Skip to content

packed_4x8_integer_dot_product#

Real-world WebGPU WGSL language feature support statistics for packed_4x8_integer_dot_product across browsers, devices, and platforms.

Adds WGSL language support for packed 4x8 integer dot-product operations. These operations combine several small integers efficiently and are useful in graphics and ML-style math.

Want to check this browser directly? Open the WebGPU Capabilities Report.

Overall Support#

This chart shows what share of recent reports support packed_4x8_integer_dot_product.

98.93%
overall support

Platform Support#

This table shows how support for packed_4x8_integer_dot_product varies by platform.
PlatformSupport
Android97.34%
Chromium OS99.57%
iOS99.86%
Linux97.38%
Mac OS99.66%
Windows99.06%

Browser Support#

This table shows how support for packed_4x8_integer_dot_product varies by browser family.
BrowserSupport
Chrome/Chromium98.87%
Edge99.82%
Firefox100%
Opera99.9%
Other89.73%
Safari99.91%
Samsung Internet97.84%

Vendor Support#

This table shows how support for packed_4x8_integer_dot_product varies by GPU vendor across all platforms.
VendorSupport
amd99.84%
apple99.98%
arm99.89%
google100%
huawei100%
img-tec100%
intel99.61%
intelr100%
microsoft85.71%
nvidia99.82%
qualcomm99.76%
samsung100%

Vendor Architecture Support#

This table shows how support for packed_4x8_integer_dot_product varies by reported GPU vendor and architecture across all platforms.
Vendor - ArchitectureSupport
amd - gcn-499.71%
amd - gcn-599.55%
amd - rdna-1100%
amd - rdna-2100%
amd - rdna-3100%
apple - apple100%
apple - metal-3100%
arm - bifrost99.93%
arm - valhall99.86%
intel - gen-1199.94%
intel - gen-12lp99.77%
intel - gen-898.09%
intel - gen-999.5%
intel - xe-2lpg100%
intel - xe-lpg100%
nvidia - ampere99.87%
nvidia - blackwell100%
nvidia - kepler98.57%
nvidia - lovelace99.85%
nvidia - maxwell99.69%
nvidia - pascal99.63%
nvidia - turing100%
qualcomm - adreno-6xx99.79%
qualcomm - adreno-7xx99.77%
qualcomm - adreno-8xx100%