On Mon, 27 Oct 2025 15:19:48 GMT, Jatin Bhateja <[email protected]> wrote:
> Add new HalffloatVector type and corresponding concrete vector classes in
> addition to existing primitive vector types, maintaining operation parity
> with the FloatVector type.
> - Add necessary inline expander support.
> - Enable intrinsification for a few vector operations, namely
> ADD/SUB/MUL/DIV/MAX/MIN/FMA.
> - Use existing Float16 vector IR and backend support.
> - Extended the existing VectorAPI JTREG test suite for the newly added
> HalffloatVector operations.
>
> The idea here is to first be at par with Float16 auto-vectorization support
> before intrinsifying new operations (conversions, reduction, etc).
>
> The following are the performance numbers for some of the selected
> HalfflotVector benchmarking kernels compared to equivalent
> Float16OperationsBenchmark kernels.
>
> <img width="837" height="650" alt="{A2BA2D85-085A-489F-8DDD-0FCFB5986EA5}"
> src="https://github.com/user-attachments/assets/2f34138e-80f4-4870-9b5f-f64bf43c1b11";
> />
>
> Initial RFP[1] was floated on the panama-dev mailing list.
>
> Kindly review the draft PR and share your feedback.
>
> Best Regards,
> Jatin
>
> [1] https://mail.openjdk.org/pipermail/panama-dev/2025-August/021100.html
We already have a lot of things in the codebase now from previous issues that
use `HF` everywhere, for example some node names, and the type. Should we maybe
rename all of them to `F16`, or something else? Open question, not sure of the
answer yet.
-------------
PR Comment: https://git.openjdk.org/jdk/pull/28002#issuecomment-3516579087
