ScaleFlux CEO:液冷之外,AI 内存与存储才是瓶颈
ScaleFlux 联合创始人兼 CEO Hao Zhong 在 AI Data Pipeline Forum 上表示,GPU 和 TPU 性能过去八年提升近 1000 倍,但外部 DRAM 与 SSD 组成的数据管道未能跟上,内存和存储正成为未来 AI 的瓶颈。ScaleFlux 今年推出两款控制器芯片,分别面向 PCIe Gen 6 SSD 和 Compute Express Link 内存,用于第二、三级数据管道。随着 AI 从训练转向推理和智能体负载,系统需要更多数据就近可用。
enser racks and liquid cooling to keep that hardware running. The data pipeline that feeds those processors has not kept up, and expensive compute ends up waiting on memory and storage. That gap is getting more attention as AI infrastructure bottlenecks become central to data center planning. The fix is to rethink the external memory and storage tiers behind the GPU, according to Hao Zhong (pictured), co-founder and chief executive officer of ScaleFlux Inc. “GPU and TPU performance has been adva
匹配字段正文