AMD Radeon Pro VII
Live eBay prices, cheapest first. The best one right now is $355.00 (used).
Specs
- VRAM
- 16 GB HBM2
- Bandwidth
- 1,024 GB/s
- FP32
- 13.1 TFLOPS
- FP16
- 26.3 TFLOPS
- Power
- 250 W
- Connector
- PCIe 4.0
- Released
- 2020
- Launch price
- $1,899
- Llama 3.1 8B at Q4 (est.)
- 74–127 tokens/s
gfx906: dropped from current ROCm; runs llama.cpp (ROCm/Vulkan) and older ROCm builds.
Price right now
Live on eBayFrom the live listings below; shipping included.
Listings
Live on eBay| Listing | Condition | Price | Checked |
|---|---|---|---|
| Used | $355.00or best offer−41%free shipping | ||
| Used | $370.00−38%shipping not listed | ||
| Used | $450.00or best offer−25%free shipping | ||
| Open box | $455.00or best offer−24%+$25.00 shipping | ||
| Used | $472.00−21%+$22.00 shipping | ||
| Used | $474.00or best offer−21%shipping not listed | ||
| New | $515.55−14%+$30.00 shipping | ||
| Used | $529.33−12%+$30.00 shipping | ||
| Used | $550.00or best offer−8%shipping not listed | ||
| New | $578.00or best offerfree shipping | ||
| New | $591.78+$8.00 shipping | ||
| New | $610.00or best offer+$5.00 shipping | ||
| New | $628.00or best offerfree shipping | ||
| New | $629.85or best offerfree shipping | ||
| New | $648.38or best offer+$8.00 shipping | ||
| Used | $650.98or best offerfree shipping | ||
| New | $675.99or best offer+$14.99 shipping | ||
| New | $710.00or best offerfree shipping | ||
| New | $727.20+$10.00 shipping | ||
| New | $780.00+$30.00 shipping | ||
| New | $780.00+$30.00 shipping | ||
| New | $917.43or best offer+$38.88 shipping |
Auctions ending soon
No auctions ending soon right now.
Prices include shipping and were read from eBay at the time shown. They can change: check the listing before you buy.
Owner reviews
No owner reviews yet. Own one? Tell others how it runs.
From signed-in owners, checked before they're published. PerGB doesn't pay for or edit reviews.
How it compares
Live on eBay- #6 of 83 PCIe GPUs with 16 GB or more for the money: $557.66 per GeForce RTX 3090 of performance.
- Software: Not on ROCm's support list (older ROCm or Vulkan).
As much VRAM, about as fast, for less
- AMD Radeon VII: $249.00, 65% of the GeForce RTX 3090, 16 GB
- AMD Instinct MI50 16 GB: $266.00, 64% of the GeForce RTX 3090, 16 GB
- NVIDIA Tesla V100 PCIe 16 GB: $325.00, 123% of the GeForce RTX 3090, 16 GB
Faster, for the least
- Intel Arc A770 8 GB: $259.99, 103% of the GeForce RTX 3090, 8 GB
- NVIDIA Tesla V100 PCIe 12 GB: $261.00, 123% of the GeForce RTX 3090, 12 GB
- NVIDIA Titan V: $394.04, 104% of the GeForce RTX 3090, 12 GB
More VRAM, for not much more
- NVIDIA Tesla P40: $270.00, 35% of the GeForce RTX 3090, 24 GB
From the live listings of every GPU, shipping included; performance against the GeForce RTX 3090 (FP16 or FP32 compute and memory bandwidth). Updated with every build.
Local AI: what it runs
| Model (Q4, 8K context) | Radeon Pro VII needed | Costs now | Speed (est.) |
|---|---|---|---|
| Llama 3.1 8B | 1 card | $355.00 | 74–127 tokens/s |
| Qwen3 14B | 1 card | $355.00 | 40–68 tokens/s |
| gpt-oss 20B | 1 card | $355.00 | 99–169 tokens/s |
| Gemma 3 27B | 2 cards | $725.00 | 22–37 tokens/s |
| Qwen3 32B | 2 cards | $725.00 | 18–31 tokens/s |
| Llama 3.3 70B | 4 cards | $1,630 | 8.4–14 tokens/s |
| gpt-oss 120B | 5 cards | $2,102 | 70–119 tokens/s |
| Qwen3 235B-A22B | more than 8 | — | — |
| DeepSeek R1 / V3 671B | more than 8 | — | — |
Cost: that many of the cheapest live listings. Speed: an estimate from the memory bandwidth for one person chatting. Try other models, sizes and contexts.
Decision models
| Model (Q4, 8K context) | Radeon Pro VII needed | Costs now | Per question (est.) |
|---|---|---|---|
| Julia-10.144B | 1 card | $355.00 | 16–34 ms |
| Laya0.421B | 1 card | $355.00 | 27–57 ms |
| Kev-4B4B | 1 card | $355.00 | 64–137 ms |
| lev4B | 1 card | $355.00 | 191–410 ms |
| OpenJev27B | 2 cards | $725.00 | 228–490 ms |
| Clef27B | 2 cards | $725.00 | 228–490 ms |
| Perplexity Decider v1.227B | 2 cards | $725.00 | 228–490 ms |
| mammoth-127B | 2 cards | $725.00 | 228–490 ms |
Decision models score the options you give them instead of writing text, so they're timed per question. The time is scaled from each model's reported speed on an RTX PRO 6000 by this card's compute and memory bandwidth, within about half either way. About the models.
Compare elsewhere
Search other shops for AMD Radeon Pro VII. These open the shop's own search; we don't show their prices here.



















