Scheduling of EU hardware threads to a subslice is "fair"

OpenCL* for CPU

Ask questions and share information on Intel® SDK for OpenCL™ Applications and OpenCL™ implementations for Intel® CPU.

Scheduling of EU hardware threads to a subslice is "fair"

347 Views

I was curious to see how a local work group was distributed across a subslice.

My assumption was that a cooperative group of SIMDxx hardware threads would be assigned to the least occupied EU.

That seems to be the case.

Here is a plot showing a SIMD8 kernel being launched with local work groups of size: 224, 128, 64, 32, 16, 8:

Link Copied

0 Replies

Community support is provided during standard business hours (Monday to Friday 7AM - 5PM PST). Other contact methods are available here.

Intel does not verify all solutions, including but not limited to any file transfers that may appear in this community. Accordingly, Intel disclaims all express and implied warranties, including without limitation, the implied warranties of merchantability, fitness for a particular purpose, and non-infringement, as well as any warranty arising from course of performance, course of dealing, or usage in trade.

For more complete information about compiler optimizations, see our Optimization Notice.