msond

11 ms·

1.
by msond 2mo ago
We'd love to know which libraries your production code falls over on! Feel free to ping us on discord ( https://discord.com/invite/KNpgGbTc38 ) or you can find my e-mail in my profile if that's preferable
2.
by msond 2mo ago
We haven’t yet released support for tcgen05, but we’ll deal with it the same way we deal with other inline PTX: parsing it and converting it to target-appropriate instructions together with the rest of the program. This is something we’ve d
3.
by msond 2mo ago
We actually support NVIDIA hardware, too. In some benchmarks, SCALE beats nvcc, and we have compiler optimizations in the pipeline that will improve those numbers over time. > If all you want is to be able to easily use non-NVidia hardwa
4.
by msond 2mo ago
We have a comparison page: https://docs.scale-lang.com/stable/manual/comparison/#zluda
5.
by msond 2mo ago
A guess would be some time next year — since our public launch our focus has generally been on API coverage and increasingly recently, on performance. While performance improvements will always remain a target, we're soon at full cover
6.
by msond 2mo ago
Actually we launched in 2024 and the last message in our discord is definitely not that: https://discord.gg/KNpgGbTc38
7.
by msond 2mo ago
We're actually targeting all of it, and not just CUDA C++.
8.
Why GPU compilers are MORE important in the agentic era(scale-lang.com)
9 points by msond 4mo ago|0 comments
9.
by msond 1y ago
Hey there! Would love to connect and hear more about what you like about SCALE and where you'd like it to go. AMD is a part of our strategy, but it's not the end-game - we envision SCALE to be vendor neutral and have plans to supp
10.
CUDA version of GROMACS is faster on AMD than HIP port(scale-lang.com)
15 points by msond 1y ago|3 comments
11.
Scale 1.3: Run CUDA Natively on MI-Instinct GPUs(docs.scale-lang.com)
7 points by msond 1y ago|1 comments
12.
by msond 1y ago
With the latest release of SCALE, wave64 AMD Instinct GPUs are supported
13.
by msond 2y ago
Very interesting, thank you for sharing! We do believe in open source software and we do want to move the GPGPU market away from fully closed languages. The future is open for discussion but regardless, the status-quo at the moment is a pro
14.
by msond 2y ago
We're still thinking about our approach but this is a nice suggestion, thank you. I'm curious, for what reasons are you interested in the source code yourself?
15.
by msond 2y ago
SCALE is not a "translation layer", it's a full source-to-target compiler from CUDA-like C++ code to AMD GPUs. See this part of the documentation for more details regarding warp sizes: https://docs.scale-lang.com&#
16.
by msond 2y ago
We have not reverse engineered any compiled code in the process of developing SCALE. It was clean-room implemented purely from the API surface and by trial-and-error with open CUDA code.
17.
by msond 2y ago
We're going to be publishing more details on later blog posts and documentation about how this works and how we've built it. Yes, we're not open source, however our license is very permissive. It's both in the software d
18.
by msond 2y ago
Actually, we use mkdocs and the excellent material for mkdocs theme: https://squidfunk.github.io/mkdocs-material/
19.
by msond 2y ago
We're putting together benchmarks to publish at a later time, and we've asked some independent third parties to work on their own additionally.
20.
by msond 2y ago
You can learn more about us on https://spectralcompute.co.uk