Question 1

Can Intel Arc A380 6GB run Llama 3.2 1B Instruct?

Accepted Answer

Yes, Intel Arc A380 6GB can run Llama 3.2 1B Instruct with a C grade (Runs well). Expected decode speed: 133.9 tok/s.

Question 2

How much VRAM does Llama 3.2 1B Instruct need?

Accepted Answer

Llama 3.2 1B Instruct (1B parameters) requires approximately 3.2 GB of memory with Q4_K_M quantization.

Question 3

What is the best quantization for Llama 3.2 1B Instruct?

Accepted Answer

The recommended quantization for Llama 3.2 1B Instruct is Q4_K_M, which balances quality and memory efficiency.

Workload	Grade	Fit	Decode	TTFT	Context
Agentic Coding	C	Runs well	91.1 tok/s	3090 ms	60K
Chat	C	Runs well	91.1 tok/s	1159 ms	15K
Coding	C	Runs well	133.9 tok/s	1446 ms	30K
RAG	C	Runs well	91.1 tok/s	3862 ms	60K
Reasoning	C	Runs well	91.1 tok/s	2510 ms	30K

Quant	Bits	VRAM	Quality	Fit
Q2_K	2	0.4 GB	Low	D31
Q3_K_S	3	0.5 GB	Low	D31
NVFP4	4

Can Intel Arc A380 6GB run Llama 3.2 1B Instruct?

Memory breakdown