llamaperf

Intel Arc Pro B60 24GB vs AMD Strix Halo 128GB for local LLMs

9 reports on the Intel Arc Pro B60 24GB and 348 on the AMD Strix Halo 128GB

Which is faster for local LLMs?

On Qwen3.8 27B at 4-bit, the Intel Arc Pro B60 24GB has 18.8 generation tokens/s (4 runs); prompt processing 339.4 tokens/s (3 runs); the AMD Strix Halo 128GB has 13.1 generation tokens/s (4 runs); prompt processing 378.2 tokens/s (3 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 1.8x higher on the Intel Arc Pro B60 24GB (456 GB/s on the Intel Arc Pro B60 24GB, 256 on the AMD Strix Halo 128GB). Across 2 shared models with generation measurements on both cards, the Intel Arc Pro B60 24GB's median is higher on 1 and the AMD Strix Halo 128GB's on 1.

Intel Arc Pro B60 24GB
Memory
24GB
Memory bandwidth
456 GB/s
FP16 compute
98.5 TFLOPS
Reports
9
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

ModelIntel Arc Pro B60 24GBAMD Strix Halo 128GB
Qwen3.8 27B4-bit18.84 runsPP 339.4 (3 runs)llama.cpp, 1 not stated512 to 150K context, 1 not stated13.14 runsPP 378.2 (3 runs)llama.cpp, 1 not stated32K context, 3 not stated
Qwen3.5 35B · 3B active4-bit8.3one runPP 112.6 (one run)llama.cppcontext not stated52.42 runsPP 1017.0 (one run)llama.cpp, 1 not stated128K 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.

ModelIntel Arc Pro B60 24GBAMD Strix Halo 128GB
Qwen3.8 27B16.5 t/sQ4_K_M3.5 t/sFP16
Qwen3.5 35B · 3B active58.5 t/sQ4_K_S12.6 t/sFP16
Qwen3.8 125B · 6B active17.8 t/sQ2_K, experts in RAM12.6 t/sQ6_K
Qwen3.6 35B · 3B active58.5 t/sQ4_K_S12.6 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 27B16.5 t/sQ4_K_M3.5 t/sFP16
Gemma 4 26B · 4B active37.2 t/sQ5_K_S9.5 t/sFP16

Change the context, RAM or model in the calculator: Intel Arc Pro B60 24GB · AMD Strix Halo 128GB

Every report on each card, with its source: Intel Arc Pro B60 24GB · AMD Strix Halo 128GB

More comparisons