Top GPU Cloud Serverless Infrastructure (Opensource) Star if you like it!
-
Updated
Nov 4, 2026
Top GPU Cloud Serverless Infrastructure (Opensource) Star if you like it!
Andrew White — Managing Technical Consultant on the AI Factory Team at CDW | NVIDIA SME
A deterministic C++20 digital twin for AI training clusters—model workloads, networks, scheduling, congestion, and failures through reproducible experiments.
Code for "When Carbon-Aware Deferral Backfires: Contention, Not Forecast Error, Drives Harm in GPU Workload Shifting" (PCDS 2026). Simulator, scheduling policies, conformal forecasters and harm attribution, replaying the Alibaba PAI GPU trace against six electricity grids.
LLM-agent planning for orbital GPU clusters, with measured A100 traces, an optical-interconnect simulator, and joint placement/scheduling tools.
[WIP] Interactive visualization of LLM training parallelism across GPU clusters
To associate your repository with the gpu-clusters topic, visit your repo's landing page and select "manage topics."