

NVIDIA GeForce RTX 5090 wins 7 of 9 benchmarks, averaging 419.1% faster.
Both cards were measured first-party on our bench, same suite, same test rig.
The gap is widest in Z-Image Turbo, where NVIDIA GeForce RTX 5090 leads by 694% (1.35 vs 10.725 images/min); the closest fight is Qwen3 4B (290% apart); VRAM decides part of this one: NVIDIA GeForce RTX 5090 runs 7 of our 12 AI workloads while the other card runs 6, models that don't fit score zero.
| Benchmark | NVIDIA GeForce RTX 4060 Ti 16GB | NVIDIA GeForce RTX 5090 | Difference |
|---|---|---|---|
| Qwen3 4B tok/s | 96.29 | 375.6 | -74% |
| Llama 3.1 8B tok/s | 57.09 | 268.14 | -79% |
| Qwen2.5-Coder 14B tok/s | 31.12 | 149.75 | -79% |
| Qwen3 32B tok/s | 0 | 71.15 | n/a |
| Llama 3.3 70B tok/s | 0 | 0 | n/a |
| Stable Diffusion XL images/min | 4.58 | 21.08 | -78% |
| Z-Image Turbo images/min | 1.35 | 10.725 | -87% |
| FLUX.1 dev images/min | 0 | 2.036 | n/a |
| Qwen-Image-Edit images/min | 0 | 0 | n/a |
How long each card takes to finish a complete pipeline, not just one model. NVIDIA GeForce RTX 4060 Ti 16GB is faster on 0 of 1; NVIDIA GeForce RTX 5090 on 1.
| Workflow | NVIDIA GeForce RTX 4060 Ti 16GB | NVIDIA GeForce RTX 5090 | Difference |
|---|---|---|---|
| Full codebase review | 32.1 min | 6.7 min | NVIDIA GeForce RTX 5090 4.81x faster |
| NVIDIA GeForce RTX 4060 Ti 16GB | NVIDIA GeForce RTX 5090 | |
|---|---|---|
| VRAM | 16GB | 32GB |
| Transistors | 22,900M | 92,200M |
| Die size | 187.8 mm² | 750 mm² |
| Process node | 4 nm | 4 nm |
| Transistor density | 121.9 M/mm² | 122.9 M/mm² |
| Architecture | Ada Lovelace | Blackwell (GB202) |
| Memory bandwidth | 288 GB/s | 1792 GB/s |
| Boost clock | 2,535 MHz | 2,407 MHz |
| TDP | 165 W | 575 W |
| Launch MSRP | $499 | $1,999 |
| Release | 2023-07-18 | 2025-01-30 |
NVIDIA GeForce RTX 4060 Ti 16GB full review · NVIDIA GeForce RTX 5090 full review · All AI & Machine Learning rankings