RTX A6000 48GB vs Ryzen AI Max 390 64GB for Local LLMs

The RTX A6000 48GB has about 257% more memory bandwidth (768 vs 215 GB/s). Batch-1 decode tracks bandwidth closely, so a model that fits both cards should decode roughly 257% faster on the RTX A6000 48GB. Both fit the same number of catalog models at Q4/8K (41 of 69; 48 vs 48 GB VRAM). MSRP is $4,500 for the RTX A6000 48GB vs $1,300 for the Ryzen AI Max 390 64GB. MSRP is a launch list price, not live retail — Amazon prices move; click through for the current price. In the Q4/8K head-to-head below (57 shared models), the RTX A6000 48GB posts the higher tok/s in 57 rows vs 0. Methodology.

Spec comparison

	RTX A6000 48GB	Ryzen AI Max 390 64GB
Vendor	NVIDIA	AMD
VRAM	48 GB	48 GB
Memory type	GDDR6	LPDDR5X
Memory bandwidth	768 GB/s	215 GB/s
TDP	300 W	120 W
MSRP (list)	$4,500	$1,300
Buy	Search Amazon (affiliate)	Search Amazon (affiliate)

MSRP is the launch list price, not live retail. Buy links are Amazon affiliate links — the price after you click is the live price.

Q4 / 8K head-to-head — top 15 of 57 shared models

Model	Params (B)	RTX A6000 48GB tok/s	RTX A6000 48GB fits	Ryzen AI Max 390 64GB tok/s	Ryzen AI Max 390 64GB fits	Winner
MiniCPM5-1B	1.08	452.5 est.	Yes	126.7 est.	Yes	RTX A6000 48GB
Qwen3.5-2B	2.0	244.4 est.	Yes	68.4 est.	Yes	RTX A6000 48GB
DeepSeek-R1-0528-Qwen3-8B-layer-mix-bpw-3.8-mlx	2.3	212.5 est.	Yes	59.5 est.	Yes	RTX A6000 48GB
Qwen3.6-35B-A3B	36.0	162.9 est.	Yes	45.6 est.	Yes	RTX A6000 48GB
Qwen3.5-35B-A3B	34.7	162.9 est.	Yes	45.6 est.	Yes	RTX A6000 48GB
GLM-4.7-Flash	31.2	162.9 est.	Yes	45.6 est.	Yes	RTX A6000 48GB
Qwen3-30B-A3B-Instruct-2507	30.5	162.9 est.	Yes	45.6 est.	Yes	RTX A6000 48GB
Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16	30.0	162.9 est.	Yes	45.6 est.	Yes	RTX A6000 48GB
Qwen3-Coder-Next	79.7	162.9 est.	Yes	45.6 est.	Yes	RTX A6000 48GB
Ornith-1.0-35B	35.0	162.9 est.	Yes	45.6 est.	Yes	RTX A6000 48GB
Qwen3-Next-80B-A3B-Instruct	81.3	162.9 est.	Yes	45.6 est.	Yes	RTX A6000 48GB
qwen35b-a3b-fable-sft-abliterated	34.7	162.9 est.	Yes	45.6 est.	Yes	RTX A6000 48GB
Llama-3.2-3B-Instruct	3.2	152.7 est.	Yes	42.8 est.	Yes	RTX A6000 48GB
gpt-oss-20b	21.5	135.8 est.	Yes	38.0 est.	Yes	RTX A6000 48GB
gpt-oss-120b	120.4	135.3 est.	No	37.9 est.	No	RTX A6000 48GB

est. = bandwidth-based estimate; measured = curated real benchmark that overrides the estimate. Fits = weights + KV(8K) + 1 GB overhead ≤ 95% of VRAM.

Fit coverage at Q4 / 8K

	RTX A6000 48GB	Ryzen AI Max 390 64GB
Catalog models that fit	41 of 69	41 of 69
Largest exclusive fit	—	—

RTX A6000 48GB details → · Ryzen AI Max 390 64GB details →