RTX 4090D 24GB (China) vs RX 7900 XT 20GB for Local LLMs
The RTX 4090D 24GB (China) has about 26% more memory bandwidth (1,008 vs 800 GB/s). Batch-1 decode tracks bandwidth closely, so a model that fits both cards should decode roughly 26% faster on the RTX 4090D 24GB (China). The RTX 4090D 24GB (China) fits more of the catalog at Q4/8K — 33 vs 32 of 61 models (24 vs 20 GB VRAM). Largest model that fits the RTX 4090D 24GB (China) but not the RX 7900 XT 20GB: gemma-3-27b-it (27.4B). MSRP is $1,599 for the RTX 4090D 24GB (China) vs $899 for the RX 7900 XT 20GB. 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 (46 shared models), the RTX 4090D 24GB (China) posts the higher tok/s in 45 rows vs 1. Methodology.