GPU Comparison

AMD Instinct MI350X vs NVIDIA RTX PRO 6000 Blackwell Server Edition

How the AMD Instinct MI350X and NVIDIA RTX PRO 6000 Blackwell Server Edition compare across memory, compute, power and form factor, with the stronger figure highlighted on each row. Vendor-specific figures (CUDA vs stream processors, Tensor vs matrix cores) are listed on their own rows.

Specification	CDNA 4 AMD Instinct MI350X NVIDIA alternative: NVIDIA B200 →	BLACKWELL NVIDIA RTX PRO 6000 Blackwell Server Edition View Product →
Pricing
Price	Not sold by NVNexus	$17,445.60
Overview
Architecture	AMD CDNA 4	Blackwell
Memory
Memory	288 GB HBM3E	96 GB GDDR7 with ECC
Memory Bandwidth	8 TB/s	Up to 1.6 TB/s
Memory Interface	8192-bit	–
Compute
CUDA Cores	–	24,064
Stream Processors (AMD)	16,384 (256 CUs)	–
Tensor Cores	–	752
Matrix Cores / AI Accelerators (AMD)	1,024	–
RT Cores / Ray Accelerators	–	188
FP32	144.2 TFLOPS	–
FP64 Tensor	72.1 TFLOPS	–
FP16/BF16 Tensor	2.3 PFLOPS / 4.6 (sparse) TFLOPS	–
FP8 Tensor	4.6 PFLOPS / 9.2 (sparse) TFLOPS	–
FP4 Tensor	9.2 PFLOPS TFLOPS	–
Platform
Interface	PCIe Gen5 x16	PCIe Gen5 x16
NVLink / Infinity Fabric	Infinity Fabric, 1,075 GB/s aggregate	–
TDP	1000W	400W – 600W (configurable)
Form Factor	OAM module (8-GPU platform)	4.4" H × 10.5" L, FHFL dual slot, passive
Cooling	Air cooled	–
Transistors	185 billion	–

Not sure which fits?

Talk it through with a specialist

Tell us your workload and we will recommend the right GPU, configuration, and sourcing options.

Talk to an Expert →

Talk it through with a specialist

More Comparisons