Question 1

Can NVIDIA GH200 96GB run Llama 3.2 3B Instruct?

Accepted Answer

Yes, NVIDIA GH200 96GB can run Llama 3.2 3B Instruct with a C grade (Runs well). Expected decode speed: 1360.0 tok/s.

Question 2

How much VRAM does Llama 3.2 3B Instruct need?

Accepted Answer

Llama 3.2 3B Instruct (3B parameters) requires approximately 13.8 GB of memory with Q5_K_M quantization.

Question 3

What is the best quantization for Llama 3.2 3B Instruct?

Accepted Answer

The recommended quantization for Llama 3.2 3B Instruct is Q5_K_M, which balances quality and memory efficiency.

Workload	Grade	Fit	Decode	TTFT	Context
Agentic Coding	C	Runs well	1360.0 tok/s	350 ms	221K
Chat	C	Runs well	1360.0 tok/s	350 ms	56K
Coding	C	Runs well	1360.0 tok/s	350 ms	112K
RAG	C	Runs well	1360.0 tok/s	350 ms	221K
Reasoning	C	Runs well	1360.0 tok/s	350 ms	112K

Quant	Bits	VRAM	Quality	Fit
Q2_K	2	1.2 GB	Low	D30
Q3_K_S	3	1.5 GB	Low	D30
NVFP4	4

Can NVIDIA GH200 96GB run Llama 3.2 3B Instruct?

Memory breakdown