Skip to content

fp32 select

Eight independent f32 lanes, unrolled 4x, each stepping x = select(x + b, x - a, x > 1.0): a bounded walk whose direction is a runtime compare. Counted as 1 op per lane step (the select; compare and add/sub not counted).

Measured in OP/s. Higher values mean better performance. Benchmark your device.

Performance distribution

 
185 MOP/s55.7 GOP/s16.8 TOP/s

Performance thresholds

99% of results achieve at least 1.74 GOP/s

95% of results achieve at least 26.6 GOP/s

90% of results achieve at least 60.6 GOP/s

Median (50th percentile) 766 GOP/s

Best performer

Highest recorded performance: 16.8 TOP/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