Home › GPUs & Macs › NVIDIA GTX 1660 SUPER vs NVIDIA RTX 3060
NVIDIA GTX 1660 SUPER vs NVIDIA RTX 3060 for local LLMs
Across 2 tests measured on both, the NVIDIA RTX 3060 is faster in 2 and generates a median 4.91× the tokens/s of the NVIDIA GTX 1660 SUPER.
On Llama 3.1 8B Instruct (Q4_K_M, LocalScore) the NVIDIA GTX 1660 SUPER generates 9.6 tokens/s and the NVIDIA RTX 3060 44.1 tokens/s (measured).
Same model, same quant, same test
| Model | Quant / test | GTX 1660 SUPER tok/s | RTX 3060 tok/s | Faster | Sources and run details |
|---|---|---|---|---|---|
| Llama 3.1 8B Instruct | Q4_K_M LocalScore | 9.6 llamafile | 44.1 llamafile | RTX 3060 +360% | LocalScore GTX 1660 SUPER · LocalScore RTX 3060 |
| Llama 3.2 1B Instruct | Q4_K_M LocalScore | 35.2 llamafile | 185 llamafile | RTX 3060 +426% | LocalScore GTX 1660 SUPER · LocalScore RTX 3060 |
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 GTX 1660 SUPER | NVIDIA RTX 3060 | |
|---|---|---|
| Memory | 6 GB | 12 GB |
| Memory bandwidth | 336 GB/s | 360 GB/s |
All results for the NVIDIA GTX 1660 SUPER · All results for the NVIDIA RTX 3060 · More comparisons