jaykrown@lemmy.world to Technology@lemmy.worldEnglish · 23 hours agoDDR6 RAM shrinks by 44% while delivering 4X DDR5's baseline bandwidth — JEDEC previews 12.8 GT/s modules with up to 1 terabyte of capacity per slotwww.tomshardware.comexternal-linkmessage-square112linkfedilinkarrow-up1360arrow-down16
arrow-up1354arrow-down1external-linkDDR6 RAM shrinks by 44% while delivering 4X DDR5's baseline bandwidth — JEDEC previews 12.8 GT/s modules with up to 1 terabyte of capacity per slotwww.tomshardware.comjaykrown@lemmy.world to Technology@lemmy.worldEnglish · 23 hours agomessage-square112linkfedilink
minus-squarebrucethemoose@lemmy.worldlinkfedilinkEnglisharrow-up2·5 hours agoGenerative ML inference isn’t so dramatically bandwidth limited anymore, either, so the “demand” for HBM isn’t as clear.
minus-squareboonhet@sopuli.xyzlinkfedilinkEnglisharrow-up3·3 hours agoEh the really big models want a lot of bandwidth still, otherwise you just don’t make use of all the compute power you have.
Generative ML inference isn’t so dramatically bandwidth limited anymore, either, so the “demand” for HBM isn’t as clear.
Eh the really big models want a lot of bandwidth still, otherwise you just don’t make use of all the compute power you have.