Ryzen AI Max+ 395 128GB vs H200 141GB SXM5 for Local LLMs

The H200 141GB SXM5 has about 2133% more memory bandwidth (4,800 vs 215 GB/s). Batch-1 decode tracks bandwidth closely, so a model that fits both cards should decode roughly 2133% faster on the H200 141GB SXM5. The H200 141GB SXM5 fits more of the catalog at Q4/8K — 55 vs 51 of 69 models (141 vs 96 GB VRAM). Largest model that fits the H200 141GB SXM5 but not the Ryzen AI Max+ 395 128GB: Qwen3-235B-A22B-Instruct-2507 (235.1B). MSRP is $2,000 for the Ryzen AI Max+ 395 128GB vs $35,000 for the H200 141GB SXM5. MSRP is a launch list price, not live retail — Amazon prices move; click through for the current price. Not retail-shoppable: H200 141GB SXM5 (datacenter-class; cloud rental is the realistic way to use it). In the Q4/8K head-to-head below (63 shared models), the H200 141GB SXM5 posts the higher tok/s in 63 rows vs 0. Methodology.

Spec comparison

	Ryzen AI Max+ 395 128GB	H200 141GB SXM5
Vendor	AMD	NVIDIA
VRAM	96 GB	141 GB
Memory type	LPDDR5X	HBM3e
Memory bandwidth	215 GB/s	4,800 GB/s
TDP	120 W	700 W
MSRP (list)	$2,000	$35,000
Buy	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 63 shared models

Model	Params (B)	Ryzen AI Max+ 395 128GB tok/s	Ryzen AI Max+ 395 128GB fits	H200 141GB SXM5 tok/s	H200 141GB SXM5 fits	Winner
MiniCPM5-1B	1.08	126.7 est.	Yes	2828.3 est.	Yes	H200 141GB SXM5
Qwen3.5-2B	2.0	68.4 est.	Yes	1527.3 est.	Yes	H200 141GB SXM5
DeepSeek-R1-0528-Qwen3-8B-layer-mix-bpw-3.8-mlx	2.3	59.5 est.	Yes	1328.1 est.	Yes	H200 141GB SXM5
Qwen3.6-35B-A3B	36.0	45.6 est.	Yes	1018.2 est.	Yes	H200 141GB SXM5
Qwen3.5-35B-A3B	34.7	45.6 est.	Yes	1018.2 est.	Yes	H200 141GB SXM5
GLM-4.7-Flash	31.2	45.6 est.	Yes	1018.2 est.	Yes	H200 141GB SXM5
Qwen3-30B-A3B-Instruct-2507	30.5	45.6 est.	Yes	1018.2 est.	Yes	H200 141GB SXM5
Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16	30.0	45.6 est.	Yes	1018.2 est.	Yes	H200 141GB SXM5
Qwen3-Coder-Next	79.7	45.6 est.	Yes	1018.2 est.	Yes	H200 141GB SXM5
Ornith-1.0-35B	35.0	45.6 est.	Yes	1018.2 est.	Yes	H200 141GB SXM5
Qwen3-Next-80B-A3B-Instruct	81.3	45.6 est.	Yes	1018.2 est.	Yes	H200 141GB SXM5
qwen35b-a3b-fable-sft-abliterated	34.7	45.6 est.	Yes	1018.2 est.	Yes	H200 141GB SXM5
Llama-3.2-3B-Instruct	3.2	42.8 est.	Yes	954.5 est.	Yes	H200 141GB SXM5
gpt-oss-20b	21.5	38.0 est.	Yes	848.5 est.	Yes	H200 141GB SXM5
gemma-4-26B-A4B-it	25.2	36.0 est.	Yes	803.8 est.	Yes	H200 141GB SXM5

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 128GB	H200 141GB SXM5
Catalog models that fit	51 of 69	55 of 69
Largest exclusive fit	—	Qwen3-235B-A22B-Instruct-2507 (235.1B)

Ryzen AI Max+ 395 128GB details → · H200 141GB SXM5 details →