Home › GPUs & Macs › NVIDIA DGX Spark (GB10) vs NVIDIA RTX 4000 SFF Ada
NVIDIA DGX Spark (GB10) vs NVIDIA RTX 4000 SFF Ada for local LLMs
Across 3 tests measured on both, the NVIDIA RTX 4000 SFF Ada is faster in 3 and generates a median 1.35× the tokens/s of the NVIDIA DGX Spark (GB10).
On Llama 3.1 8B Instruct (Q4_K_M, LocalScore) the NVIDIA DGX Spark (GB10) generates 34.2 tokens/s and the NVIDIA RTX 4000 SFF Ada 44.4 tokens/s (measured).
Same model, same quant, same test
| Model | Quant / test | DGX Spark (GB10) tok/s | RTX 4000 SFF Ada tok/s | Faster | Sources and run details |
|---|---|---|---|---|---|
| Llama 3.1 8B Instruct | Q4_K_M LocalScore | 34.2 llamafile | 44.4 llamafile | RTX 4000 SFF Ada +30% | LocalScore DGX Spark (GB10) · LocalScore RTX 4000 SFF Ada |
| Llama 3.2 1B Instruct | Q4_K_M LocalScore | 132 llamafile | 189 llamafile | RTX 4000 SFF Ada +43% | LocalScore DGX Spark (GB10) · LocalScore RTX 4000 SFF Ada |
| Qwen2.5 14B Instruct | Q4_K_M LocalScore | 18.3 llamafile | 24.7 llamafile | RTX 4000 SFF Ada +35% | LocalScore DGX Spark (GB10) · LocalScore RTX 4000 SFF Ada |
Generation speed in tokens/s, best measured backend for each chip (shown under the number). Backends can differ between the two chips; check them before drawing conclusions. Measured results only, nothing estimated.
Specs
| NVIDIA DGX Spark (GB10) | NVIDIA RTX 4000 SFF Ada | |
|---|---|---|
| Memory | 128 GB | 20 GB |
| Memory bandwidth | 273 GB/s | 280 GB/s |
All results for the NVIDIA DGX Spark (GB10) · All results for the NVIDIA RTX 4000 SFF Ada · More comparisons