swerner

6 ms·

1.
by swerner 8d ago
Proximity to borders leaves the government the possibility that if it really, really wanted to search a property, they'd just have to cede some land in a 100mile radius to a foreign country and instantly create an international border
2.
by swerner 12d ago
If your stack is open source, why not? I’ve had good results with pointing an LLM to a debug build of the GPU compiler when I saw it crash. It just seems you need to periodically remind them that gdb exists to prevent it from littering your
3.
by swerner 13d ago
Wait, I have an idea: What is your qualification for talking about graphics programming? If you cannot provide credentials, I will have no other choice than to claim you don’t know what you’re talking about, based on the criteria by which
4.
by swerner 13d ago
Not everything is AI and dot products of massive vectors, there are still applications that do other maths on GPUs My thinking was rather that most of our current programming languages put memory layout fully into the programmer’s responsib
5.
by swerner 13d ago
Vulkan compute is not really designed or intended to be a CUDA competitor, its feature set is much more restricted, and Vulkan host side code is much more verbose than CUDA. OpenCL or SYCL are much closer in features to CUDA. I found that w
6.
by swerner 13d ago
For a performant portable language, we’d have to go to a higher level where you describe what to do and leave the how to do to the compiler. It would then need to be able to adjust memory layout, access patterns, data type choice to the und
7.
by swerner 13d ago
Yes, unfortunately. Otherwise we could just implement all of SYCL and OpenCL on top of Vulkan and live happily ever after.
8.
by swerner 13d ago
TIL, I didn’t know that. I always assumed it came from level zero”, the Intel computer layer that ZLUDA was translating to before its developer was hired by AMD to target HIP.
9.
by swerner 13d ago
What domain do you think I was talking about?
10.
by swerner 13d ago
Oh, absolutely. Nvidia is the modern day “nobody gets fired for buying IBM”. The reason we’re still using Unix is not because it’s the best, but because it had to much inertia to let any alternative become its successor. Similarly, C and H
11.
by swerner 13d ago
Unfortunately, many of very good ideas end with “it’s open source, anyone can contribute” because very few actually do.
12.
by swerner 13d ago
Most modern graphics is compute. Pixar, Dreamworks, Sony, etc do not use Vulkan to render their movies. It’s CPUs or CUDA. “graphics people do not (and cannot) write CUDA kernels/shaders” is just not true at all. All it would take to v
13.
by swerner 13d ago
Outsider to what?
14.
by swerner 13d ago
If it helps you at all: “if you've never debugged a GPU synchronization error or kernel panic due to some GPU firmware bug or fought absolute shit profilers hunting for perf” I have done all of those things. As part of my full time job
15.
by swerner 13d ago
I don't think we're at a point yet where anyone would trust ZLUDA enough to ship commercial products that rely on it. I would be delighted though, if anyone can prove me wrong.
16.
by swerner 13d ago
I could post the GitHub URLs of all the shader code I wrote that’s running on countless GPUs right now, but what would it change? I’m still just a random guy on the internet with an opinion that happens to be different from your opinion. Yo
17.
by swerner 13d ago
To my knowledge, SPIR-V on Intel will stay, and be it only because it’s part of the OpenCL and Vulkan standards.
18.
by swerner 13d ago
They were maintained by Codeplay - paid for my Intel. Nvidia can make contributions anytime they want, and here is the problem: Nvidia does not want to. Until each vendor starts pitching in with contributing their backend to an open standar
19.
by swerner 13d ago
I’m too old to participate in internet pissing contests.
20.
by swerner 13d ago
If that is your standard, I do have an idea what I’m talking about.
21.
by swerner 13d ago
The problem with OneAPI is naming. It leads people to believe that is another competing standard where in fact is is simply just an implementation of a standard compliant SYCL compiler. If it just had been named “Intel SYCL compiler”, simil
22.
by swerner 13d ago
What do you mean by holistic? SYCL is an open versioned standard that allows for vendor specific extensions. The problem is not that there isn’t a proper standard, the problem is that many hardware vendors - or software developers simply do
23.
by swerner 13d ago
AI will take down Nvidia’s moat. When it becomes trivial to translate CUDA/PTX to HIP, SYCL or Metal, CUDA is no longer the moat, it becomes the intermediate representation.
24.
by swerner 16d ago
Life saving feature for ppl with ADHD. M
25.
by swerner 1mo ago
It may have a different connotation in the US than in Europe. In US media I’ve only heard the term being used recently, in Europe it’s been around for decades. In Europe, “Antifascist” is a label that people groups use to describe themselve
26.
by swerner 1mo ago
They’ve been doing this since the post-9/11 DHS initiative, at least.
27.
by swerner 1mo ago
I tried to get Copilot to port the Linux drivers for my HP CP1025 to macOS, but it failed. Even modern Linux can’t print to it any more, I had to put an older distribution on a RasPi and now I’m printing at reduced quality. It was crazy how
28.
by swerner 2mo ago
Cycles is getting direct code contributions from all major hardware vendors.
29.
by swerner 2mo ago
Unfortunately, Vulkan Compute doesn’t to all the things that OpenCL, SYCL, HIP or CUDA do.
30.
by swerner 5mo ago
https://burningman.org/global-events-groups/

More ›