Mojo function

warpgroup_reg_alloc

warpgroup_reg_alloc[count: Int]()

Allocates additional registers for the executing warp group.

This function provides a hint to the system to increase the maximum number of per-thread registers owned by the executing warp. It requests additional registers to increase the absolute per-thread maximum register count from its current value to the specified count.

Note: - Only supported on NVIDIA SM90+ GPUs - This is a performance optimization hint that may be ignored by the hardware - Should be paired with `warpgroup_reg_dealloc() when extra registers are no longer needed

Parameters:

• ​count (Int): The desired number of registers per thread. Must be: - A multiple of 8 - Between 24 and 256 (inclusive).

Was this page helpful?

Thank you! We'll create more content like this.

Thank you for helping us improve!

😔 What went wrong?

Some code doesn’t work

It includes inaccurate information

It's missing information I need

It was difficult to understand

Other

Submit