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

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

Rig A

AWS · cloud-model

Change

Rig B

NVIDIA · edge

Change

0 of 31 fields comparable

Spec	AWS Trainium2	NVIDIA Jetson AGX Thor
Bf16 Tflops PER Chip	667	—
Chips PER Trn2 Instance	16	—
Chips PER Ultraserver	64	—
CPU Arch	—	Arm Neoverse-V3AE 64-bit
CPU Cores	—	14
CPU L2 Cache PER Core MB	—	1
CPU L3 Cache MB	—	16
CPU MAX Freq GHZ	—	2.6
Cuda Cores	—	2,560
Devkit Networking	—	1x 5GbE RJ45 + 1x QSFP28 (4x 25GbE)
Devkit Storage	—	1TB NVMe M.2 Key M
FP4 Tflops Sparse	—	2,070
FP8 Dense Petaflops PER Chip	1.3	—
FP8 Sparse Petaflops PER Chip	5.2	—
FP8 Tflops Sparse Unverified	—	1,035
GPU Arch	—	Blackwell, 5th-gen Tensor Cores
GPU MAX Freq GHZ	—	1.57
GPU Tpcs	—	10
HBM Bandwidth TB S PER Chip	2.9	—
Hbm3e Capacity GB PER Chip	96	—
Memory (GB)	—	128
Memory bandwidth (GB/s)	—	273
Memory type	—	LPDDR5X 256-bit
Module	—	Jetson T5000
MSRP (USD)↓ better	—	3,499
Neuroncores PER Chip	8	—
Power Range Watts	—	40-130
Trn2 Instance FP8 Petaflops	20.8	—
Ultraserver FP8 Petaflops	83.2	—
Ultraserver Total HBM Bandwidth TB S	185	—
Ultraserver Total HBM TB	6	—