RTX 5090-Ready Models
These popular AI models can run on the RTX 5090. Actual requirements vary by model, engine, precision, and configuration.
textvision
Qwen3.8 27B
Qwen3.8 27BDense 27B vision-language model with hybrid Gated DeltaNet attention and 262K context
textvision
Gemma 4 31B IT
Gemma 4 31B ITGemma 4 31B dense vision-language model by Google with 256K context and thinking mode
text
Olmo 3.1 32B Instruct
Fully open dense 32B instruct language model by Ai2 with strong math, reasoning, and coding performance
RTX 5090 GPU Details
| Category | RTX 5090 |
|---|---|
| GPU Name | GB202 |
| Architecture | Blackwell 2.0 |
| Process Size | 5 nm |
| Transistors | 92,200 million |
| Release Date | Jan 30th, 2025 |
| Launch Price | 1,999 USD |
| Base Clock | 2017 MHz |
| Boost Clock | 2407 MHz |
| Memory Size | 32 GB |
| Memory Type | GDDR7 |
| Bandwidth | 1.79 TB/s |
| Tensor Cores | 680 |
| FP16 (half) | 104.8 TFLOPS (1:1) |
| FP32 (float) | 104.8 TFLOPS |