← AI Hashrate 中文

Ryzen AI Max+ 395 64GB vs RTX 4090D 24GB (China) for Local LLMs

The RTX 4090D 24GB (China) has about 369% more memory bandwidth (1,008 vs 215 GB/s). Batch-1 decode tracks bandwidth closely, so a model that fits both cards should decode roughly 369% faster on the RTX 4090D 24GB (China). The Ryzen AI Max+ 395 64GB fits more of the catalog at Q4/8K — 41 vs 35 of 69 models (48 vs 24 GB VRAM). Largest model that fits the Ryzen AI Max+ 395 64GB but not the RTX 4090D 24GB (China): Qwen3-Next-80B-A3B-Instruct (81.3B). MSRP is $1,500 for the Ryzen AI Max+ 395 64GB vs $1,599 for the RTX 4090D 24GB (China). MSRP is a launch list price, not live retail — Amazon prices move; click through for the current price. Not retail-shoppable: RTX 4090D 24GB (China) (datacenter-class; cloud rental is the realistic way to use it). In the Q4/8K head-to-head below (51 shared models), the RTX 4090D 24GB (China) posts the higher tok/s in 51 rows vs 0. Methodology.

Spec comparison

Ryzen AI Max+ 395 64GBRTX 4090D 24GB (China)
VendorAMDNVIDIA
VRAM48 GB24 GB
Memory typeLPDDR5XGDDR6X
Memory bandwidth215 GB/s1,008 GB/s
TDP120 W425 W
MSRP (list)$1,500$1,599
BuySearch 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 51 shared models

ModelParams (B)Ryzen AI Max+ 395 64GB tok/sRyzen AI Max+ 395 64GB fitsRTX 4090D 24GB (China) tok/sRTX 4090D 24GB (China) fitsWinner
MiniCPM5-1B1.08126.7 est.Yes593.9 est.YesRTX 4090D 24GB (China)
Qwen3.5-2B2.068.4 est.Yes320.7 est.YesRTX 4090D 24GB (China)
DeepSeek-R1-0528-Qwen3-8B-layer-mix-bpw-3.8-mlx2.359.5 est.Yes278.9 est.YesRTX 4090D 24GB (China)
Qwen3.6-35B-A3B36.045.6 est.Yes213.8 est.YesRTX 4090D 24GB (China)
Qwen3.5-35B-A3B34.745.6 est.Yes213.8 est.YesRTX 4090D 24GB (China)
GLM-4.7-Flash31.245.6 est.Yes213.8 est.YesRTX 4090D 24GB (China)
Qwen3-30B-A3B-Instruct-250730.545.6 est.Yes213.8 est.YesRTX 4090D 24GB (China)
Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF1630.045.6 est.Yes213.8 est.YesRTX 4090D 24GB (China)
Ornith-1.0-35B35.045.6 est.Yes213.8 est.YesRTX 4090D 24GB (China)
qwen35b-a3b-fable-sft-abliterated34.745.6 est.Yes213.8 est.YesRTX 4090D 24GB (China)
Llama-3.2-3B-Instruct3.242.8 est.Yes200.5 est.YesRTX 4090D 24GB (China)
gpt-oss-20b21.538.0 est.Yes178.2 est.YesRTX 4090D 24GB (China)
Qwen3-Coder-Next79.745.6 est.Yes174.1 est.NoRTX 4090D 24GB (China)
gemma-4-26B-A4B-it25.236.0 est.Yes168.8 est.YesRTX 4090D 24GB (China)
Qwen3-Next-80B-A3B-Instruct81.345.6 est.Yes167.5 est.NoRTX 4090D 24GB (China)

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 64GBRTX 4090D 24GB (China)
Catalog models that fit41 of 6935 of 69
Largest exclusive fitQwen3-Next-80B-A3B-Instruct (81.3B)

Ryzen AI Max+ 395 64GB details → · RTX 4090D 24GB (China) details →