Improve performance of vote() for SIMD_WIDTH = 16
Between the two optimizations, performance of decompressing a very large file on my Meteor Lake laptop was improved by ~11%.
Author ianromanick not on autobuild list. Waiting for curator authorization before starting CI build.
CI Vulkan-ExtensionLayer build queued with queue ID 442243.
CI Vulkan-ExtensionLayer build # 1050 running.
CI Vulkan-ExtensionLayer build # 1050 passed.
@vkushwaha-nv Do these changes look okay to you?
It has been a month with no activity. It is very frustrating that an 11% performance increase is being allowed to sit and rot.
Sorry I missed this earlier. Change looks good to me.
Going ahead with the merge since it has approval.
Thank you! :)