Can Qwen3.5 397B A17B run on B100 192GB?

YES — With Q2_K

D40Poor

Estimated from fit model

Qwen3.5 397B A17B needs ~221.8 GB VRAM. B100 192GB has 192.0 GB. With Q2_K quantization, expect ~25 tok/s.

Runtime: OllamaCapacity: OffloadBandwidth: HighStack: BasicBottleneck: 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.

Qwen3.5 397B A17B at Q4_K_M needs 309.1 GB — too much for B100 192GB (192.0 GB). Runs at Q2_K (221.8 GB) with low quality.

Capabilities:

Select quantization to explore

Q4_K_M (Medium quality) — 309.1 GB, exceeds 192.0 GB available

309.1 GB required192.0 GB available

161% VRAM needed

117.1 GB over capacity — needs offload or smaller quantization

Fit status

Too heavy

Decode

10.8 tok/s

TTFT

17945 ms

Safe context

Memory

309.1 GB / 192.0 GB

Offload

40%

Memory breakdown

Weights242.2 GB

KV Cache46.5 GB

Runtime1.2 GB

Headroom19.2 GB

See how fast it feels

With memory offload — actual speed may be lower

See how fast it feelsQwen3.5 397B A17B on B100 192GB
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: 10.8 tok/s decode · 17.9s TTFT (warm) · 27 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 20.8 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	F	Too heavy	12.3 tok/s	8603 ms	4K
Coding	F	Too heavy	10.8 tok/s	17945 ms	4K
Agentic Coding	F	Too heavy	8.6 tok/s	32896 ms	4K
Reasoning	F	Too heavy	10.8 tok/s	21207 ms	4K
RAG	F	Too heavy	8.6 tok/s	41120 ms	4K

Quantization options

How Qwen3.5 397B A17B (397B params) fits at each quantization level on B100 192GB (192.0 GB usable).

Quant	Bits	VRAM	Quality	Fit
Q2_K	2	154.8 GB	Low	F0
Q3_K_S	3	194.5 GB	Low	F0
NVFP4	4	222.3 GB	Medium	F0
Q4_K_M	4	242.2 GB	Medium	F0
Q5_K_M	5	285.8 GB	High	F0
Q6_K	6	325.5 GB	High	F0
Q8_0	8	424.8 GB	Very High	F0
F16	16	813.8 GB	Maximum	F0

Get started

Copy-paste commands to run Qwen3.5 397B A17B on your machine.

Run

lms load hf-unsloth--qwen3-5-397b-a17b-gguf && lms server start

Upgrade-Optionen

Hardware, die Qwen3.5 397B A17B gut ausführt

AMD Instinct MI350X 288GBBestes Preis-Leistungs-Verhältnis

288 GB VRAM (+96)

Makes the model fit on the accelerator instead of staying completely out of reach.14.6 tok/s Dekodierung

Makes the model fit on the accelerator instead of staying completely out of reach.

Raises estimated decode speed by about 35%.

ca. $8,000 MSRP

Frequently asked questions

Can B100 192GB run Qwen3.5 397B A17B?

Yes, B100 192GB can run Qwen3.5 397B A17B at Q2_K quantization (Very compromised (needs ~20.8 GB host RAM)). The recommended Q4_K_M requires 309.1 GB which exceeds available memory, but at Q2_K it needs only 221.8 GB. Expected decode speed: 24.8 tok/s.

How much VRAM does Qwen3.5 397B A17B need?

Qwen3.5 397B A17B (397B parameters) requires approximately 309.1 GB at Q4_K_M quantization. On B100 192GB, it fits at Q2_K using 221.8 GB.

What is the best quantization for Qwen3.5 397B A17B?

The recommended quantization is Q4_K_M, but on B100 192GB the best fitting quantization is Q2_K, which uses 221.8 GB.

What speed will Qwen3.5 397B A17B run at on B100 192GB?

On B100 192GB, Qwen3.5 397B A17B achieves approximately 24.8 tokens per second decode speed with a time-to-first-token of 7804ms using Q2_K quantization.

Can B100 192GB run Qwen3.5 397B A17B for coding?

For coding workloads, Qwen3.5 397B A17B on B100 192GB receives a F grade with 10.8 tok/s and 4K context.

What context window can Qwen3.5 397B A17B use on B100 192GB?

On B100 192GB, Qwen3.5 397B A17B can safely use up to 6K tokens of context at Q2_K quantization. The model's official context limit is —, but available memory constrains the safe maximum.

What should I upgrade first if Qwen3.5 397B A17B feels slow on B100 192GB?

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 B100 192GB See all hardware for Qwen3.5 397B A17B

Embed this result▼

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

<iframe src="https://willitrunai.com/embed/hf-unsloth--qwen3-5-397b-a17b-gguf-on-b100-192gb" width="400" height="180" frameborder="0" style="border:none;border-radius:12px;overflow:hidden;" title="Will It Run AI — fit result"></iframe>

Preview: