[-] Capsicones 2 points 2 days ago

I'm stealing your idea.

[-] Capsicones 22 points 2 days ago

"compromise xz utils."

[-] Capsicones 51 points 2 days ago

I feel bad for Taiwanese politicians. They can't afford to provoke a major tantrum from Trump. They still need to purchase American arms, even if the US can't be counted on to help in an invasion. Some performative flattery and groveling might be in order. If it was me, I'd swallow my pride and cite the impossibility of moving entire chip production chains instead.

294
submitted 2 days ago* (last edited 2 days ago) by Capsicones to c/world@lemmy.world

To be America's enemy is dangerous; to be its friend, fatal.

[-] Capsicones 16 points 2 days ago* (last edited 2 days ago)

Russian soldiers come from poor regions: Siberia, the Far East, Kalmykia, etc. To many poor recruits the pay far exceeds anything they could get in civilian life. It's the same strategy as the US military: go to impoverished states and get those desperate poor schmucks to fight and die for you.

Russia has expanded the draft and seemingly can't count on volunteers as much as before. So the situation may change in the future.

[-] Capsicones 54 points 4 months ago

That's very good. Once I wanted to compile Firefox myself for some reason I no longer remember, but their Mercurial-based system was a hassle to work with. Most of us are already familiar with git. So, I know I'm going to be more inclined to make code contributions now that it uses git.

Just wish they could've chosen another git-based option like Codeberg, or even an internally-hosted server. I'm rather wary of GitHub/Microsoft swallowing up so many open source projects.

[-] Capsicones 21 points 7 months ago* (last edited 7 months ago)

The paper was published by IEEE and with professors as co-authors. Only the second author is a student. And I wouldn't dismiss it out of hand like that because of a magazine article. Students come up with breakthroughs all the time. The paper itself says it disproves Yao's conjecture. I personally plan to implement and benchmark this because the results seem so good. It could be another fibonacci heap situation, but maybe not. Hash tables are so widely used, that it might even be worthwhile to make special hardware to use this on servers, if our current computer architecture is only thing that holds back the performance.

Edit: author sequence

[-] Capsicones 22 points 8 months ago

Some commenters on this post are clearly not aware of PTX being a part of the CUDA environment. If you know this, you aren't who I'm trying to inform.

[-] Capsicones 124 points 8 months ago

There seems to be some confusion here on what PTX is -- it does not bypass the CUDA platform at all. Nor does this diminish NVIDIA's monopoly here. CUDA is a programming environment for NVIDIA GPUs, but many say CUDA to mean the C/C++ extension in CUDA (CUDA can be thought of as a C/C++ dialect here.) PTX is NVIDIA specific, and sits at a similar level as LLVM's IR. If anything, DeepSeek is more dependent on NVIDIA than everyone else, since PTX is tightly dependent on their specific GPUs. Things like ZLUDA (effort to run CUDA code on AMD GPUs) won't work. This is not a feel good story here.

[-] Capsicones 24 points 9 months ago

And I thank you, Norah. As an Asian woman, any Linux space can feel pretty unwelcoming sometimes. Most of the time it's the sexism, but this insistence on saying "ricing" is just another reminder that many in this space enjoy a bit of racism on the side, too.

I don't usually say anything; I'm personally too afraid of being dragged into an endless "debate". Perhaps a bit cowardly on my part. So, I appreciate you pointing it out first.

[-] Capsicones 20 points 9 months ago

It's still racist when you make it an acronym. We know the term originally came from a racist term for Asian vehicles.

[-] Capsicones 27 points 2 years ago* (last edited 2 years ago)

Machine learning pays my bills, and I never had a choice on my graphics card brand. To be sure, I wanted an AMD for the open source drivers, but CUDA remains essential to me. RocM support from AMD is a joke, and isn't anywhere close to an alternative. Reseachers release code that only runs on CUDA for a good reason. To say that I don't get to complain is going too far

view more: next ›

Capsicones

joined 2 years ago