mlsanityData to 14 Oct 2026

Towards Easy and Realistic Network Infrastructure Testing for Large-scale Machine Learning

arXiv 2504.20854 ↗First posted 29 Apr 2025cs.NI cs.AI cs.DCcited by 3

Not on the map. Why?

Abstract (arXiv)

This paper lays the foundation for Genie, a testing framework that captures the impact of real hardware network behavior on ML workload performance, without requiring expensive GPUs. Genie uses CPU-initiated traffic over a hardware testbed to emulate GPU to GPU communication, and adapts the ASTRA-sim simulator to model interaction between the network and the ML workload.

Key facts. Cited by 3 papers in the map (data through 14 Oct 2026). Builds on 12 of its 35 references that are in the map. First posted on arXiv 29 Apr 2025.

Builds on

12 of 35 references.

References · 12

  1. 2021Efficient Large-Scale Language Model Training on GPU Clusters Using Megatron-LMcited by 632
  2. 2024MegaScale: Scaling Large Language Model Training to More Than 10,000 GPUscited by 112
  3. 2022Accelerating Distributed MoE Training and Inference with Linacited by 109
  4. 2023ASTRA-sim2.0: Modeling Hierarchical Networks and Disaggregated Systems for Large-model Training at Scalecited by 47
  5. 2021TACCL: Guiding Collective Algorithm Synthesis using Communication Sketchescited by 38
  6. 2022TopoOpt: Co-optimizing Network Topology and Parallelization Strategy for Distributed Training Jobscited by 30
  7. 2023CASSINI: Network-Aware Job Scheduling in Machine Learning Clusterscited by 26
  8. 2020An In-Depth Analysis of the Slingshot Interconnectcited by 13
  9. 2023Chakra: Advancing Performance Benchmarking and Co-design using Standardized Execution Tracescited by 13
  10. 2022Impact of RoCE Congestion Control Policies on Distributed Training of DNNscited by 6
  11. 2023TACOS: Topology-Aware Collective Algorithm Synthesizer for Distributed Machine Learningcited by 4
  12. 2024Towards a Standardized Representation for Deep Learning Collective Algorithmscited by 3

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 1; 2026: 2.

3 papers.

  1. 2025Scalable Synthesis of distributed LLM workloads through Symbolic Tensor Graphscited by 1
  2. 2026MLCommons Chakra: Advancing Performance Benchmarking and Co-design using Standardized Execution Tracescited by 0
  3. 2026ASTRA-sim 3.0: Next-Level Distributed Machine Learning Simulations via High-Fidelity GPU and Infrastructure Modelingcited by 0