AMD Radeon Pro W7800 48 GB
Live eBay prices, cheapest first. The best one right now is $5,090 (new).
Other sizes: 32 GB
Specs
- VRAM
- 48 GB GDDR6
- Bandwidth
- 576 GB/s
- FP32
- 45.2 TFLOPS
- FP16
- 90 TFLOPS
- Power
- 260 W
- Connector
- PCIe 4.0
- Released
- 2023
- Llama 3.1 8B at Q4 (est.)
- 42–71 tokens/s
Price right now
Live on eBayFrom the live listings below; shipping included.
Listings
Live on eBay| Listing | Condition | Price | Checked |
|---|---|---|---|
| New | $5,090or best offerfree shipping | ||
| New | $5,126or best offerfree 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- Software: ROCm (on AMD's support list).
As much VRAM, about as fast, for less
- NVIDIA Quadro RTX 8000: $2,000, 115% of the GeForce RTX 3090, 48 GB
- Intel Arc Pro B60 48 GB: $2,049, 82% of the GeForce RTX 3090, 48 GB
- NVIDIA A16: $2,200, 93% of the GeForce RTX 3090, 64 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 Tesla V100 PCIe 16 GB: $325.00, 123% of the GeForce RTX 3090, 16 GB
More VRAM, for not much more
- NVIDIA A16: $2,200, 93% of the GeForce RTX 3090, 64 GB
- AMD Instinct MI210: $3,799, 211% of the GeForce RTX 3090, 64 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 W7800 48 GB needed | Costs now | Speed (est.) |
|---|---|---|---|
| Llama 3.1 8B | 1 card | $5,090 | 42–71 tokens/s |
| Qwen3 14B | 1 card | $5,090 | 22–39 tokens/s |
| gpt-oss 20B | 1 card | $5,090 | 55–95 tokens/s |
| Gemma 3 27B | 1 card | $5,090 | 12–21 tokens/s |
| Qwen3 32B | 1 card | $5,090 | 10–17 tokens/s |
| Llama 3.3 70B | 1 card | $5,090 | 4.7–8.1 tokens/s |
| gpt-oss 120B | 2 cards | $10,216 | 39–67 tokens/s |
| Qwen3 235B-A22B | 4 cards | not enough listings | 9.1–16 tokens/s |
| 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 W7800 48 GB needed | Costs now | Per question (est.) |
|---|---|---|---|
| Julia-10.144B | 1 card | $5,090 | 11–25 ms |
| Laya0.421B | 1 card | $5,090 | 19–41 ms |
| Kev-4B4B | 1 card | $5,090 | 46–98 ms |
| lev4B | 1 card | $5,090 | 138–295 ms |
| OpenJev27B | 1 card | $5,090 | 165–353 ms |
| Clef27B | 1 card | $5,090 | 165–353 ms |
| Perplexity Decider v1.227B | 1 card | $5,090 | 165–353 ms |
| mammoth-127B | 1 card | $5,090 | 165–353 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 W7800 48 GB. These open the shop's own search; we don't show their prices here.
