Show HN: Hardware-software co-design for MoE models to bypass NCCL bottlenecks github.com 3 points by PJHkorea 1 day ago
PJHkorea 1 day ago Imagine virtual memory, but instead of using your local storage, you’re pooling a whole bunch of network servers into a single address space
Imagine virtual memory, but instead of using your local storage, you’re pooling a whole bunch of network servers into a single address space