search

oneapi-src / oneDPL

oneAPI DPC++ Library (oneDPL) https://software.intel.com/content/www/us/en/develop/tools/oneapi/components/dpc-library.html
Apache License 2.0
715 stars 112 forks source link

Some improvements of struct reduce_over_group #1666

Closed SergeyKopienko closed 1 week ago

SergeyKopienko commented 1 week ago

@julianmi could you please check some perf improvement with these changes?

julianmi commented 1 week ago

@julianmi could you please check some perf improvement with these changes?

The performance on PVC seems to be unchanged for 4-byte data. Large data types shouldn't see a change in performance either since they are passed by reference like on mainline.

SergeyKopienko commented 1 week ago

@julianmi could you please check some perf improvement with these changes?

The performance on PVC seems to be unchanged for 4-byte data. Large data types shouldn't see a change in performance either since they are passed by reference like on mainline.

Thanks a lot, @julianmi, for this update!. The idea, unfortunately, was not useful.

SergeyKopienko commented 1 week ago

The idea, unfortunately, was not useful.