On Thu, 6 Aug 2026 03:31:58 GMT, Jatin Bhateja <[email protected]> wrote:
>> Patch optimizes Float16 to integral conversion operations. Currently, its a >> two step process where by first a Float16 value is >> converted to a single precision floating point value followed by a >> conversion to an integral value. >> >> x86 targets supporting AVX512-FP16 feature (Intel Sapphire Rapids+ and >> upcoming AMD Zen6) provides direct instruction to convert a Float16 value to >> integral value. >> >> Following are the performance numbers of micro benchmark included with the >> patch on Granite Rapids with and without auto-vectorization. >> >> <img width="1125" height="636" alt="image" >> src="https://github.com/user-attachments/assets/ca6e6757-1579-475f-8307-9454c7c025c1"; >> /> >> >> Kindly review and share your feedback. >> >> Best Regards, >> Jatin >> >> --------- >> - [x] I confirm that I make this contribution in accordance with the >> [OpenJDK Interim AI Policy](https://openjdk.org/legal/ai). > > Jatin Bhateja has updated the pull request with a new target base due to a > merge or a rebase. The incremental webrev excludes the unrelated changes > brought in by the merge/rebase. The pull request contains nine additional > commits since the last revision: > > - Merge branch 'master' of http://github.com/openjdk/jdk into JDK-8382523 > - Review comments resolution > - Merge branch 'master' of http://github.com/openjdk/jdk into JDK-8382523 > - Review comments resolution > - Review comments resolution > - Review comments resolution > - Review comments resolution > - Review comments resolution > - 8382523: Optimize Float16 to integral conversion operations for > AVX512-FP16 targets test/micro/org/openjdk/bench/jdk/incubator/vector/Float16ToIntegralConvBenchmark.java line 60: > 58: i -> { > 59: if ((i % 100) == 0) { > 60: fp16inp[i] = float16ToRawShortBits(specialValues[i % > specialValues.length]); This comment comes a bit too late, but you will only ever have `i = k * 100`. And `specialValues.length = 5`. So you are really always picking `specialValues[0] = Float.NaN` here. It should be `specialValues[(i / 100) % specialValues.length]` instead. @jatin-bhateja @sviswa7 @missa-prime Could this have an impact on the performance measurement? I'm especially thinking about NaN/infinity fixup. ------------- PR Review Comment: https://git.openjdk.org/jdk/pull/30928#discussion_r3735927775
