NVIDIA A100 SXM4 80 GB
Live eBay prices, cheapest first. The best one right now is $10,450 (used).
Not a PCIe card. SXM module: needs an SXM-to-PCIe adapter board, its own cooling and power to run in a normal PC. The PCIe version: NVIDIA A100 PCIe 80 GB, from $10,999 →
Specs
- VRAM
- 80 GB HBM2e
- Bandwidth
- 2,039 GB/s
- FP32
- 19.5 TFLOPS
- FP16
- 312 TFLOPS
- Power
- 400 W
- Connector
- SXM4 (PCIe 4.0)
- Released
- 2020
- CUDA compute
- 8.0
- Llama 3.1 8B at Q4 (est.)
- 210–315 tokens/s
Price right now
Live on eBayFrom the live listings below; shipping included.
Listings
Live on eBay| Listing | Condition | Price | Checked |
|---|---|---|---|
| Used | $10,450−25%shipping not listed | ||
| Used | $10,500or best offer−25%free shipping | ||
| Used | $12,064or best offer−14%free shipping | ||
| Used | $12,068or best offer−14%free shipping | ||
| Used | $13,995or best offerfree shipping | ||
| Used | $13,995or best offerfree shipping | ||
| Used | $13,995free shipping | ||
| Used | $13,995or best offerfree shipping | ||
| Used | $13,995or best offerfree shipping | ||
| Used | $16,999or best offerfree shipping | ||
| Open box | $19,800or best offerfree shipping | ||
| Open box | $20,290or best offerfree shipping | ||
| Used | $23,550or best offerfree shipping | ||
| Used | $26,880or best offerfree shipping | ||
| Used | $26,898or 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- #3 of 9 GPUs with 80 GB or more for the money: $3,378 per GeForce RTX 3090 of performance.
- Software: CUDA 11.0 and newer.
As much VRAM, about as fast, for less
- AMD Instinct MI250: $4,995, 422% of the GeForce RTX 3090, 128 GB
Faster, for the least
- NVIDIA RTX PRO 6000 Blackwell: $13,100, 367% of the GeForce RTX 3090, 96 GB
- NVIDIA H100 PCIe: $26,935, 477% of the GeForce RTX 3090, 80 GB
- AMD Instinct MI300X: $27,550, 1021% of the GeForce RTX 3090, 192 GB
More VRAM, for not much more
- AMD Instinct MI250: $4,995, 422% of the GeForce RTX 3090, 128 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) | A100 SXM4 80 GB needed | Costs now | Speed (est.) |
|---|---|---|---|
| Llama 3.1 8B | 1 card | $10,450 | 210–315 tokens/s |
| Qwen3 14B | 1 card | $10,450 | 114–170 tokens/s |
| gpt-oss 20B | 1 card | $10,450 | 280–420 tokens/s |
| Gemma 3 27B | 1 card | $10,450 | 61–92 tokens/s |
| Qwen3 32B | 1 card | $10,450 | 51–77 tokens/s |
| Llama 3.3 70B | 1 card | $10,450 | 24–36 tokens/s |
| gpt-oss 120B | 1 card | $10,450 | 198–297 tokens/s |
| Qwen3 235B-A22B | 2 cards | $20,950 | 46–69 tokens/s |
| DeepSeek R1 / V3 671B | 6 cards | $73,072 | 27–41 tokens/s |
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) | A100 SXM4 80 GB needed | Costs now | Per question (est.) |
|---|---|---|---|
| Julia-10.144B | 1 card | $10,450 | 2.5–5.3 ms |
| Laya0.421B | 1 card | $10,450 | 4.2–8.9 ms |
| Kev-4B4B | 1 card | $10,450 | 10.0–21 ms |
| lev4B | 1 card | $10,450 | 30–64 ms |
| OpenJev27B | 1 card | $10,450 | 36–77 ms |
| Clef27B | 1 card | $10,450 | 36–77 ms |
| Perplexity Decider v1.227B | 1 card | $10,450 | 36–77 ms |
| mammoth-127B | 1 card | $10,450 | 36–77 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 NVIDIA A100 SXM4 80 GB. These open the shop's own search; we don't show their prices here.














