NVIDIA RTX 4060 Ti 8GB vs NVIDIA RTX 5090 for local LLMs
3 reports on the NVIDIA RTX 4060 Ti 8GB and 114 on the NVIDIA RTX 5090
Which is faster for local LLMs?
Nobody has reported the same model at the same quant on both cards yet, so there's no measured head-to-head. Memory bandwidth, which caps how fast a card can write, is 6.2x higher on the NVIDIA RTX 5090 (288 GB/s on the NVIDIA RTX 4060 Ti 8GB, 1,792 on the NVIDIA RTX 5090). The estimates below compare them on the models people run most.
- Memory
- 8GB
- Memory bandwidth
- 288 GB/s
- FP16 compute
- 88 TFLOPS
- Reports
- 3
- Memory
- 32GB
- Memory bandwidth
- 1,792 GB/s
- FP16 compute
- 419 TFLOPS
- Reports
- 114
Estimated by the calculator
The quant the calculator recommends for each card and its estimated speed at a 32,768-token context with 32 GB of system RAM, from memory bandwidth and the model's shape.
| Model | NVIDIA RTX 4060 Ti 8GB | NVIDIA RTX 5090 |
|---|---|---|
| Qwen3.8 27B | 1.6 t/sQ6_K, layers on CPU | 44.9 t/sQ6_K |
| Qwen3.8 125B · 6B active | doesn't fit at these settings | 20.3 t/sQ2_K, experts in RAM |
| DeepSeek V4 Flash 284B · 13B active | doesn't fit at these settings | doesn't fit at these settings |
| Qwen3.6 35B · 3B active | 14.7 t/sQ5_K_M, experts in RAM | 163.0 t/sQ5_K_M |
| Qwen3.6 27B | 1.6 t/sQ6_K, layers on CPU | 44.9 t/sQ6_K |
| Gemma 4 26B · 4B active | 9.4 t/sQ6_K, experts in RAM | 114.5 t/sQ6_K |
Change the context, RAM or model in the calculator: NVIDIA RTX 4060 Ti 8GB · NVIDIA RTX 5090
Every report on each card, with its source: NVIDIA RTX 4060 Ti 8GB · NVIDIA RTX 5090