Ryzen AI Max+ 395 64GB vs DGX Spark (GH200 480GB) for Local LLMs

The DGX Spark (GH200 480GB) has about 133% more memory bandwidth (500 vs 215 GB/s). Batch-1 decode tracks bandwidth closely, so a model that fits both cards should decode roughly 133% faster on the DGX Spark (GH200 480GB). The DGX Spark (GH200 480GB) fits more of the catalog at Q4/8K — 66 vs 41 of 69 models (480 vs 48 GB VRAM). Largest model that fits the DGX Spark (GH200 480GB) but not the Ryzen AI Max+ 395 64GB: DeepSeek-V4-Pro (861.6B). MSRP is $1,500 for the Ryzen AI Max+ 395 64GB vs $3,000 for the DGX Spark (GH200 480GB). 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 DGX Spark (GH200 480GB) posts the higher tok/s in 57 rows vs 0. Methodology.

Spec comparison

	Ryzen AI Max+ 395 64GB	DGX Spark (GH200 480GB)
Vendor	AMD	NVIDIA
VRAM	48 GB	480 GB
Memory type	LPDDR5X	LPDDR5X
Memory bandwidth	215 GB/s	500 GB/s
TDP	120 W	170 W
MSRP (list)	$1,500	$3,000
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)	Ryzen AI Max+ 395 64GB tok/s	Ryzen AI Max+ 395 64GB fits	DGX Spark (GH200 480GB) tok/s	DGX Spark (GH200 480GB) fits	Winner
MiniCPM5-1B	1.08	126.7 est.	Yes	294.6 est.	Yes	DGX Spark (GH200 480GB)
Qwen3.5-2B	2.0	68.4 est.	Yes	159.1 est.	Yes	DGX Spark (GH200 480GB)
DeepSeek-R1-0528-Qwen3-8B-layer-mix-bpw-3.8-mlx	2.3	59.5 est.	Yes	138.3 est.	Yes	DGX Spark (GH200 480GB)
Qwen3.6-35B-A3B	36.0	45.6 est.	Yes	106.1 est.	Yes	DGX Spark (GH200 480GB)
Qwen3.5-35B-A3B	34.7	45.6 est.	Yes	106.1 est.	Yes	DGX Spark (GH200 480GB)
GLM-4.7-Flash	31.2	45.6 est.	Yes	106.1 est.	Yes	DGX Spark (GH200 480GB)
Qwen3-30B-A3B-Instruct-2507	30.5	45.6 est.	Yes	106.1 est.	Yes	DGX Spark (GH200 480GB)
Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16	30.0	45.6 est.	Yes	106.1 est.	Yes	DGX Spark (GH200 480GB)
Qwen3-Coder-Next	79.7	45.6 est.	Yes	106.1 est.	Yes	DGX Spark (GH200 480GB)
Ornith-1.0-35B	35.0	45.6 est.	Yes	106.1 est.	Yes	DGX Spark (GH200 480GB)
Qwen3-Next-80B-A3B-Instruct	81.3	45.6 est.	Yes	106.1 est.	Yes	DGX Spark (GH200 480GB)
qwen35b-a3b-fable-sft-abliterated	34.7	45.6 est.	Yes	106.1 est.	Yes	DGX Spark (GH200 480GB)
Llama-3.2-3B-Instruct	3.2	42.8 est.	Yes	99.4 est.	Yes	DGX Spark (GH200 480GB)
gpt-oss-20b	21.5	38.0 est.	Yes	88.4 est.	Yes	DGX Spark (GH200 480GB)
gemma-4-26B-A4B-it	25.2	36.0 est.	Yes	83.7 est.	Yes	DGX Spark (GH200 480GB)

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

	Ryzen AI Max+ 395 64GB	DGX Spark (GH200 480GB)
Catalog models that fit	41 of 69	66 of 69
Largest exclusive fit	—	DeepSeek-V4-Pro (861.6B)

Ryzen AI Max+ 395 64GB details → · DGX Spark (GH200 480GB) details →