AMD Radeon 780M vs AMD Strix Halo 128GB for local LLMs
8 reports on the AMD Radeon 780M and 348 on the AMD Strix Halo 128GB
Which is faster for local LLMs?
On Qwen3.6 35B · 3B active at 8-bit, the AMD Radeon 780M has 19.7 generation tokens/s (2 runs); prompt processing 299.4 tokens/s (2 runs); the AMD Strix Halo 128GB has 50.4 generation tokens/s (4 runs); prompt processing 1353.3 tokens/s (4 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 2.1x higher on the AMD Strix Halo 128GB (120 GB/s on the AMD Radeon 780M, 256 on the AMD Strix Halo 128GB). Across 2 shared models with generation measurements on both cards, the AMD Radeon 780M's median is higher on 0 and the AMD Strix Halo 128GB's on 2.
- Memory
- shared memory unified
- Memory bandwidth
- not on record
- FP16 compute
- not on record
- Reports
- 8
- Memory
- 128GB unified
- Memory bandwidth
- 256 GB/s
- FP16 compute
- 59 TFLOPS
- Reports
- 348
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 | AMD Radeon 780M | AMD Strix Halo 128GB |
|---|---|---|
| Qwen3.6 35B · 3B active8-bit | 19.72 runsPP 299.4 (2 runs)llama.cppcontext not stated | 50.44 runsPP 1353.3 (4 runs)llama.cpp, Gufo, 2 not stated512K context, 4 not stated |
| Gemma 4 26B · 4B active4-bit | 25.0one runPP 208.7 (one run)llama.cppcontext not stated | 67.32 runsPP 4061.0 (2 runs)Gufocontext 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 | AMD Radeon 780M | AMD Strix Halo 128GB |
|---|---|---|
| Qwen3.6 35B · 3B active | not in the calculator yet | 12.6 t/sFP16 |
| Gemma 4 26B · 4B active | not in the calculator yet | 9.5 t/sFP16 |
| Qwen3.8 27B | not in the calculator yet | 3.5 t/sFP16 |
| Qwen3.8 125B · 6B active | not in the calculator yet | 12.6 t/sQ6_K |
| DeepSeek V4 Flash 284B · 13B active | not in the calculator yet | 33.5 t/sQ2_K |
| DeepSeek V4.1 Flash 552B · 16B active | not in the calculator yet | doesn't fit at these settings |
| Qwen3.6 27B | not in the calculator yet | 3.5 t/sFP16 |
| GLM-5.3 320B · 18B active | not in the calculator yet | 11.1 t/sQ2_K |
Change the context, RAM or model in the calculator: AMD Radeon 780M · AMD Strix Halo 128GB
Every report on each card, with its source: AMD Radeon 780M · AMD Strix Halo 128GB