525 citations · 680 across the 8 of their papers we have counts for
Showing cs.NIShow all
3 papers · 1 filter
cs.NI2026
HCCL: Collective Communication for Meta Training and Inference Accelerators
Wesley Bland, Tiago Antunes, Lars Paul Huse +63
We present HCCL, a collective communication library co-designed with Meta's MTIA 300 accelerator, the first Meta chip to integrate backend networking directly on chip package. MTIA…
cs.NI2026★ 2 cited
Resilient AI Supercomputer Networking using MRC and SRv6
Joao Araujo, Alex Chow, Mark Handley +47
Tail latency dominates the performance of synchronous pretraining jobs when running at very large scales. We describe a three-pronged approach: (1) a new RDMA-based transport proto…
cs.NI2012★ 1 cited
Cross-Layer Design to Maintain Earthquake Sensor Network Connectivity After Loss of Infrastructure
Greg Steinbrecher
We present the design of a cross-layer protocol to maintain connectivity in an earthquake monitoring and early warning sensor network in the absence of communications infrastructur…