AMD Radeon Pro W6600 vs NVIDIA RTX 3090 for local LLMs
0 reports on the AMD Radeon Pro W6600 and 215 on the NVIDIA RTX 3090
Which is faster for local LLMs?
Nobody has reported the same model at the same quant on both cards yet, so there's no measured head-to-head. Memory bandwidth, which caps how fast a card can write, is 4.2x higher on the NVIDIA RTX 3090 (224 GB/s on the AMD Radeon Pro W6600, 936 on the NVIDIA RTX 3090). The estimates below compare them on the models people run most.
AMD Radeon Pro W6600
- Memory
- 8GB
- Memory bandwidth
- 224 GB/s
- FP16 compute
- 20.8 TFLOPS
- Reports
- 0
NVIDIA RTX 3090
- Memory
- 24GB
- Memory bandwidth
- 936 GB/s
- FP16 compute
- 142 TFLOPS
- Reports
- 215
Estimated by the calculator
The quant the calculator recommends for each card and its estimated speed at a 32,768-token context with 32 GB of system RAM, from memory bandwidth and the model's shape.
| Model | AMD Radeon Pro W6600 | NVIDIA RTX 3090 |
|---|---|---|
| Qwen3.8 27B | 1.5 t/sQ6_K, layers on CPU | 33.9 t/sQ4_K_M |
| Qwen3.8 125B · 6B active | doesn't fit at these settings | 19.4 t/sQ2_K, experts in RAM |
| Qwen3.6 35B · 3B active | 13.8 t/sQ5_K_M, experts in RAM | 120.1 t/sQ4_K_S |
| DeepSeek V4 Flash 284B · 13B active | doesn't fit at these settings | doesn't fit at these settings |
| Qwen3.6 27B | 1.5 t/sQ6_K, layers on CPU | 33.9 t/sQ4_K_M |
| Gemma 4 26B · 4B active | 8.8 t/sQ6_K, experts in RAM | 76.4 t/sQ5_K_S |
Change the context, RAM or model in the calculator: AMD Radeon Pro W6600 · NVIDIA RTX 3090
Every report on each card, with its source: AMD Radeon Pro W6600 · NVIDIA RTX 3090