The RTX 3090 24GB has about 8% more memory bandwidth (936 vs 864 GB/s). Batch-1 decode tracks bandwidth closely, so a model that fits both cards should decode roughly 8% faster on the RTX 3090 24GB. The L40S 48GB fits more of the catalog at Q4/8K — 39 vs 33 of 61 models (48 vs 24 GB VRAM). Largest model that fits the L40S 48GB but not the RTX 3090 24GB: Qwen3-Coder-Next (79.7B). MSRP is $8,000 for the L40S 48GB vs $1,499 for the RTX 3090 24GB. 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 (47 shared models), the RTX 3090 24GB posts the higher tok/s in 37 rows vs 10. Methodology.