s41.7 GFLOP/s7.48 TFLOP/s
233 MFLOP/s41.7 GFLOP/s7.48 TFLOP/s
Performance thresholds
99% of results achieve at least 489 MFLOP/s
95% of results achieve at least 12.3 GFLOP/s
90% of results achieve at least 31.7 GFLOP/s
Median (50th percentile) 401 GFLOP/s
Best performer
Highest recorded performance: 7.48 TFLOP/s. View benchmark profile.
Recorded 9/14/2026, 2:52:24 PM
WebGPU device
Vendornvidia
Device—
Architectureblackwell
Description—
WebGL device
VendorGoogle Inc. (NVIDIA)
RendererANGLE (NVIDIA, NVIDIA GeForce RTX 5090 (0x00002B85) Direct3D11 vs_5_0 ps_5_0, D3D11)
Browser
NameChrome
Version153.0.0.0
Operating system
NameWindows
Version10