On Tue, 17 Dec 2024 11:05:27 GMT, Jatin Bhateja <jbhat...@openjdk.org> wrote:
>> Are we sure that the rounding behaviour of float is the correct behaviour >> for Float16? I would like to see some examples where rounding matters. > > FP16 has 11 bits precision and FP32 has 24 bit precision, thus as per [2P > rule ](https://dl.acm.org/doi/pdf/10.1145/221332.221334) the operation is > innocuous to double rounding effects. In addition, fall back implementation > of Float16.divide also takes the same route of performing the operation at > FP32 granularity. I understand. Thanks for the explanation. But the backend `Float16` instructions do not convert to float, right? So we need to be sure, and test that rounding works correctly in all cases. Can you point me to a test for that? ------------- PR Review Comment: https://git.openjdk.org/jdk/pull/22754#discussion_r1888882880
