mightsleep opened a new issue, #11542: URL: https://github.com/apache/arrow-rs/issues/11542
Follow-up to #11322. In the LLVM lowering of the same two forms (llvm/llvm-project#227292), Neoverse-N2 ran 64-bit pext in 5.97 ns with the whole-word network against 8.49 ns bytewise (llvm-mca: 19 vs 35 cycles on N2, 19 vs 33 on V2). Apple M1 went the other way. To check inside compress on Neoverse and pick by target_arch. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
