threelonmusketeers@sh.itjust.worksM to Tesla@lemmy.worldEnglish · 6 months agoElon Musk asked Nvidia to prioritize GPU shipments to X over Tesla, emails revealelectrek.coexternal-linkmessage-square7fedilinkarrow-up140arrow-down13
arrow-up137arrow-down1external-linkElon Musk asked Nvidia to prioritize GPU shipments to X over Tesla, emails revealelectrek.cothreelonmusketeers@sh.itjust.worksM to Tesla@lemmy.worldEnglish · 6 months agomessage-square7fedilink
minus-squareEndmaker@lemmy.worldlinkfedilinkEnglisharrow-up4·6 months agoHow are AMD GPUs useful though? Last time I’ve heard, CUDA (and CuDNN) is still an Nvidia-only thing.
minus-squareichbinjasokreativ@lemmy.worldlinkfedilinkEnglisharrow-up7·6 months agoThere are compatibility layers for cuda to run on AMD, and everything AI can also natively run on ROCm. It’s a choice to use nvidia, not mandatory.
minus-squarenotfromhere@lemmy.mllinkfedilinkEnglisharrow-up2·5 months agoWhat is the best working compatibility layer to run cuda on AMD? ROCm seems to drop support pretty quickly after release though so it’s hard for it to get a foothold. As Karparhy has shown, doing low level C++ has some amazing results…
How are AMD GPUs useful though? Last time I’ve heard, CUDA (and CuDNN) is still an Nvidia-only thing.
There are compatibility layers for cuda to run on AMD, and everything AI can also natively run on ROCm. It’s a choice to use nvidia, not mandatory.
Oh wow. TIL
What is the best working compatibility layer to run cuda on AMD? ROCm seems to drop support pretty quickly after release though so it’s hard for it to get a foothold. As Karparhy has shown, doing low level C++ has some amazing results…