Question 1

Can the NVIDIA GeForce RTX 4060 Ti 16GB run LLAMA4 Maverick 400B A17B?

Accepted Answer

LLAMA4 Maverick 400B A17B needs about 245.75GB working set (weights 245GB + KV cache 0.75GB at 4k context). The NVIDIA GeForce RTX 4060 Ti 16GB offers ~16GB usable — verdict: no. 245.8 GB vs 16 GB usable — does not fit

Question 2

How fast is LLAMA4 Maverick 400B A17B on the NVIDIA GeForce RTX 4060 Ti 16GB?

Accepted Answer

~0.65–1 tok/s on the NVIDIA GeForce RTX 4060 Ti 16GB. computed roofline: 288 GB/s × 0.55–0.88 efficiency window / 245.0GB (Q4_K_M) — bands, never points.

Question 3

What should I run LLAMA4 Maverick 400B A17B on instead?

Accepted Answer

ASUS ExpertCenter Pro ET900N G3 (GB300) runs it fully ($60000).

Can the NVIDIA GeForce RTX 4060 Ti 16GB run LLAMA4 Maverick 400B A17B?

Run LLAMA4 Maverick 400B A17B on instead