NVIDIA A10 - 24GB usable
24GB single-slot at 150W, the easy way to add capacity to a dense server. Only 600 GB/s, so generation trails the A30 and 3090.
Specifications
| Brand | NVIDIA |
|---|---|
| Model | A10 |
| Usable VRAM | 24GB |
| Architecture | Ampere |
| CUDA / Stream Processors | 9,216 |
| Memory Bandwidth | 600 GB/s |
| TDP | 150W |
| FP32 TFLOPS | 31.2 |
Current Offers
Used from $3500New from $3299
Prices last updated:
GPUDojo is reader-supported. When you buy through links on our site, we may earn an affiliate commission.
Price History
- eBay$3500mid-range
- Newegg$3299at high
For AI / LLM Use
Solid choice for 30B models and comfortable 14B inference. Datacenter card with no display output, may need aftermarket cooling.
What Models Can It Run?
- 30B Q4_K_M, 14B full precision, 70B Q2 (tight)
- 14B Q6_K, 30B Q3_K (tight)
- 14B Q4_K_M, 7B full precision
- 7B Q6_K, 14B Q3_K (tight)
- 7B Q4_K_M only
Estimated Performance
Generation: ~45 tokens/sec
Prefill: ~557 tokens/sec
Recommended Quantisations
- Q4_K_M recommended for 30B models
- Q6_K or Q8 for 14B and below
- Full precision for 7B
Pros & Cons
Pros
- 24GB usable VRAM: handles large models
- Only 150W TDP: power efficient
- Ampere architecture: broad CUDA software support
Cons
- Moderate memory bandwidth: not the fastest for inference
- No display output: headless only
- May need aftermarket cooling solution
Community Verdict
No community reviews yet for the A10. Know a good review? Let us know.