Analyzing Persistent Alltoallv RMA Implementations for High-Performance MPI Communication

2026-04-08T08:52:25Z54826ba817781be4b71d1c83fc7bacef3e5b146fddbd4dc8322a6d6953eb3a1a
AlltoallvCUDAGPUGTaPGoHPCJAXKubernetesMPINode.jsOpenFaaSPythonRMAdistributed-systems-theoryfence-synchronizationfork-join-runtime','pragma-interface','work-stealing','EPAQ','Cofriends-of-friendsk-nearest-neighborslinearizabilitylock-synchronizationmessage-passingperformance-optimizationpersistent-communicationserverlesstree-traversal

What happened

Feed of recent arXiv CS (distributed computing / systems / ML infrastructure) papers (2026-04-08) covering practical and theoretical advances in HPC and ML systems. Key contributions include: a persistent RMA Alltoallv design (fence/lock variants) that reduces large-message MPI runtimes up to ~44% and improves scalability; an OpenFaaS-on-Kubernetes performance study favoring Go and K3s/Kubeadm; theoretical lower bounds and message-chain requirements for linearizable registers in asynchronous message-passing systems; JZ-Tree, a Morton plane-based GPU-friendly tree layout providing >10x speedups

Why it matters

A reviewed impact interpretation has not been published for this record.

Evidence and limitations

Source ID
arxiv_cs_dc
Record identifier
54826ba817781be4b71d1c83fc7bacef3e5b146fddbd4dc8322a6d6953eb3a1a
Enrichment time
2026-04-08T08:52:25Z
AI-assisted enrichment
Yes

This record may overlap with other records. Its enrichment can be incomplete or wrong, and machine assistance was used. Validate consequential decisions against the linked source and your own environment.

Record · Analyzing Persistent Alltoallv RMA Implementations for High-Performance MPI Communication · Baitaphish