AMD Instinct MI210 64GB vs NVIDIA RTX PRO 6000 Blackwell — Specs & Performance Compared

Side-by-side GPU comparison of the AMD Instinct MI210 64GB and the NVIDIA RTX PRO 6000 Blackwell. Released 2021 (AMD Instinct MI210 64GB), 2025 (NVIDIA RTX PRO 6000 Blackwell). MSRP range: $13,000 · $8,499. Specs, synthetic benchmarks, AI inference tok/s, and Amazon availability below; each column header links to that part's benchmark page.

AMD · CDNA2

AMD Instinct MI210 64GB

64 GB VRAM300W TDP$13,000 MSRP2021

Full AMD Instinct MI210 64GB benchmarks →

NVIDIA · Blackwell Pro

NVIDIA RTX PRO 6000 Blackwell

96 GB VRAM300W TDP$8,499 MSRP2025

Full NVIDIA RTX PRO 6000 Blackwell benchmarks →

*Price sourced from Amazon.com. Price and availability subject to change.

Where to buy

A listing for each part; the current price and stock are on the retailer's page.

Specs side-by-side

Spec	AMD Instinct MI210 64GB	NVIDIA RTX PRO 6000 Blackwell
vram gb	64	96
tdp w	300	300
cuda cores	—	20480

Synthetic benchmarks

Each row says whether a higher or a lower number is better (ranks and times are lower-is-better). Scores come from 3DMark, TechSpot, StorageReview, Chips and Cheese, AMD Official Specs and PassMark; each row links its sources, and where they differ the cell shows their median. Empty cells mean we don't yet have that benchmark on file for the SKU.

Benchmark	AMD Instinct MI210 64GB	NVIDIA RTX PRO 6000 Blackwell
3DMark Steel Nomad (higher is better) Source: 3DMark	—	15,939 points
3DMark Time Spy (higher is better) Source: TechSpot	—	51,776 points
3DMark Time Spy Extreme (higher is better) Source: TechSpot	—	28,009 points
Blender Benchmark - Classroom (higher is better) Source: StorageReview	—	4,041 samples/min
Blender Benchmark - Junkshop (higher is better) Source: StorageReview	—	4,159 samples/min
Blender Benchmark - Monster (higher is better) Source: StorageReview	—	7,870 samples/min
HBM2e L2 Cache Bandwidth (micro-benchmark) (higher is better) Source: Chips and Cheese	3,700 GB/s	—
HBM2e Peak Memory Bandwidth (theoretical) (higher is better) Source: AMD Official Specs	1,638 GB/s	—
PassMark G3D Mark (higher is better) Source: PassMark	—	37,970 points
Procyon AI Text Generation - Llama2 (higher is better) Source: StorageReview	—	8,008 points
Procyon AI Text Generation - Llama3 (higher is better) Source: StorageReview	—	6,501 points
Procyon AI Text Generation - Mistral (higher is better) Source: StorageReview	—	7,346 points

AI inference (tokens/sec)

Top community-reported tokens-per-second results for popular LLMs whose weights fit the card's memory, each linked to its source. See the per-SKU benchmark page for the full set including quantization and prompt length.

AMD Instinct MI210 64GB

llama2:7b: 718.1 tok/s (EmbeddedLLM)
llama2:7b: 718.1 tok/s (EmbeddedLLM)
llama2:13b: 595.1 tok/s (EmbeddedLLM)

All AMD Instinct MI210 64GB AI benchmarks →

NVIDIA RTX PRO 6000 Blackwell

qwen3:14b: 1743.8 tok/s (DatabaseMart)
deepseek-r1:32b: 829.0 tok/s (DatabaseMart)
Llama 3.1 8B FP8: 380.0 tok/s (Runpod)

All NVIDIA RTX PRO 6000 Blackwell AI benchmarks →

Quick answers

What did the AMD Instinct MI210 64GB and the NVIDIA RTX PRO 6000 Blackwell cost at launch?

The AMD Instinct MI210 64GB launched at a $13,000 MSRP and the NVIDIA RTX PRO 6000 Blackwell at $8,499. Street prices drift from launch pricing; the Where to buy section links current listings.