Oh look, #Nvidia makes CPUs now! And I got my hands on one! 
Today I benchmarked #FluidX3D on Nvidia's #GH200, both #GPU and #CPU with #PoCL. Finally I can answer the question: How does that exotic 2-chip #HPC APU show up in #OpenCL?
--> It's 2 separate devices, a GPU with 94GB @ 4TB/s and a 72-core CPU with 480GB @ 384GB/s. The NVLink interconnect between the two is much faster than PCIe, achieving ~380GB/s host<->device bandwidth, only limited by poor misaligned VRAM BW on the GPU or RAM BW.
When I say #FluidX3D #CFD runs on every toaster, I mean it. I finally got it running on the AMD Athlon X2 QL-65 dual-core CPU of my very first computer, a Toshiba Satellite L500D I got in 2009. The CPU itself is from 2008, a year before #OpenCL even existed. Modern #PoCL makes it compatible. Does close to 3 MLUPs/s! 
https://opencl.gpuinfo.org/listreports.php?devicename=cpu-x86-64-AMD+Athlon%28tm%29+X2+Dual-Core+QL-65&platform=
So I actually tried to give the #HSA #PoCL driver a go, and while I didn't actually get support for my #AMD integrated #GPU (but it should be doable) I actually discovered something interesting, which I hadn't noticed since I didn't compare the `clinfo` output for my iGP between #Rusticl and the proprietary driver.
So here's the weird thing: they report a different number of compute units!
I'm still moderately annoyed by the fact that there's no single #OpenCL platform to drive all computer devices on this machine. #PoCL comes close because it supports both the CPU and the #NVIDIA dGPU through #CUDA, but the not the #AMD iGPU (there's an #HSA device, but). #Rusticl supports the iGP (radeonsi) and the CPU (llvmpipe), but not the dGPU (partly because I'm running that on proprietary drivers for CUDA). Everything else has at best one supported device out of three available.
#PoCL 4.0 @openclapi Implementation Released With @IntelGraphics #oneAPI Level Zero Driver
https://www.phoronix.com/news/PoCL-4.0-Released
Original tweet : https://twitter.com/phoronix/status/1671820647540957185
9/9
The source code for the experimental @FluidX3D P2P is available in this branch on #GitHub: https://github.com/ProjectPhysX/FluidX3D/tree/experimental-p2p
The PR for #PoCL with cudaMemcpy is available here: https://github.com/pocl/pocl/pull/1189
Credit and many thanks to Jan Solanti from Tampere University for visiting me at University of Bayreuth and testing this together with me, in his endeavour to implement/optimize #PoCL-Remote.
Thanks to @ShmarvDogg for testing P2P mode on his 2x A770 16GB "bigboi" PC!8/9
As #Apple's #OpenCL implementation got deprecated and —I'm told— started degenerating towards uselessness, more people have started installing alternative ones through #Homebrew —typically PortableCL #PoCL.
This in turn returns an OpenCL library that is a #ICD loader —so people also install ocl-icd (again from Homebrew).
This poses an interesting conundrum for me as the #clinfo maintainer.
#PoCL 3.1 Released - Improved #SPIRV For CPU & #NVIDIA CUDA Drivers, WIP @VulkanAPI Driver
https://www.phoronix.com/news/PoCL-3.1-Released
Original tweet : https://twitter.com/phoronix/status/1599755854789562368
#PoCL 3.1-RC1 Released With Improved SPIR-V Support For CPU & CUDA Drivers, @VulkanAPI WIP
-- This "portable OpenCL" implementation continues improving.
https://www.phoronix.com/news/PoCL-3.1-RC1-Released
Original tweet : https://twitter.com/phoronix/status/1595366799540994049
