Home › GPUs & Macs › NVIDIA A100 80GB PCIe vs NVIDIA RTX 5090
NVIDIA A100 80GB PCIe vs NVIDIA RTX 5090 for local LLMs
Across 4 tests measured on both, the NVIDIA RTX 5090 is faster in 4 and generates a median 1.63× the tokens/s of the NVIDIA A100 80GB PCIe.
On Llama 2 7B (Q4_0, llama-bench) the NVIDIA A100 80GB PCIe generates 201 tokens/s and the NVIDIA RTX 5090 300 tokens/s (measured).
Same model, same quant, same test
| Model | Quant / test | A100 80GB PCIe tok/s | RTX 5090 tok/s | Faster | Sources and run details |
|---|---|---|---|---|---|
| Llama 2 7B | Q4_0 llama-bench | 201 CUDA | 300 CUDA | RTX 5090 +50% | llama.cpp #15013 A100 80GB PCIe · llama.cpp #15013 RTX 5090 |
| Llama 3.1 8B Instruct | Q4_K_M LocalScore | 37.1 llamafile | 66.3 llamafile | RTX 5090 +79% | LocalScore A100 80GB PCIe · LocalScore RTX 5090 |
| Llama 3.2 1B Instruct | Q4_K_M LocalScore | 81.9 llamafile | 120 llamafile | RTX 5090 +46% | LocalScore A100 80GB PCIe · LocalScore RTX 5090 |
| Qwen2.5 14B Instruct | Q4_K_M LocalScore | 20.8 llamafile | 41.8 llamafile | RTX 5090 +101% | LocalScore A100 80GB PCIe · LocalScore RTX 5090 |
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 A100 80GB PCIe | NVIDIA RTX 5090 | |
|---|---|---|
| Memory | 80 GB | 32 GB |
| Memory bandwidth | 1935 GB/s | 1792 GB/s |
All results for the NVIDIA A100 80GB PCIe · All results for the NVIDIA RTX 5090 · More comparisons