How fast does Qwen3.5 35B-A3B run on the Intel Arc Pro B70?

Qwen3.5 35B-A3B runs at about 54.5 tokens/s on the Intel Arc Pro B70 (measured, Q4_K_M, SYCL, llama-bench). Based on 1 measured run(s), each linked to its public source.

Does Qwen3.5 35B-A3B Q4_K_M fit in the Intel Arc Pro B70's memory?

Estimate: the Q4_K_M file is 22.0 GB and needs about 23.7 GB with an ~8K context; the Intel Arc Pro B70 has about 31.5 GB usable, so it fits.

Home › Models › Qwen3.5 35B-A3B › Intel Arc Pro B70

Qwen3.5 35B-A3B on the Intel Arc Pro B70

Qwen3.5 35B-A3B runs at about 54.5 tokens/s on the Intel Arc Pro B70 (measured, Q4_K_M, SYCL, llama-bench).

Best measured generation54.5tok/s

Q4_K_M · SYCL · FA on

Prompt processing618tok/s

same run

For comparison, our size-scaling estimate for this combination is 121 tok/s.

Every run

Quant	Backend	Generate tok/s	Prompt tok/s	Test	Build / by / notes	Source
Q4_K_M	SYCL FA on	54.5	618	tg128	ec6f7a6a5c PMZFX SYCL on one Arc Pro B70, context 4096, KV f16/f16; build b8840-12-gec6f7a6a5-dirty: the author's notes say the local change is a Q2_K DMMV work-in-progress patch	PMZFX intel-arc-pro-b70-benchmarks (MIT)

tg128 = generating 128 tokens; pp512 = reading a 512-token prompt (llama-bench). LocalScore averages several prompt and output lengths, so its numbers run lower than tg128. "decode, llama-server" = single-stream generation speed measured against a running llama-server (no speculative decoding).

Check prices

Amazon Newegg

As an Amazon Associate I earn from qualifying purchases.

No card yet?

Try it on a rented GPU (RunPod)

Affiliate link: if you sign up, RigBench may earn a referral credit. It never affects rankings or numbers.

All results for the Intel Arc Pro B70 · Qwen3.5 35B-A3B on other hardware

Embed this result

HTML

<a href="https://rigbench.pages.dev/model/qwen3-5-35b-a3b/on/intel-arc-pro-b70/"><img src="https://rigbench.pages.dev/badge/qwen3-5-35b-a3b--intel-arc-pro-b70.svg" alt="Qwen3.5 35B-A3B · Intel Arc Pro B70: 54.5 tok/s (Q4_K_M, measured)"></a>

Markdown (GitHub README, forums)

[![Qwen3.5 35B-A3B · Intel Arc Pro B70: 54.5 tok/s (Q4_K_M, measured)](https://rigbench.pages.dev/badge/qwen3-5-35b-a3b--intel-arc-pro-b70.svg)](https://rigbench.pages.dev/model/qwen3-5-35b-a3b/on/intel-arc-pro-b70/)