Can I Run Bonsai 27B on RTX 5090 Desktop (32 GB VRAM, 64 GB RAM)?

Written by Jakub Rusinowski · Last updated July 15, 2026

Yes, but it is tight

Yes, but it is tight — 1-bit Bonsai 27B at Q8_0 needs about 31.2 GB of the 32 GB usable on RTX 5090 Desktop (32 GB VRAM, 64 GB RAM), leaving only ~0.8 GB before the runtime starts swapping. Expect ~41.3 tok/s (estimated), with room for about 8,192 tokens of context.

Confidence: low · Recommended quantization: Q8_0 · Estimated speed: ~41.3 tok/s

Get a personalized upgrade path →

Buy this hardware NVIDIA GeForce RTX 5090 32GB — 32 GB VRAM · 575 W board power Deploy in the cloud now NVIDIA A40 on RunPod — from $0.44/hr · rate checked 2026-08

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.

RTX 5090 Desktop (32 GB VRAM, 64 GB RAM) — what it gives a model

Usable memory for models	32 GB
Memory bandwidth	1792 GB/s
Form factor	Desktop
Operating system	Windows or Linux
Memory upgradeable	Yes

Bonsai 27B on RTX 5090 Desktop (32 GB VRAM, 64 GB RAM): memory by quantization

Quant	Memory needed	Fits 32 GB?	Max context	Est. speed	Download
F16	56.5 GB	✗ No	—	—	54 GB
Q8_0	31.2 GB	✓ Yes	8K	~41.3 tok/s	28.7 GB
Q6_K	24.7 GB	✓ Yes	32K	~51.6 tok/s	22.1 GB
Q5_K_M	21.7 GB	✓ Yes	32K	~58.4 tok/s	19.1 GB
Q4_K_M	18.8 GB	✓ Yes	64K	~66.6 tok/s	16.3 GB
Q3_K_M	14.1 GB	✓ Yes	64K	~87.3 tok/s	11.5 GB
Q2_K	11.4 GB	✓ Yes	64K	~105.3 tok/s	8.9 GB

Which Bonsai 27B sizes fit

Variant	Needs at Q4_K_M	Fits?	Est. speed
1-bit Bonsai 27B	18.8 GB	✓ Fits	~66.6 tok/s
Ternary Bonsai 27B	18.8 GB	✓ Fits	~66.6 tok/s

What to watch out for

Only ~0.8 GB of headroom at Q8_0: a longer context or a second application can push this into swapping.
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.

RTX 5090 desktop limitations

575 W board power — budget for a 1000 W+ PSU and the heat it puts into the room.
Models larger than 32 GB must offload to system RAM, which costs roughly an order of magnitude in speed.

Recommended setup

Ollama or llama.cpp (CUDA/ROCm) — vLLM if you need concurrent serving

How these numbers are calculated

Memory need = quantized weights + KV cache at 8,192 tokens (f16) + 0.8 GB runtime overhead.
32 GB of VRAM on the NVIDIA GeForce RTX 5090 at 1792 GB/s.
64 GB of system RAM available for CPU offload when a model exceeds VRAM.
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