F

Farmgpu

PRODUCTS.FARMGPU.COMManage supplier listing

FarmGPU (Haystack fleet telemetry platform) — Workload-classified GPU fleet telemetry: a catalog of 50,000+ distinct GPU workloads over a 90-day window, each a ~100-field profile joining hardware telemetry (utilisation, memory, power draw, temperature, NVLink, ECC error counts) with container-level metrics and an LLM-assigned semantic workload label derived from image name, metric shape, and GitHub/Docker Hub repo lookup. Uniquely, it pairs measured utilisation with what is actually running and why — a demand-composition and efficiency asset rather than another price feed — plus a quantified …

Datasets

1

Workload-classified GPU fleet telemetry: a catalog of 50,000+ distinct GPU workloads over a 90-day window, each a ~100-field profile joining hardware telemetry (utilisation, memory, power draw, temperature, NVLink, ECC error counts) with container-level metrics and an LLM-assigned semantic workload label derived from image name, metric shape, and GitHub/Docker Hub repo lookup. Uniquely, it pairs measured utilisation with what is actually running and why — a demand-composition and efficiency asset rather than another price feed — plus a quantified waste layer (GPU-hours burned below 10 percent utilisation, worth $500K+ annually on the reference fleet) and an observed workload-mix distribution across development, image generation, inference, video generation, fine-tuning and training.

Scale 50Coverage Information Technology · FinancialsAsset class Equities · Fixed income

Buyer room

Sample, licence terms, pricing and eval results when Farmgpu publishes them. Until then, discover alternatives today with a 7-day trial.

Start 7-day trial →

What buyers ask about Farmgpu, answered from this page.

What does Farmgpu do?

FarmGPU (Haystack fleet telemetry platform) — Workload-classified GPU fleet telemetry: a catalog of 50,000+ distinct GPU workloads over a 90-day window, each a ~100-field profile joining hardware telemetry (utilisation, memory, power draw, temperature, NVLink, ECC error counts) with container-level metrics and an LLM-assigned semantic workload label derived from image name, metric shape, and GitHub/Docker Hub repo lookup. Uniquely, it pairs measured utilisation with what is actually running and why — a demand-composition and efficiency asset rather than another price feed — plus a quantified …

What data does Farmgpu offer?

Farmgpu offers (Alternative, Supply Chain, Geospatial) — 50,000+ cataloged workloads × ~100 fields per profile ≈ 5M+ field-level observations over 90 days, across 5,000+ unique container images; underlying DCGM + cAdvisor + Node Exporter time series at Prometheus scrape resolution (typically 15-60s) per GPU per node, so the raw metric volume is orders of magnitude larger than the catalogued profile layer; plus 11 dashboard views and CSV/JSON export.

What is Farmgpu's data used for?

Ground-truth AI demand composition — the observed workload mix (development 36%, image generation 18%, inference 9%, video generation 6%, fine-tuning 2.5%, training 0.4%) is a direct measurement of where accelerator capacity actually goes, a leading indicator for whether AI spend is shifting from experimentation to production inference; fleet efficiency benchmarking and an utilisation-based efficiency score for data-centre operators; inference-vs-training ratio as an AI-monetisation signal tradable against AI-infrastructure equities; electricity and cooling intensity modelling via power and temperature telemetry per workload type; ECC-error and thermal-failure forensics as a hardware-reliability dataset for accelerator vendors and insurers; GPU waste/spend-reduction auditing; and residual-value and utilisation-underwriting inputs for GPU-collateralised lending, where measured utilisation is the missing variable in terminal-value assumptions.

What does Farmgpu cover?

Coverage spans US; IT Consulting & Other Services, Financial Exchanges & Data; alternative, supply_chain, geospatial; equities, fixed_income.

How do I get a sample from Farmgpu?

The supplier's own access route is a contact form. Request a sample through Brickroad to start.

Farmgpu has not added their own details yet. Not yet on file:

  • Summary & mission
  • People
  • Use cases & buyers
  • FAQ