vegeta@lemmy.world to Technology@lemmy.worldEnglish · 2 years agoDeepSeek's AI breakthrough bypasses industry-standard CUDA, uses assembly-like PTX programming insteadwww.tomshardware.comexternal-linkmessage-square36linkfedilinkarrow-up1341arrow-down113cross-posted to: technology@lemmy.zip
arrow-up1328arrow-down1external-linkDeepSeek's AI breakthrough bypasses industry-standard CUDA, uses assembly-like PTX programming insteadwww.tomshardware.comvegeta@lemmy.world to Technology@lemmy.worldEnglish · 2 years agomessage-square36linkfedilinkcross-posted to: technology@lemmy.zip
minus-squarefilister@lemmy.worldlinkfedilinkEnglisharrow-up61arrow-down3·2 years agoWhat is amazing in this case is that they achieved spending a fraction of the inference cost that OpenAI is paying. Plus they are a lot cheaper too. But I am pretty sure that the American government will ban them in no time, citing national security concerns, etc. Nevertheless, I think we need more open source models. Not to mention that NVIDIA also needs to be brought to earth.
minus-squaredemesisx@infosec.pubBanned from communitylinkfedilinkEnglisharrow-up27arrow-down5·edit-25 months agodeleted by creator
minus-squaretoffi@feddit.orglinkfedilinkEnglisharrow-up19·2 years agoNever forget kids the market can stay irrational much longer than you can stay solvent.
minus-squaredemesisx@infosec.pubBanned from communitylinkfedilinkEnglisharrow-up15·edit-25 months agodeleted by creator
minus-squarePieisawesome@lemmy.worldlinkfedilinkEnglisharrow-up13·2 years agoIt’s written in nvidia instruction set PTX which is part of CUDA ecosystem. Hardly going to affect nvidia
minus-squareEager Eagle@lemmy.worldlinkfedilinkEnglisharrow-up3·2 years agoI wish that was true, but this doesn’t threaten any monopoly
minus-squaredemesisx@infosec.pubBanned from communitylinkfedilinkEnglisharrow-up13arrow-down1·edit-25 months agodeleted by creator
minus-squareEager Eagle@lemmy.worldlinkfedilinkEnglisharrow-up6·edit-22 years agomate, that means they are using PTX directly. If anything, they are more dependent to NVIDIA and the CUDA platform than anyone else. to simplify: they are bypassing the CUDA API, not the NVIDIA instruction set architecture and not CUDA as a platform.
minus-squaredemesisx@infosec.pubBanned from communitylinkfedilinkEnglisharrow-up13·edit-25 months agodeleted by creator
minus-squareEager Eagle@lemmy.worldlinkfedilinkEnglisharrow-up5·edit-22 years ago Until last week, you absolutely NEEDED an NVidia GPU equipped with CUDA to run all AI models. also not true
minus-squaredemesisx@infosec.pubBanned from communitylinkfedilinkEnglisharrow-up12·edit-25 months agodeleted by creator
What is amazing in this case is that they achieved spending a fraction of the inference cost that OpenAI is paying.
Plus they are a lot cheaper too. But I am pretty sure that the American government will ban them in no time, citing national security concerns, etc.
Nevertheless, I think we need more open source models.
Not to mention that NVIDIA also needs to be brought to earth.
deleted by creator
Never forget kids the market can stay irrational much longer than you can stay solvent.
deleted by creator
It’s written in nvidia instruction set PTX which is part of CUDA ecosystem.
Hardly going to affect nvidia
I wish that was true, but this doesn’t threaten any monopoly
deleted by creator
mate, that means they are using PTX directly. If anything, they are more dependent to NVIDIA and the CUDA platform than anyone else.
to simplify: they are bypassing the CUDA API, not the NVIDIA instruction set architecture and not CUDA as a platform.
deleted by creator
also not true
deleted by creator