4 hours agoMemory executives expect RAM shortage to continue through 2028Ars Technica · T1Infra & Chips#memory-shortage#hbm#dram◆AI generated82◀
4 days agoNvidia opens NVLink Fusion to third-party accelerators in bid to lock in the AI data centerArs Technica · T1Infra & Chips#gpu#interconnect#cuda◆AI generated60◀
4 days agoNVLink Fusion is Nvidia's play to become the connective tissue of every AI clusterTechCrunch · T2Infra & Chips#gpu#interconnect#cuda◆AI generated52◀
4 days agoWhy interconnects, not GPUs, may decide who wins the next decade of AI hardwareMIT Technology Review · T2Infra & Chips#gpu#interconnect#cuda◆AI generated52◀
6 days agoNew export rules quietly redraw the map for AI accelerator salesArs Technica · T1Infra & Chips#chips#supply-chain◆AI generated52◀
6 days agoChipmakers say the fine print is worse than the headlineTechCrunch · T2Infra & Chips#chips#supply-chain◆AI generated44◀
3 hours agoLiquid cooling retrofits hit the 100kW-per-rack wallArs Technica · T1Infra & Chips#liquid-cooling#data-center#rack-density◆AI generated