Pith. sign in

REVIEW 5 cited by

Empirical Measurements of AI Training Power Demand on a GPU-Accelerated Node

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2412.08602 v2 pith:BLZZD66L submitted 2024-12-11 cs.AR

Empirical Measurements of AI Training Power Demand on a GPU-Accelerated Node

classification cs.AR
keywords energypowertrainingconsumptiondemanddrawduringinfrastructure
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

The expansion of artificial intelligence (AI) applications has driven substantial investment in computational infrastructure, especially by cloud computing providers. Quantifying the energy footprint of this infrastructure requires models parameterized by the power demand of AI hardware during training. We empirically measured the instantaneous power draw of an 8-GPU NVIDIA H100 HGX node during the training of open-source image classifier (ResNet) and large-language models (Llama2-13b). The maximum observed power draw was approximately 8.4 kW, 18% lower than the manufacturer-rated 10.2 kW, even with GPUs near full utilization. Holding model architecture constant, increasing batch size from 512 to 4096 images for ResNet reduced total training energy consumption by a factor of 4. These findings can inform capacity planning for data center operators and energy use estimates by researchers. Future work will investigate the impact of cooling technology and carbon-aware scheduling on AI workload energy consumption.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 5 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Bit2Watt: A Cyber-Physical Vulnerability Exploiting GPU Workloads Across Power and Computing Infrastructures

    cs.CR 2026-07 conditional novelty 7.0

    Coordinated GPU workload manipulation by unprivileged cloud tenants can induce high-frequency power modulations that destabilize inverter-dominated grids, causing harmonic distortion, negative damping, and potential c...

  2. Grid Integration of Gigawatt-Scale AI Data Centers under Connect-and-Manage

    eess.SY 2026-05 conditional novelty 6.0

    A hierarchical request-acceptance protocol with learning-based planning and robust TSO evaluation reduces curtailment for GW-scale AI data centers from 9.1% to 2.8% while preserving 98.1% of frontier training workload.

  3. Composite Control of Grid-Following Inverters for Stabilizing AI-Induced Fast Power Disturbances

    eess.SY 2026-04 unverdicted novelty 5.0

    Singular perturbation analysis derives physically implementable droop control for inverters from reduced-system stability requirements to reject bounded-rate AI-induced power disturbances, providing explicit gain boun...

  4. Measurement of Generative AI Workload Power Profiles for Whole-Facility Data Center Infrastructure Planning

    eess.SY 2026-04 unverdicted novelty 5.0

    High-resolution power profiles for AI workloads on H100 GPUs are measured and scaled to whole-facility energy demand using a bottom-up model, with the dataset made public.

  5. Wide-Area Power System Oscillations from Large-Scale AI Workloads

    eess.SY 2025-08 unverdicted novelty 5.0

    AI datacenter workloads produce sustained power fluctuations that act as forcing inputs capable of amplifying local and inter-area oscillation modes in simulated grids.