Can Qwen3.5 35B A3B run on AMD Instinct MI210 64GB?

YES — Runs Great

C51Usable

○Estimated from fit model

Qwen3.5 35B A3B needs ~32.8 GB VRAM. AMD Instinct MI210 64GB has 64.0 GB. With Q4_K_M quantization, expect ~52 tok/s.

Runtime: llama.cppCapacity: RoomyBandwidth: HighStack: StandardBottleneck: Balanced

Operating mode

Choose the run profile you care about

Interactive favors responsiveness, while light API and scale-out lean harder on serving readiness. The fit stays the same, but the recommendation lens changes.

Current mode

Balanced

Balanced for general local use. Keeps the ranking neutral across personal and serving workflows.

Capabilities:

Select quantization to explore

Q4_K_M (Medium quality) — 32.8 GB, 52.2 tok/s, Runs well

32.8 GB required64.0 GB available

51% VRAM used

Fit status

Runs well

Decode

52.2 tok/s

TTFT

3711 ms

Safe context

138K

Memory

32.8 GB / 64.0 GB

Memory breakdown

Weights21.3 GB

KV Cache4.1 GB

Runtime0.9 GB

Headroom6.4 GB

See how fast it feels

See how fast it feelsQwen3.5 35B A3B on AMD Instinct MI210 64GB
1st promptCold start — includes initialization
>What is local AI inference?
Local AI inference means running an AI language model directly on your own hardware — your laptop, desktop, or server — instead of sending requests to a remote cloud API.

When you run inference locally the model weights are loaded into your GPU or unified memory. Each token you generate requires reading those weights from memory, so memory bandwidth is the main bottleneck for decode speed.

Key benefits of running locally:
 - Full privacy: your prompts never leave your machine
 - No per-token cost or rate limits
 - Works offline once the model is downloaded
 - Latency depends only on your hardware
2nd promptWarm — model ready, no init overhead
>How much VRAM do I need?
It depends on the model size and quantization level. A rough rule of thumb:

 Model size Q4 (4-bit) Q8 (8-bit) FP16
 7B params ~4.3 GB ~7.5 GB ~14 GB
 13B params ~7.9 GB ~13.9 GB ~26 GB
 70B params ~42.7 GB ~74.9 GB ~140 GB

Most people use 4-bit quantization (Q4_K_M) which gives 90-95% of full quality at a fraction of the memory. A 24 GB GPU can comfortably run most 7B-13B models.
Estimated: 52.2 tok/s decode · 3.7s TTFT (warm) · 130 tok/s prefill

What limits this setup

This setup is broadly balanced for this model.

No major red flags

This recommendation has enough memory headroom and acceptable estimated speed for the selected workload.

Best improvement path

Performance by workload

Workload	Grade	Fit	Decode	TTFT	Context
Chat	C	Runs well	52.2 tok/s	2024 ms	138K
Coding	C	Runs well	52.2 tok/s	3711 ms	138K
Agentic Coding	C	Runs well	52.2 tok/s	5398 ms	138K
Reasoning	C	Runs well	52.2 tok/s	4386 ms	138K
RAG	C	Runs well	52.2 tok/s	6747 ms	138K

Quantization options

How Qwen3.5 35B A3B (35B params) fits at each quantization level on AMD Instinct MI210 64GB (64.0 GB usable).

Quant	Bits	VRAM	Quality	Fit
Q2_K	2	13.7 GB	Low	C42
Q3_K_S	3	17.2 GB	Low	C43
NVFP4	4

Get started

Copy-paste commands to run Qwen3.5 35B A3B on your machine.

Run

docker run --rm -it ghcr.io/ggerganov/llama.cpp:full \
 --hf-repo "lmstudio-community/Qwen3.5-35B-A3B-GGUF" \
 --hf-file "Qwen3.5-35B-A3B-GGUF-Q4_K_M.gguf" \
 -c 4096 -ngl 99

Upgrade options

Hardware that runs Qwen3.5 35B A3B well

👁 NVIDIA
NVIDIA H800 80GBBudget pick

80 GB VRAM (+16)3000 GB/s (+1362)

Raises estimated decode speed by about 118%.113.8 tok/s decode

Raises estimated decode speed by about 118%.

Adds memory headroom for longer context windows and future model growth.

~$30,000 MSRP

👁 NVIDIA
NVIDIA H100 80GBBest value

80 GB VRAM (+16)3350 GB/s (+1712)

Raises estimated decode speed by about 152%.131.8 tok/s decode

Raises estimated decode speed by about 152%.