ppanchad-amd

Results 674 comments of ppanchad-amd

Inline asm translation

@microyybar Is this issue resolved? If so, please close ticket. Thanks!

[Question] gfx1032 support

@dc-dc-dc gfx 1032 is not supported on HIP SDK with the latest version 5.7.1. Please see (https://rocm.docs.amd.com/projects/install-on-windows/en/latest/reference/system-requirements.html). Thanks!

question about cache line size

@jinz2014 Is this issue resolved for you? Thanks!

Compiler warnings for g++13.1

@torrance Can you please test with latest ROCm 6.1.0 (HIP 6.1)? If resolved, please close ticket. Thanks!

Query Regarding Fast-Math Optimization Support on AMD GPUs

@xinyi-li7 Can you please test with latest ROCm 6.1.0 (HIP 6.1)? If resolved, please close ticket. Thanks!

Query Regarding Fast-Math Optimization Support on AMD GPUs

@xinyi-li7 -ffast-math is now available in the latest ROCm 6.1.1 Thanks!

`__half2` implementation does not match CUDA one, resulting in a bug in `__half2float` calls

@fxmarty Can you please test with latest ROCm 6.1.0 (HIP 6.1)? If resolved, please close ticket. Thanks!

hipblasdgemm not getting close to peak

@JorgeG94 Can you please test with the latest ROCm 6.1.2? If your issue is resolved, please close the ticket. Thanks!

undefined symbol: Tensile::TypedContractionInputs<Tensile::Half>::TypedContractionInputs()

@xiakubaobaore Apologies for the lack of response. Can you please test with the latest ROCm 6.1.2? If issue is resolved, please close the ticket. Thanks!

kernel.cpp without assembly kernel implement

@DoubleClark Do you still need assistance with this ticket? If not, please close the ticket. Thanks!