Can I Run Qwen 3.5 on 192 GB system RAM?

作者： Jakub Rusinowski · 最后更新： 2026年2月24日

Yes

Yes — Qwen 3.5 122B-A10B at Q8_0 needs about 133.3 GB of the 153.6 GB usable on 192 GB system RAM (~20.3 GB spare), at ~5.5 tok/s (estimated), with room for about 65,536 tokens of context.

Confidence: medium · Recommended quantization: Q8_0 · Estimated speed: ~5.5 tok/s

Buy This HardwareApple Mac Studio M2 Ultra — 192 GB VRAM · 60 W board power Deploy in the Cloud NowRunPod

or compare on Vast.ai

As an Amazon Associate we earn from qualifying purchases. Cloud GPU links are referral links — we may earn a commission at no extra cost to you.

192 GB system RAM — what it gives a model

Usable memory for models	153.6 GB
Memory bandwidth	90 GB/s

Qwen 3.5 on 192 GB system RAM: memory by quantization

Quant	Memory needed	Fits 153.6 GB?	Max context	Est. speed	Download
F16	247.7 GB	✗ No	—	—	244 GB
Q8_0	133.3 GB	✓ Yes	64K	~5.5 tok/s	129.6 GB
Q6_K	103.7 GB	✓ Yes	128K	~6.9 tok/s	100 GB
Q5_K_M	90.1 GB	✓ Yes	128K	~7.8 tok/s	86.5 GB
Q4_K_M	77.3 GB	✓ Yes	128K	~8.8 tok/s	73.7 GB
Q3_K_M	55.7 GB	✓ Yes	128K	~11.5 tok/s	52 GB
Q2_K	43.8 GB	✓ Yes	128K	~13.8 tok/s	40.1 GB

Which Qwen 3.5 sizes fit

Variant	Needs at Q4_K_M	Fits?	Est. speed
Qwen 3.5 397B-A17B	244.7 GB	✗ Too large	—
Qwen 3.5 122B-A10B	77.3 GB	✓ Fits	~8.8 tok/s
Qwen 3.5 35B-A3B	23.8 GB	✓ Fits	~23.2 tok/s
Qwen 3.5 27B	18.8 GB	✓ Fits	~3.9 tok/s
Qwen 3.5 9B	7.4 GB	✓ Fits	~10.9 tok/s
Qwen 3.5 4B	4.1 GB	✓ Fits	~22.3 tok/s
Qwen 3.5 2B	2.7 GB	✓ Fits	~38.9 tok/s
Qwen 3.5 0.8B	1.8 GB	✓ Fits	~74.5 tok/s

What to watch out for

1 larger variant of Qwen 3.5 does not fit and would need CPU offload or different hardware.
This model does not publish its full attention configuration, so the KV-cache share of these figures is inferred from its parameter count rather than computed exactly.
These figures assume CPU-only inference. Any discrete GPU, even an 8 GB one, will be several times faster for models that fit in its VRAM.

Recommended setup

llama.cpp (CPU build) or Ollama — both run without a GPU

How these numbers are calculated

Memory need = quantized weights + KV cache at 8,192 tokens (f16) + 0.8 GB runtime overhead.
153.6 GB of the 192 GB is treated as usable for model weights (80% — the rest is the OS and running applications).
DDR5-5600 dual channel at 89.6 GB/s peak. CPU decode is assumed to sustain 35% of that peak, because CPU inference is not purely bandwidth-bound — it also spends real time in compute and thread synchronisation. This figure is an assumption, not a fitted constant: no CPU measurement is in the calibration set.
CPU-only inference: no GPU is assumed. A GPU of any size will beat these figures substantially.
Throughput is a memory-bandwidth roofline estimate, not a measurement. It is labelled "estimated" everywhere it appears.
KV cache is inferred from the parameter count because this model does not publish its full attention configuration.

FAQ