Head to head · October 2026
RTX 3090 vs Tesla P40: which is better for local AI?
24GB against 24GB, 936 against 346 GB/s. What each runs, how fast, and which one to buy.
The verdict
Both have 24GB of memory, so they run the same models. The RTX 3090 generates text about 230% faster thanks to 936 GB/s of memory bandwidth versus 346 GB/s. The Tesla P40 costs about $820 less ($230 used vs $1,050 used).
- Get the RTX 3090 if speed matters most.
- Get the Tesla P40 if you want the most for your money or you care about power draw, heat, and noise.
We may earn a commission when you buy through links on this page. It never changes what we recommend. How we make money
Performance
Model by model
Estimated single-user generation speed for popular models, at the best quantization each card can hold with 8K context.
| Model | RTX 3090 | Tesla P40 |
|---|---|---|
| Llama 3.1 8B Instruct | ~73 tok/sQ8_0 | ~22 tok/sQ8_0 |
| Qwen3 14B | ~41 tok/sQ8_0 | ~12 tok/sQ8_0 |
| gpt-oss-20b | ~120 tok/sMXFP4 | ~41 tok/sMXFP4 |
| Qwen3 30B-A3B (2507) | ~135 tok/sQ4_K_M | ~47 tok/sQ4_K_M |
| Gemma 3 27B | ~39 tok/sQ4_K_M | ~11 tok/sQ4_K_M |
| Qwen3 32B | ~41 tok/sQ3_K_M | ~12 tok/sQ3_K_M |
| Llama 3.3 70B Instruct | ~1.8 tok/sQ4_K_M Partial offload | ~1.6 tok/sQ4_K_M Partial offload |
| gpt-oss-120b | ~21 tok/sMXFP4 Partial offload | ~18 tok/sMXFP4 Partial offload |
Estimates from our bandwidth model; see methodology. Measured results, where we have them, are on each GPU's page.
Specifications
Spec by spec
| Spec | RTX 3090 | Tesla P40 |
|---|---|---|
| Memory | 24GB GDDR6X | 24GB GDDR5 |
| Bandwidth | 936 GB/s● | 346 GB/s |
| Power | 350W | 250W● |
| Price | ~$1,050 used | ~$230 used● |
| Price per GB | $44 | $10● |
| Architecture | Ampere | Pascal |
| Released | 2020 | 2016 |
| Software | CUDA, Vulkan | CUDA, Vulkan |
Our take
Bottom line on each
The default value pick for 24GB of CUDA memory and the backbone of most multi-GPU builds.
The cheapest 24GB card, if you can cool it and accept Pascal's slow compute.
Frequently asked questions
Is the RTX 3090 or Tesla P40 better for running LLMs?
Both have 24GB of memory, so they run the same models. The RTX 3090 generates text about 230% faster thanks to 936 GB/s of memory bandwidth versus 346 GB/s. The Tesla P40 costs about $820 less ($230 used vs $1,050 used).
How much faster is the RTX 3090 than the Tesla P40 for AI?
About 230% faster at generating text in our estimates, because token generation is limited by memory bandwidth (936 GB/s vs 346 GB/s). Prompt processing depends more on compute and can differ by more.
Keep comparing