Can Qwen 3.5 397B A17B run on AMD Instinct MI325X 256GB?

Q: What should I upgrade first if Qwen 3.5 397B A17B feels slow on AMD Instinct MI325X 256GB?

Remove offload with more accelerator memory. Prioritize a GPU or unified-memory tier that fits the whole model natively. Removing offload usually helps more than small compute gains.

YES — With Offload

A84Great

Estimated from fit model

Qwen 3.5 397B A17B needs ~271.5 GB VRAM. AMD Instinct MI325X 256GB has 256.0 GB. With Q4_K_M quantization, expect ~39 tok/s.

Runtime: llama.cppCapacity: OffloadBandwidth: HighStack: StandardBottleneck: Host offload

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) — 271.5 GB, 39.2 tok/s, Runs with offload (needs ~13.9 GB host RAM)

271.5 GB required256.0 GB available

106% VRAM needed

15.5 GB over capacity — needs offload or smaller quantization

Fit status

Runs with offload (needs ~13.9 GB host RAM)

Decode

39.2 tok/s

TTFT

4938 ms

Safe context

Memory

271.5 GB / 256.0 GB

Offload

10%

Memory breakdown

Weights242.2 GB

KV Cache2.9 GB

Runtime0.9 GB

Headroom25.6 GB

See how fast it feels

See how fast it feelsQwen 3.5 397B A17B on AMD Instinct MI325X 256GB
1st promptCold start — includes initialization
>Write a Python function to sort a list
Here's a Python function that sorts a list using quicksort:

def quicksort(arr):
    if len(arr) <= 1:
        return arr
    pivot = arr[len(arr) // 2]
    left   = [x for x in arr if x < pivot]
    middle = [x for x in arr if x == pivot]
    right  = [x for x in arr if x > pivot]
    return quicksort(left) + middle + quicksort(right)

# Example usage
numbers = [3, 6, 8, 10, 1, 2, 1]
print(quicksort(numbers))  # [1, 1, 2, 3, 6, 8, 10]

The algorithm picks a pivot, partitions the list into smaller, equal, and larger elements, then recursively sorts each partition and concatenates the results.
2nd promptWarm — model ready, no init overhead
>Now add type hints and a docstring
def quicksort(arr: list[int]) -> list[int]:
    """Sort a list using the quicksort algorithm.

    Args:
        arr: List of integers to sort.

    Returns:
        A new sorted list.
    """
    if len(arr) <= 1:
        return arr
    pivot = arr[len(arr) // 2]
    left   = [x for x in arr if x < pivot]
    middle = [x for x in arr if x == pivot]
    right  = [x for x in arr if x > pivot]
    return quicksort(left) + middle + quicksort(right)

Added type hints for the input and return type, plus a Google-style docstring.
Estimated: 39.2 tok/s decode · 4.9s TTFT (warm) · 98 tok/s prefill

What limits this setup

It fits through host-memory offload, and offload is the main reason performance drops.

CPU or host-memory offload is active

About 10% of the working set spills out of accelerator memory, which usually hurts latency and sustained decode throughput.

Very little memory headroom

You can run the model, but there is not much room left for longer context, bigger batches, extra apps, or future model updates.

Best improvement path

Remove offload with more accelerator memory

Prioritize a GPU or unified-memory tier that fits the whole model natively. Removing offload usually helps more than small compute gains.

Buy headroom, not only minimum fit

A slightly larger memory tier gives you safer context growth and makes the recommendation more future-proof.

Increase host RAM if you keep offloading

This setup may need roughly 13.9 GB of extra host RAM just for the offloaded portion, before OS and other tools.

Performance by workload

Workload	Grade	Fit	Decode	TTFT	Context
Chat	A	Runs with offload (needs ~12.6 GB host RAM)	39.6 tok/s	2663 ms	4K
Coding	A	Runs with offload (needs ~13.9 GB host RAM)	39.2 tok/s	4938 ms	4K
Agentic Coding	A	Runs with offload (needs ~16.2 GB host RAM)	38.4 tok/s	7343 ms	4K
Reasoning	A	Runs with offload (needs ~13.9 GB host RAM)	39.2 tok/s	5835 ms	4K
RAG	A	Runs with offload (needs ~16.2 GB host RAM)	38.4 tok/s

Quantization options

How Qwen 3.5 397B A17B (397B params) fits at each quantization level on AMD Instinct MI325X 256GB (256.0 GB usable).

Quant	Bits	VRAM	Quality	Fit
Q2_K	2	154.8 GB	Low	S91
Q3_K_SBest for your GPU	3	194.5 GB	Low	S91

Get started

Copy-paste commands to run Qwen 3.5 397B A17B on your machine.

Run

docker run --rm -it ghcr.io/ggerganov/llama.cpp:full \
  --hf-repo "Qwen/Qwen3.5-397B-A17B-Instruct" \
  --hf-file "Qwen3.5-397B-A17B-Instruct-Q4_K_M.gguf" \
  -c 4096 -ngl 99

Frequently asked questions

Can AMD Instinct MI325X 256GB run Qwen 3.5 397B A17B?

Yes, AMD Instinct MI325X 256GB can run Qwen 3.5 397B A17B with a A grade (Runs with offload (needs ~13.9 GB host RAM)). Expected decode speed: 39.2 tok/s.

How much VRAM does Qwen 3.5 397B A17B need?

Qwen 3.5 397B A17B (397B parameters) requires approximately 271.5 GB of memory with Q4_K_M quantization.

What is the best quantization for Qwen 3.5 397B A17B?

The recommended quantization for Qwen 3.5 397B A17B is Q4_K_M, which balances quality and memory efficiency.

What speed will Qwen 3.5 397B A17B run at on AMD Instinct MI325X 256GB?

On AMD Instinct MI325X 256GB, Qwen 3.5 397B A17B achieves approximately 39.2 tokens per second decode speed with a time-to-first-token of 4938ms using Q4_K_M quantization.

Can AMD Instinct MI325X 256GB run Qwen 3.5 397B A17B for coding?

For coding workloads, Qwen 3.5 397B A17B on AMD Instinct MI325X 256GB receives a A grade with 39.2 tok/s and 4K context.

What context window can Qwen 3.5 397B A17B use on AMD Instinct MI325X 256GB?

On AMD Instinct MI325X 256GB, Qwen 3.5 397B A17B can safely use up to 4K tokens of context. The model's official context limit is 131K, but available memory constrains the safe maximum.

See all results for AMD Instinct MI325X 256GB See all hardware for Qwen 3.5 397B A17B

Embed this result▼

Paste this snippet into any page to show a live fit card.

<iframe src="https://willitrunai.com/embed/qwen-3.5-397b-a17b-on-instinct-mi325x-256gb" width="400" height="180" frameborder="0" style="border:none;border-radius:12px;overflow:hidden;" title="Will It Run AI — fit result"></iframe>

Preview: