Can GB200 NVL72 run Llama 3.2 3B?

Yes, the GB200 NVL72 has 13824GB VRAM and Llama 3.2 3B requires approximately 1.7GB VRAM for Q4 quantization. Estimated inference speed: 1675 tokens/second.

✅ — GB200 NVL72 Can Run Llama 3.2 3B

The GB200 NVL72 packs 13824GB VRAM. Llama 3.2 3B needs ~1.7GB. You're good to go.

✅ Plenty of headroom

Hardware Specs

GPU: GB200 NVL72
VRAM: 13824 GB
Bandwidth: 3350 GB/s
MSRP: $0

Model Requirements

Model: Llama 3.2 3B
Parameters: 3B
VRAM (Q4): ~1.7 GB
Context: 128,000

Estimated Performance

1675 tok/s

Estimated inference speed (Q4 quantization)

Run it locally with Ollama

ollama run llama3.2:3b

Requires Ollama installed on your machine

View on Ollama

Llama 3.2 3B Details GB200 NVL72 Specs

Try Llama 3.2 3B on a different GPU

Try another model on GB200 NVL72

Full compatibility matrix

✅ — GB200 NVL72 Can Run Llama 3.2 3B

Hardware Specs

Model Requirements

Estimated Performance

Run it locally with Ollama

Try Llama 3.2 3B on a different GPU

Try another model on GB200 NVL72

Related Checks

More tools

VRAM Calculator

Cloud GPU Pricing

GPU Comparison

Compatibility Lab

✅ — GB200 NVL72 Can Run Llama 3.2 3B

Hardware Specs

Model Requirements

Estimated Performance

Run it locally with Ollama

Try Llama 3.2 3B on a different GPU

Try another model on GB200 NVL72

Related Checks

More tools

VRAM Calculator

Cloud GPU Pricing

GPU Comparison

Compatibility Lab