[pjmlp]

Most of these "alternatives" focus on CUDA C++, and overlook what actually makes CUDA interesting.

Already in 2020,

https://developer.nvidia.com/blog/cuda-refresher-the-gpu-com...

[mschuetz]

> Ease of programming and a giant leap in performance is one of the key reasons for the CUDA platform’s widespread adoption

This, so much. Other platforms continue to ignore developer UX, but it's one of the main things that get's new users onboard and keeps old users around.

[msond]

We're actually targeting all of it, and not just CUDA C++.

[pjmlp]

Including stuff like Fortran, Haskell, Java, .NET via PTX, Python JIT, IDE tooling integration with major IDEs, graphical GPU debugging and profiling, libraries and co?

Then I guess all the best.

[zorked]

This post has some serious peanut-gallery vibes.

[pjmlp]

Peanut-gallery is happily using CUDA, and needs actual sound reasons to move.

[account42]

Then the peanut gallery has nothing to complain when Nvidia jacks up prices.

[anon291]

They don't though. Money for ai is pretty much free in America to anyone who can demonstrate a modicum of competence in the field.

The only people who are without access are students or hobbyists really.

[pjmlp]

Do you see me complaining?

Here is a tip, you don't always need to suffer from FOMO and get the very latest model card.

In fact, contrary to the competition, one can play with CUDA even on laptops, go figure.

[ychen306]

How do you deal with target-specific inline asm like tcgen05.mma?

[msond]

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 done already for the hopper-class tensorcore instructions, and the blackwell ones will map similarly, though likely with a kernel launch involved.

[embedding-shape]

Ambitious but neat, good luck if nothing else :)

If you were to guess, when do you think your Nsight Compute alternative might be ready with your own toolchain?

[msond]

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 coverage of the core CUDA APIs and will be shifting an increasing amount of effort towards developer tooling.