UXL Foundation readying alternative to Nvidia's CUDA for this year The UXL Foundation is readying its open standard accelerator programming model, touted by some as an alternative to Nvidia's CUDA platform, for "a spec release in Q4." Announced last year, the Unified Acceleration (UXL) Foundation is a group of companies operating under the aegis of the Linux Foundation to develop an open …
It is and has been for some time obvious that AMDs ROCm is a failure. After Lisa Su came aboard in 2015 or so and gutted the Fusion and HSA program to start over with ROCm it has been a disaster. Yes she has shepherd in Ryzen/EPYC, Infinity Architecture and Xilinx. AMD hardware is second to none in the X86-64 world. But their software work is very sub-par. I recently stated on Phoronix that at this point AMD should just abandon ROCm and adopt Intel’s oneAPI and their entire compute stack. Maybe UXL can be the bridge in which AMD walks away from ROCm and cuts their losses.
AMDs software has been crap, to the extent of fundamental bugs in the fft lib being reported and unfixed for 2years.
They can't get onboard with Intel's API if Intel are competing with AMD for graphics card sales.
And nobody is going to get behind an open standard that doesn't give them the advantage everyone else implementing the standard, so it becomes a box tick 'opencl support(*)' in the same way that Windows 'supported' OpenGL and POSIX.
* for minimum values of support
> And nobody is going to get behind an open standard that doesn't give them the advantage everyone else implementing the standard
Erm, Vulcan? It was obvious to everyone that multiple diverging standards in the graphics space was hurting everyone.
What you say may be true for AI compute at the moment with there being still rapid development, but I don't think it will be long before the software stack starts to become somewhat commoditised and both developers and manufacturers will want converge on something to avoid the ongoing cost of maintaining it. For NVidia, CUDA is a means to an end and that end is selling hardware. It's only really popular now because it is the only big game in town and people have invested in it.
Once something useful emerges that is cross-platform, people will move to it in droves. Vulcan was like that. It was in the pipeline for quite a while, then once it became mature, everyone important jumped onto it.
Vulkan is only relevant on Android, as it is its main 3D API (nowadays even GL ES runs on top of it), and Linux.
Zero usage on game consoles PS uses LibGNM and XBox naturally DX, with exception of the Switch where the main API is NVN anyway.
On Windows, it is mostly used by folks that also want to target The Linux Desktop with their products.
And on Apple there is Metal.
It is basically Android, and Valve emulating DirectX via Vulkan on SteamDeck.
And the worst part is that I'd love to see an alternative to CUDA to exist, but our desires should not impede an impartial analysis...
The most glaring example of past initiatives failing is OpenCL. Between 1.0 and 2.2 it was painfully slow, so slow in fact, that Apple deprecated it, after all the work they did to spearhead it, and integrat it as a frist class citizen in OSX/MacOS (for instance, one could use Grand Central dispatch to issue OpenCL tasks as easily for OpenCL as for CPU).
And then came the clusterFSCK that was OpenCL 3.0, where the mandatory baseline is OpenCL 1.2 (that is, in fact, a regression) and then everything else is OPTIONAL. Which means, is super hard to write hardware agnostic code in OpenCL 3.0, as not all functions will be supported by all manufacturers.
And remember, beyond AMD, Intel and nVidia, there is Qualcomm (Adreno), ARM (Mali), PowerVR, VIA (as S3), and the Only-China duo of Imaginnation Technologies (similar but not equal to PowerVR) and MooreThreads. Pleanty of hardware to choose if you want to write OpenCL code. And please also remmeber GPU Accelerated code is not only used for ML/AI, or CFD, or HPC, but also for day to day tasks (Like calculating indexes and hash codes in databases). Also, remmeber that many of the "Lesser" GPU architectures are used in ARM servers
Similar examples exist with things like SYCL, and, as the previous comentor wrote, AMD doing FOSS GRaphics and CUDA-Killers as knee-jerk reactions and changing things every few years... AMD has been bussy changing their "CUDA Killer" architecture (and the Accelerated Graphics one too) every few years. Close To Metal, Mantle, Stream, GPUOpen, HIP, ROCm... ¿Does any one of those ring a bell?
Instead of AMD's engineers heeding the suba divers advice:
Stop-Think-Act. If I can do that 500m inside a Cavern, why AMD's fellow engineers can not do it sitting in a conference room is beyond me...
So, TLDR: While I would like to have an alternative to CUDA, past experience says that these kind of initiatives fail, Is likely that this one will fail as well.
Apart from your scathing takedown of AMD which I alluded to above and wholeheartedly agree, I disagree with you that UXL will fail . Here’s why. Google…Intel…ARM. Between those three you cover the entirety of X86 and ARM ISA for hyperscalers down to Edge and local computing and IoT. There is actually very little outside Gigawatt Cloud Warehouses that can support Nvidia’s wares but there are quite literally billions and billions more to come low power compute platforms that need acceleration and support for on board if not on die accelerators where there is absolutely no need for CUDA or the power requirements that an Nvidia GPU requires to run said stack. It would be trivial for AMD to walk away from yet ANOTHER in house GPU/Accelerator/Compute stack because over the last 30 years they’ve become good at it ( remember 3D Now! from the 90s ?) and retool for UXL and by extension Intel’s oneAPI stack which is the basis for UXL for both X86 and ARM. Which makes sense seeing as how Intel’s FPGAs and AMDs Xilinx FPGA’s are ARM based. AMD can still compete with Intel over hardware optimizations that give better performance with Intel’s own software than even Intel can do such as what EPYC already does. And cheaper as well.
I suspect it's like the days of Unix workstations, Sun / SGI / HP-UX / IBM RS6000 all competed to be the best proprietary Unix to lock you into their propriety hardware.
And all got wiped out by Linux - but it would have been instant death if any of them had embraced Linux and still allowed it to run on other people's cheap hardware. But Linux only really replaced them when you had powerful commodity h/w to run it on - which only happened because Windows Servers created the market.
So to replace CUDA with OpenCL HIP or Open-flavour-of-the-day we need a market of commodity massively powerful non-NVidia GPUs to run it on, and if the only maker of massively powerful GPUs is NVidia cos of CUDA it's a bit of a Catch-0x16
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
