Hugging Face's logo

Datasets:
alpha-one-index
/
ai-infra-index

Tasks:
Tabular Classification
Tabular Regression
Languages:
English
Size:
n<1K
Tags:
gpu
cloud-computing
pricing
hardware
ai-infrastructure
benchmarks
Libraries:
License:
Dataset card Files
xet
 Community
Dataset Viewer
Duplicate
The dataset viewer is not available for this split.
Cannot load the dataset split (in streaming mode) to extract the first rows.
Error code:   StreamingRowsError
Exception:    ValueError
Message:      Bad split: cloud_pricing. Available splits: ['train']
Traceback:    Traceback (most recent call last):
                File "/src/services/worker/src/worker/utils.py", line 99, in get_rows_or_raise
                  return get_rows(
                         ^^^^^^^^^
                File "/src/libs/libcommon/src/libcommon/utils.py", line 272, in decorator
                  return func(*args, **kwargs)
                         ^^^^^^^^^^^^^^^^^^^^^
                File "/src/services/worker/src/worker/utils.py", line 61, in get_rows
                  ds = load_dataset(
                       ^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/load.py", line 1705, in load_dataset
                  return builder_instance.as_streaming_dataset(split=split)
                         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 1196, in as_streaming_dataset
                  raise ValueError(f"Bad split: {split}. Available splits: {list(splits_generators)}")
              ValueError: Bad split: cloud_pricing. Available splits: ['train']

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

AI Infrastructure Index

Dataset Description

The AI Infrastructure Index is a comprehensive open-source reference for AI hardware specifications, cloud GPU pricing, and infrastructure intelligence. It catalogs major AI hardware platforms currently in production, covering data center GPUs, custom AI accelerators (TPUs, LPUs, IPUs, WSEs), cloud pricing, benchmarks, and cost optimization data.

Dataset Summary

This dataset provides structured, machine-readable data on:

  • Cloud GPU Pricing — Real-time pricing from 12 cloud providers (Azure, RunPod, Lambda Labs, CoreWeave, Together AI, Vast.ai, etc.)
  • GPU Specifications — Detailed specs for NVIDIA (H100, H200, B200, GB200), AMD (MI300X, MI325X), and Intel (Gaudi 3) data center accelerators.
  • Training Costs — Model training cost estimates and TCO analysis.
  • Inference Benchmarks — MLPerf results and throughput data.
  • Model-GPU Sizing — Guides for matching GPU configurations to model requirements.

Languages

The documentation and data field names are in English (en).

Data Fields

Cloud Pricing (data/cloud-pricing.json)

  • provider: Cloud service provider name.
  • gpu_name: Hardware model (e.g., "H100 NVL").
  • price_per_hour_usd: Hourly rental cost in USD.
  • pricing_type: On-demand, Reserved, or Spot.
  • region: Geographic region or "Global".
  • timestamp: Last update time (ISO 8601).

GPU Specifications (data/gpu-specs.json)

  • model: GPU model name.
  • architecture: Hardware architecture (e.g., "Hopper", "Blackwell").
  • memory_capacity: HBM capacity in GB.
  • memory_bandwidth: Bandwidth in TB/s.
  • fp8_perf: Peak FP8 performance (PFLOPS).
  • tdp: Thermal Design Power in Watts.

Data Splits

The dataset is organized by file type rather than traditional ML splits:

  • cloud_pricing: Contains current market rates.
  • gpu_specs: Contains static hardware specifications.

Dataset Creation

Curation Rationale

No single open-source resource previously combined GPU hardware specifications, multi-provider cloud pricing, and model-sizing benchmarks into a unified, machine-readable format for automated procurement and system design.

Source Data

All data is independently verified against official vendor datasheets and pricing pages. Primary sources include:

  • Official NVIDIA, AMD, and Intel datasheets.
  • Cloud provider pricing APIs (e.g., Azure Retail Prices API) and public pricing tables.
  • MLCommons (MLPerf) benchmark results.

Personal and Sensitive Information

This dataset contains no personal or sensitive information. All data consists of publicly available technical specifications and commercial pricing.

Licensing Information

The dataset is licensed under the MIT License.

Citation Information 

@misc{aiinfraindex2026,
  title = {AI Infrastructure Index},
  author = {Alpha One Index},
  year = {2026},
  url = {https://github.com/alpha-one-index/ai-infra-index},
  note = {Open-source AI hardware specifications and cloud pricing data}
}

Contributions

Contributions are welcome via Pull Requests. Please see CONTRIBUTING.md for data format standards.
Downloads last month
38
Total file size:
30.7 kB