News for hackers. Stuff that matters.
Re:Simplifying GPU Application Development with HMM by pjmlp on Tuesday August 29, 2023 @09:07AM (#37307278)
> This new ability to directly read or write to the full application memory address space will significantly improve programmer productivity for all programming models built on top of CUDA: CUDA C++, Fortran, standard parallelism in Python, ISO C++, ISO Fortran, OpenACC, OpenMP, and many others.

This is the part of CUDA alternatives always miss when their models only support C and some C++ subset.

Re:Simplifying GPU Application Development with HMM by mratsim on Tuesday August 29, 2023 @12:08PM (#37309979)
Apple Metal does this though.
Re:Simplifying GPU Application Development with HMM by liuliu on Tuesday August 29, 2023 @12:31PM (#37310324)
Only recently. mmap file and directly use in Metal kernels is actually not supported until iOS 16 / macOS 13. Also, there are limited optimization opportunities around that and the recommended way seems still to use the specific Metal APIs to stream load assets from disk.
Re:Simplifying GPU Application Development with HMM by capableweb on Tuesday August 29, 2023 @12:19PM (#37310161)
Although Apple holds a <10% of total market share when it comes to computers, so not sure how helpful it is.
Re:Simplifying GPU Application Development with HMM by Me1000 on Tuesday August 29, 2023 @01:48PM (#37311378)
Well Nvidia holds like 90% of the GPU marketshare, so any reply mentioning a competitor would have this property.
Re:Simplifying GPU Application Development with HMM by pjmlp on Tuesday August 29, 2023 @01:18PM (#37310973)
You missed the polyglot description regarding which workloads CUDA supports.