DEPLOYDatabase

Chip comparison

AWS Inferentia 2 vs AWS Trainium (Trn1)

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.

FieldAWS Inferentia 2AWS Trainium (Trn1)
Process nodeTSMC N7TSMC N7
TDP250 W500 W
Memory32 GB HBM332 GB HBM2e
Memory bandwidth820 GB/s820 GB/s
BF16 (dense)190 TFLOPS210 TFLOPS
FP8 (dense)380 TFLOPS
Form factorInf2 EC2 instanceTrn1 EC2 instance
Announced2022-12-012020-12-01
Released2023-04-012022-10-01

Robots running AWS Inferentia 2

No robots publicly running it yet.

Robots running AWS Trainium (Trn1)

No robots publicly running it yet.

Data centers with AWS Inferentia 2

No data centers publicly running it yet.

Data centers with AWS Trainium (Trn1)

No data centers publicly running it yet.

Common questions

AWS Inferentia 2 vs AWS Trainium (Trn1): which is faster for training?

AWS Inferentia 2 has 1.81x the dense FP16/BF16 throughput of the other (AWS Inferentia 2: 380 TFLOPS; AWS Trainium (Trn1): 210 TFLOPS). Real-world training speed also depends on memory bandwidth, interconnect, and software stack: see the spec diff for those inputs.

AWS Inferentia 2 vs AWS Trainium (Trn1): which has more memory?

AWS Trainium (Trn1) carries more HBM (AWS Inferentia 2: 32 GB HBM3; AWS Trainium (Trn1): 32 GB HBM2e). Memory capacity sets the largest model that fits on one chip without partitioning; memory bandwidth sets inference throughput.

AWS Inferentia 2 vs AWS Trainium (Trn1): which is more power-efficient?

Dense FP16/BF16 TFLOPS per watt: AWS Inferentia 2 1.52 (380 ÷ 250 W); AWS Trainium (Trn1) 0.42 (210 ÷ 500 W). AWS Inferentia 2 wins at the die level; system-level efficiency also depends on cooling and interconnect.

See also: every chip comparison · AWS Inferentia 2 full page · AWS Trainium (Trn1) full page.