NVIDIA CMP 170HX 64GB (unlocked) vs NVIDIA RTX 3090 for local LLMs
15 reports on the NVIDIA CMP 170HX 64GB (unlocked) and 271 on the NVIDIA RTX 3090
Which is faster for local LLMs?
On Qwen3.6 27B at 4-bit, the NVIDIA CMP 170HX 64GB (unlocked) has 38.0 generation tokens/s (one run); prompt processing 907.0 tokens/s (one run); the NVIDIA RTX 3090 has 42.0 generation tokens/s (one run); prompt processing 1020.0 tokens/s (one run). Those are community reports whose engines and settings differ, so the gap is a rough sign, not a controlled measure of the two cards. Memory bandwidth, which caps how fast a card can write, is 1.6x higher on the NVIDIA CMP 170HX 64GB (unlocked) (1,493 GB/s on the NVIDIA CMP 170HX 64GB (unlocked), 936 on the NVIDIA RTX 3090).
- Memory
- 64GB
- Memory bandwidth
- 1,493 GB/s
- FP16 compute
- not on record
- Reports
- 15
- Memory
- 24GB
- Memory bandwidth
- 936 GB/s
- FP16 compute
- 142 TFLOPS
- Reports
- 271
Measured on both
Generation and prompt processing speeds from plain runs (one device, one request, no speculative decoding, the whole model in the device's memory) of the same model size at the same bit class. Each side shows the engines and reported context behind its median. Context can mean a configured window or prompt depth, so these are community comparisons with differing settings. How to read these
| Model | NVIDIA CMP 170HX 64GB (unlocked) | NVIDIA RTX 3090 |
|---|---|---|
| Qwen3.6 27B4-bit | 38.0one runPP 907.0 (one run)llama.cppcontext not stated | 42.0one runPP 1020.0 (one run)llama.cpp100K context |
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. Where a model was also measured above, the measurement wins.
| Model | NVIDIA CMP 170HX 64GB (unlocked) | NVIDIA RTX 3090 |
|---|---|---|
| Qwen3.6 27B | 30.4 t/sQ8_0 | 33.9 t/sQ4_K_M |
| Qwen3.8 27B | 30.4 t/sQ8_0 | 33.9 t/sQ4_K_M |
| Qwen3.8 125B · 6B active | 127.0 t/sQ3_K_S | 19.4 t/sQ2_K, experts in RAM |
| Qwen3.6 35B · 3B active | 109.5 t/sQ8_0 | 120.1 t/sQ4_K_S |
| DeepSeek V4 Flash 284B · 13B active | doesn't fit at these settings | doesn't fit at these settings |
| DeepSeek V4.1 Flash 552B · 16B active | doesn't fit at these settings | doesn't fit at these settings |
| Gemma 4 26B · 4B active | 82.1 t/sQ8_0 | 76.4 t/sQ5_K_S |
Change the context, RAM or model in the calculator: NVIDIA CMP 170HX 64GB (unlocked) · NVIDIA RTX 3090
Every report on each card, with its source: NVIDIA CMP 170HX 64GB (unlocked) · NVIDIA RTX 3090