NVIDIA RTX 4090 vs AMD Strix Halo 128GB for local LLMs
41 reports on the NVIDIA RTX 4090 and 111 on the AMD Strix Halo 128GB
Which is faster for local LLMs?
On Qwen3.8 27B at 4-bit, the NVIDIA RTX 4090 runs at 31.5 tokens per second (2 runs) and the AMD Strix Halo 128GB at 14.0 (2 runs). 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 3.9x higher on the NVIDIA RTX 4090 (1,008 GB/s on the NVIDIA RTX 4090, 256 on the AMD Strix Halo 128GB).
- Memory
- 24GB
- Memory bandwidth
- 1,008 GB/s
- FP16 compute
- 330 TFLOPS
- Reports
- 41
- Memory
- 128GB unified
- Memory bandwidth
- 256 GB/s
- FP16 compute
- 59 TFLOPS
- Reports
- 111
Measured on both
Median tokens per second 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 RTX 4090 | AMD Strix Halo 128GB |
|---|---|---|
| Qwen3.8 27B4-bit | 31.52 runsllama.cpp128K to 194K context | 14.02 runsllama.cpp32K context, 1 not stated |
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 RTX 4090 | AMD Strix Halo 128GB |
|---|---|---|
| Qwen3.8 27B | 36.5 t/sQ4_K_M | 3.5 t/sFP16 |
| Qwen3.8 125B · 6B active | 19.6 t/sQ2_K, experts in RAM | 12.6 t/sQ6_K |
| DeepSeek V4 Flash 284B · 13B active | doesn't fit at these settings | 33.5 t/sQ2_K |
| Qwen3.6 35B · 3B active | 129.3 t/sQ4_K_S | 12.6 t/sFP16 |
| Qwen3.6 27B | 36.5 t/sQ4_K_M | 3.5 t/sFP16 |
| GLM-5.3 320B · 18B active | doesn't fit at these settings | 11.1 t/sQ2_K |
| Gemma 4 26B · 4B active | 82.3 t/sQ5_K_S | 9.5 t/sFP16 |
Change the context, RAM or model in the calculator: NVIDIA RTX 4090 · AMD Strix Halo 128GB
Every report on each card, with its source: NVIDIA RTX 4090 · AMD Strix Halo 128GB