llamaperf

AMD Radeon Pro W7800 48GB vs AMD Strix Halo 128GB for local LLMs

2 reports on the AMD Radeon Pro W7800 48GB and 348 on the AMD Strix Halo 128GB

Which is faster for local LLMs?

On Qwen3.8 125B · 6B active at 3-bit, the AMD Radeon Pro W7800 48GB has 90.0 generation tokens/s (one run); the AMD Strix Halo 128GB has 27.4 generation tokens/s (2 runs); prompt processing 873.0 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 3.4x higher on the AMD Radeon Pro W7800 48GB (864 GB/s on the AMD Radeon Pro W7800 48GB, 256 on the AMD Strix Halo 128GB).

AMD Radeon Pro W7800 48GB
Memory
48GB
Memory bandwidth
864 GB/s
FP16 compute
90.4 TFLOPS
Reports
2
AMD Strix Halo 128GB
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

ModelAMD Radeon Pro W7800 48GBAMD Strix Halo 128GB
Qwen3.8 125B · 6B active3-bit90.0one runllama.cpp128K context27.42 runsPP 873.0 (one run)Strata, 1 not stated25K 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.

ModelAMD Radeon Pro W7800 48GBAMD Strix Halo 128GB
Qwen3.8 125B · 6B active80.1 t/sQ2_K12.6 t/sQ6_K
Qwen3.8 27B17.6 t/sQ8_03.5 t/sFP16
Qwen3.6 35B · 3B active55.4 t/sQ8_012.6 t/sFP16
DeepSeek V4 Flash 284B · 13B activedoesn't fit at these settings33.5 t/sQ2_K
DeepSeek V4.1 Flash 552B · 16B activedoesn't fit at these settingsdoesn't fit at these settings
Qwen3.6 27B17.6 t/sQ8_03.5 t/sFP16
Gemma 4 26B · 4B active47.5 t/sQ8_09.5 t/sFP16

Change the context, RAM or model in the calculator: AMD Radeon Pro W7800 48GB · AMD Strix Halo 128GB

Every report on each card, with its source: AMD Radeon Pro W7800 48GB · AMD Strix Halo 128GB

More comparisons