Chip comparison
AWS Trainium 2 vs Intel Gaudi 3
Side-by-side specs
Straight from each vendor's datasheet. Winner in bold: higher throughput, memory, or transistor count; lower power draw or price. Sparse (2:4) throughput roughly doubles the FP8/INT8 numbers shown.
| Field | AWS Trainium 2 | Intel Gaudi 3 |
|---|---|---|
| Process node | TSMC N5 | TSMC N5 |
| TDP | 500 W | 900 W |
| Memory | 96 GB HBM3 | 128 GB HBM2e |
| Memory bandwidth | 2,900 GB/s | 3,675 GB/s |
| FP16 (dense) | — | 1,835 TFLOPS |
| BF16 (dense) | 667 TFLOPS | 1,835 TFLOPS |
| FP8 (dense) | 1,300 TFLOPS | 1,835 TFLOPS |
| INT8 (dense) | 1,300 TOPS | — |
| Form factor | AWS Neuron 2 (Trn2 UltraServer with 64 chips per unit) | OAM |
| Announced | 2024-12-03 | 2024-04-09 |
| Released | 2025-01-01 | 2024-09-01 |
Robots running AWS Trainium 2
No robots publicly running it yet.
Robots running Intel Gaudi 3
No robots publicly running it yet.
Data centers with AWS Trainium 2
Data centers with Intel Gaudi 3
No data centers publicly running it yet.
Common questions
AWS Trainium 2 vs Intel Gaudi 3: which is faster for training?
Intel Gaudi 3 has 1.41x the dense FP8 throughput of the other (AWS Trainium 2: 1,300 TFLOPS; Intel Gaudi 3: 1,835 TFLOPS). Real-world training speed also depends on memory bandwidth, interconnect, and software stack: see the spec diff for those inputs.
AWS Trainium 2 vs Intel Gaudi 3: which has more memory?
Intel Gaudi 3 carries more HBM (AWS Trainium 2: 96 GB HBM3; Intel Gaudi 3: 128 GB HBM2e). Memory capacity sets the largest model that fits on one chip without partitioning; memory bandwidth sets inference throughput.
AWS Trainium 2 vs Intel Gaudi 3: which is more power-efficient?
Dense FP8 TFLOPS per watt: AWS Trainium 2 2.60 (1,300 ÷ 500 W); Intel Gaudi 3 2.04 (1,835 ÷ 900 W). AWS Trainium 2 wins at the die level; system-level efficiency also depends on cooling and interconnect.
AWS Trainium 2 vs Intel Gaudi 3: which is more widely used?
AWS Trainium 2: 0 robots and 3 data centers publicly running it. Intel Gaudi 3: 0 robots and 0 data centers. Empty here means no operator has said publicly, not zero adoption in the market.
See also: every chip comparison · AWS Trainium 2 full page · Intel Gaudi 3 full page.