Accelerators and local compute profiles used for AI training, fine-tuning, and inference.
Curated coverage priority across current AI accelerator and local-device alternatives from official vendor documentation; this is not a market-share estimate.
Ranked alternatives
Popularity-ranked entries linked back to graph nodes and deterministic sources.
| Rank | Component | Kind | Openness | Popularity metric | Tasks | Source |
|---|---|---|---|---|---|---|
| 1 | NVIDIA B200infrastructure:nvidia-b200 | infrastructure | proprietary | 100Coverage priority as of 2026-06-10 | training, inference, fine-tuning | source |
| 2 | NVIDIA H200infrastructure:nvidia-h200 | infrastructure | proprietary | 99Coverage priority as of 2026-06-10 | training, inference, fine-tuning | source |
| 3 | NVIDIA H100infrastructure:nvidia-h100 | infrastructure | proprietary | 98Coverage priority as of 2026-06-10 | training, inference, fine-tuning | source |
| 4 | NVIDIA GeForce RTX 5090infrastructure:nvidia-rtx-5090 | infrastructure | proprietary | 97Coverage priority as of 2026-06-10 | local-inference, image-generation, fine-tuning | source |
| 5 | AMD Instinct MI325Xinfrastructure:amd-mi325x | infrastructure | proprietary | 96Coverage priority as of 2026-06-10 | training, inference, fine-tuning | source |
| 6 | AMD Instinct MI300Xinfrastructure:amd-mi300x | infrastructure | proprietary | 95Coverage priority as of 2026-06-10 | training, inference, fine-tuning | source |
| 7 | Intel Gaudi 3infrastructure:intel-gaudi-3 | infrastructure | proprietary | 94Coverage priority as of 2026-06-10 | training, inference | source |
| 8 | Google Cloud TPU v6e Trilliuminfrastructure:google-cloud-tpu-v6e-trillium | infrastructure | proprietary | 93Coverage priority as of 2026-06-10 | training, inference, fine-tuning | source |
| 9 | AWS Trainium3infrastructure:aws-trainium3 | infrastructure | proprietary | 92Coverage priority as of 2026-06-10 | training, inference, fine-tuning | source |
| 10 | Apple M5 Maxinfrastructure:apple-m5-max | infrastructure | proprietary | 91Coverage priority as of 2026-06-10 | local-inference, developer-workstation | source |