Side-by-side, with the numbers that matter.

Two rigs, one table. Higher numbers win in cobalt; TDP and price flip.

Rig A

Intel · accelerator

Change

Rig B

AWS · cloud-model

Change

0 of 32 fields comparable

Spec	Intel Gaudi 3 HL-325L	AWS Trainium2
Bf16 Tflops	1,835	—
Bf16 Tflops PER Chip	—	667
Chips PER Trn2 Instance	—	16
Chips PER Ultraserver	—	64
Ethernet Port Gbps	200	—
Ethernet Ports	24	—
Form factor	OAM	—
FP8 Dense Petaflops PER Chip	—	1.3
FP8 Sparse Petaflops PER Chip	—	5.2
FP8 Tflops	1,835	—
FP8 Tflops Intel WP 2025 Table	1,678	—
HBM (GB)	128	—
HBM Bandwidth TB S PER Chip	—	2.9
HBM Capacity GB PER Chip	—	96
HBM Type	HBM2E	HBM3
Memory bandwidth (TB/s)	3.7	—
MME Engines	8	—
Neuroncores PER Chip	—	8
Process Node NM	5	—
Sram MB	96	—
TDP (W)↓ better	900	—
TPC Engines	64	—
Trainium3 GA	—	2025-12-02
Trainium3 HBM Bandwidth TB S	—	4.9
Trainium3 Hbm3e GB PER Chip	—	144
Trn2 Instance FP8 Petaflops	—	20.8
Trn2 Instance HBM TB	—	1.5
Trn3 Ultraserver FP8 Pflops	—	362
Trn3 Ultraserver MAX Chips	—	144
Ultraserver FP8 Petaflops	—	83.2
Ultraserver Total HBM Bandwidth TB S	—	185
Ultraserver Total HBM TB	—	6