Autor: Jakub Rusinowski · Ostatnia aktualizacja: 10 lipca 2026
The RTX 5090 delivers 213 tokens/sec versus the 4090's 165 — a 29% speed boost with 33% more VRAM. Here's who should upgrade and who should wait.
NVIDIA's RTX 5090 landed in January 2025 with a $1,999 price tag and bold claims about AI performance. But if you already own an RTX 4090 — or are choosing between them on the second-hand market — is the upgrade actually worth it for running local LLMs? Let's look at the real numbers. | | RTX 4090 | RTX 5090 | |---|---|---| | VRAM | 24 GB GDDR6X | 32 GB GDDR7 | | Memory Bandwidth | 1,008 GB/s | 1,792 GB/s | | TDP | 450 W | 575 W | | Launch MSRP | $1,599 | $1,999 | | Architecture | Ada Lovelace | Blackwell | The single most important spec for LLM inference isn't CUDA cores or clock speed — it's…