Search Results for author:
Found 2 papers, 0 papers with code
Swing: Short-cutting Rings for Higher Bandwidth Allreduce
The allreduce collective operation accounts for a significant fraction of the runtime of workloads running on distributed systems.
HammingMesh: A Network Topology for Large-Scale Deep Learning
Numerous microarchitectural optimizations unlocked tremendous processing power for deep neural networks that in turn fueled the AI revolution.
Cannot find the paper you are looking for? You can
Submit a new open access paper.
