Can Qwen3-Coder 480B A35B Instruct run on AMD Instinct MI350X 288GB?

BARELY — Tight on Memory

A79Great

Estimated from fit model

Qwen3-Coder 480B A35B Instruct needs ~325.4 GB VRAM. AMD Instinct MI350X 288GB has 288.0 GB. With Q4_K_M quantization, expect ~35 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) — 325.4 GB, 35.3 tok/s, Very compromised (needs ~33.6 GB host RAM)

325.4 GB required288.0 GB available

113% VRAM needed

37.4 GB over capacity — needs offload or smaller quantization

Fit status

Very compromised (needs ~33.6 GB host RAM)

Decode

35.3 tok/s

TTFT

5482 ms

Safe context

Memory

325.4 GB / 288.0 GB

Offload

10%

Memory breakdown

Weights292.8 GB

KV Cache2.9 GB

Runtime0.9 GB

Headroom28.8 GB

See how fast it feels

See how fast it feelsQwen3-Coder 480B A35B Instruct on AMD Instinct MI350X 288GB
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: 35.3 tok/s decode · 5.5s TTFT (warm) · 88 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 33.6 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	Very compromised (needs ~32.5 GB host RAM)	35.6 tok/s	2962 ms	4K
Coding	A	Very compromised (needs ~33.6 GB host RAM)	35.3 tok/s	5482 ms	4K
Agentic Coding	A	Very compromised (needs ~35.9 GB host RAM)	34.7 tok/s	8122 ms	4K
Reasoning	A	Very compromised (needs ~33.6 GB host RAM)	35.3 tok/s	6478 ms	4K
RAG	A	Very compromised (needs ~35.9 GB host RAM)	34.7 tok/s	10152 ms	4K

Inference speed

Qwen3-Coder 480B A35B Instruct inference speed — tokens per second by GPU & Mac

Estimated decode speed (tokens/sec) for Qwen3-Coder 480B A35B Instruct at Q4_K_M across popular GPUs and Apple Silicon, including multi-GPU rigs, using the fastest local runtime per device. Fastest is Mac Studio M3 Ultra 256GB at ~4 tok/s. Speed is memory-bandwidth bound, so cards that fit the whole model in VRAM run far faster than ones that offload to system RAM.

GPU / Mac	Memory	Quant	Speed (tok/s)	Fits?
Mac Studio M3 Ultra 256GB	256 GB	Q4_K_M	4.4	Too big
Mac Studio M2 Ultra 128GB	128 GB	Q4_K_M	3.4	Too big
Mac Studio M1 Ultra 128GB	128 GB	Q4_K_M	3.2	Too big
MacBook Pro M4 Max 128GB	128 GB	Q4_K_M	2.5	Too big
MacBook Pro M4 Max 64GB	64 GB	Q4_K_M	2.5	Too big
RTX 5090 32GB	32 GB	Q4_K_M	2.0	Too big
RTX 4090 24GB	24 GB	Q4_K_M	2.0	Too big
RTX 4080 Super 16GB	16 GB	Q4_K_M	2.0	Too big
RTX 3090 24GB	24 GB	Q4_K_M	2.0	Too big
RTX 4070 12GB	12 GB	Q4_K_M	2.0	Too big
RTX 3060 12GB	12 GB	Q4_K_M	2.0	Too big
RTX 4060 8GB	8 GB	Q4_K_M	2.0	Too big
RX 7900 XTX 24GB	24 GB	Q4_K_M	2.0	Too big
MacBook Pro M3 Max 64GB	64 GB	Q4_K_M	2.0	Too big
MacBook Pro M1 Max 64GB	64 GB	Q4_K_M	2.0	Too big
MacBook Pro M4 Pro 48GB	48 GB	Q4_K_M	2.0	Too big
2× RTX 4090 24GB	48 GB	Q4_K_M	2.0	Too big
2× RTX 3090 24GB	48 GB	Q4_K_M	2.0	Too big
2× RX 7900 XTX 24GB	48 GB	Q4_K_M	2.0	Too big
4× RTX 3060 12GB	48 GB	Q4_K_M	2.0	Too big

Estimates for single-stream decoding at Q4_K_M; real tokens/sec varies with prompt length, context, batch size, and runtime build. Prompt processing (prefill) is faster than the decode figures shown here.

Quantization options

How Qwen3-Coder 480B A35B Instruct (480B params) fits at each quantization level on AMD Instinct MI350X 288GB (288.0 GB usable).

Quant	Bits	VRAM	Quality	Fit
Q2_KBest for your GPU	2	187.2 GB	Low	S86
Q3_K_S	3	235.2 GB	Low	F0
NVFP4	4	268.8 GB	Medium	F0
Q4_K_M	4	292.8 GB	Medium	F0
Q5_K_M	5	345.6 GB	High	F0
Q6_K	6	393.6 GB	High	F0
Q8_0	8	513.6 GB	Very High	F0
F16	16	984.0 GB	Maximum	F0

Get started

Copy-paste commands to run Qwen3-Coder 480B A35B Instruct on your machine.

Run

lms load Qwen3-Coder-480B-A35B-Instruct && lms server start

Frequently asked questions

Can AMD Instinct MI350X 288GB run Qwen3-Coder 480B A35B Instruct?

Yes, AMD Instinct MI350X 288GB can run Qwen3-Coder 480B A35B Instruct with a A grade (Very compromised (needs ~33.6 GB host RAM)). Expected decode speed: 35.3 tok/s.

How much VRAM does Qwen3-Coder 480B A35B Instruct need?

Qwen3-Coder 480B A35B Instruct (480B parameters) requires approximately 325.4 GB of memory with Q4_K_M quantization.

What is the best quantization for Qwen3-Coder 480B A35B Instruct?

The recommended quantization for Qwen3-Coder 480B A35B Instruct is Q4_K_M, which balances quality and memory efficiency.

What speed will Qwen3-Coder 480B A35B Instruct run at on AMD Instinct MI350X 288GB?

On AMD Instinct MI350X 288GB, Qwen3-Coder 480B A35B Instruct achieves approximately 35.3 tokens per second decode speed with a time-to-first-token of 5482ms using Q4_K_M quantization.

Can AMD Instinct MI350X 288GB run Qwen3-Coder 480B A35B Instruct for coding?

For coding workloads, Qwen3-Coder 480B A35B Instruct on AMD Instinct MI350X 288GB receives a A grade with 35.3 tok/s and 4K context.

What context window can Qwen3-Coder 480B A35B Instruct use on AMD Instinct MI350X 288GB?

On AMD Instinct MI350X 288GB, Qwen3-Coder 480B A35B Instruct can safely use up to 4K tokens of context. The model's official context limit is 256K, but available memory constrains the safe maximum.

What should I upgrade first if Qwen3-Coder 480B A35B Instruct feels slow on AMD Instinct MI350X 288GB?

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.

See all results for AMD Instinct MI350X 288GB See all hardware for Qwen3-Coder 480B A35B Instruct

Embed this result▼

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

<iframe src="https://willitrunai.com/embed/qwen-3-coder-480b-a35b-on-instinct-mi350x-288gb" width="400" height="180" frameborder="0" style="border:none;border-radius:12px;overflow:hidden;" title="Will It Run AI — fit result"></iframe>

Preview: