On Mon, 22 Jun 2026 09:37:01 GMT, Andrew Dinn <[email protected]> wrote:
> However, it might still be the case that when employed in combination with > the multiply intrinsic it is of benefit - possibly also depending on what > hardware we are running on. I agree with that, but it's surely worth a look at the generated code so see _why_ hand-coded `benchMultiply` runs slower. Looking at the Java implementation, it might simply be that C2 generates near-perfect code. I just had a look, and it's not at all bad. ------------- PR Comment: https://git.openjdk.org/jdk/pull/30941#issuecomment-4771497820
