fp16 rsqrt
Eight independent scalar f16 inverseSqrt-add chains per thread, unrolled 4x. Same shape as f32-rsqrt but entirely in half precision; compare against f16-sqrt to see the gap.
Measured in FLOP/s. Higher values mean better performance. Benchmark your device.
Performance distribution
47.3 MFLOP/s26.2 GFLOP/s14.5 TFLOP/s
47.3 MFLOP/s26.2 GFLOP/s14.5 TFLOP/s
Performance thresholds
99% of results achieve at least 3.48 GFLOP/s
95% of results achieve at least 31 GFLOP/s
90% of results achieve at least 80.2 GFLOP/s
Median (50th percentile) 713 GFLOP/s
Best performer
Highest recorded performance: 14.5 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