Skip to content
Preprint

Provisioning Networks for AI Supercomputers: A Trace-Driven Study of Performance Sensitivity at Unprecedented Scale

Oct 2026 · 0 citations · 49 references
Computer Science

TL;DR

This work presents a trace-driven study of network sensitivity for training and inference workloads at scales up to 4,096 GPUs, and combines signature-based trace decomposition with linear programming to make broad network-parameter sweeps feasible.

Abstract

AI workloads on supercomputers make interconnect performance a primary design constraint. Hardware specifications and microbenchmarks alone do not show how real workloads respond to latency, bandwidth, and jitter. We present a trace-driven study of network sensitivity for training and inference workloads at scales up to 4,096 GPUs. Our methodology combines signature-based trace decomposition with linear programming to make broad network-parameter sweeps feasible. It preserves the recorded NCCL schedules and program dependencies, and estimates a lower bound on runtime under the LogGP model. We validate the compositional analysis against monolithic linear programs, LogGOPSim, and hardware measurements at feasible scales. The resulting latency-bandwidth envelopes identify low-sensitivity regions and performance cliffs that vary substantially across workloads. They also show when bandwidth upgrades provide meaningful returns and when additional network performance has little effect. These results translate execution traces into practical guidance for evaluating network configurations while making the model assumptions explicit.

View source

Similar papers

#artificial intelligence Preprint Oct 2026

LLM-PowerTrace: A Dataset of Time-Resolved GPU Power Traces for AI Compute

AI compute is becoming large and fast-changing electrical loads. Its impacts on data center infrastructure and power grids depend not only on total energy consumption, but also on how their power draw moves over sub-second to minute timescales, including peaks, ramps, and synchronized fluctuations across accelerators....

Jian-Sheng Wang, J. Singh, Hao Xu et al. · 0 citations
Open access Sep 2026

A Characterization of Google Workload Traces

In this work, we perform an in-depth characterization of the microarchitectural performance of modern datacenter workloads. Google recently released hundreds of gigabytes of trace data from production, offering a unique view into current-day warehouse-scale workload behavior. Through simulating 359 cores spanning 12 wo...

Matthew Giordano, Akanksha Jain, E. Deiana et al. · 1 citation
Book Open access Sep 2026

LMTracer: Fine-Grained and Real-Time Performance Profiling for Production LLM Systems

LMTracer is presented, a fine-grained and real-time performance profiling framework for production LLM services that embed profiling logic into the execution through graph-embedded probing and streaming buffered profiling data to CPUs on demand to keep the execution of user kernels uninterrupted.

Wei Liu, Yong-Chao He, Bo-Han Zhao et al. · 0 citations
Book Open access Nov 2026

ML Inference Acceleration via Scalable In-network Computing

Modern inference engines have significantly advanced the scalability of large models through innovations in serving techniques, proving instrumental in driving the widespread adoption of AI. However their true potential is constrained by traditional server architecture where reliance on host CPUs and operating systems...

Shaunak Galvankar · 0 citations
Sep 2026

NDS: Programmer-Free Offload of High-Performance Near Data Strands

Near Data Processing (NDP) has the potential to significantly improve system performance and energy by alleviating data movement bottlenecks. However, most NDP proposals pose heavy requirements for the software stack, data layout, and/or the underlying hardware. To broaden NDP adoption, this work focuses on a modular h...

Shreyas Singh, Pratyush Nandi, Lin Jia et al. · 0 citations
Book Open access Sep 2026

StreamTrace: Fast Trace Analysis for Large-Scale Parallel Applications on a Single Node

StreamTrace is presented, a streaming trace analysis system that enables efficient analysis of massive traces on a single computing node with bounded memory consumption and introduces two key techniques: communication pattern-aware chunk partitioning that minimizes cross-chunk dependence and dynamic priority-based chun...

Yu-Yang Jin, Ji-Dong Zhai · 0 citations

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.