Nvidia has officially launched its next-generation GeForce RTX 50 series, featuring the new Blackwell architecture.
The series includes the RTX 5090, RTX 5080, RTX 5070 Ti, and RTX 5070, available in both desktop and mobile variants.
These GPUs are designed for high performance, with improvements across various specs, including faster memory, better ray tracing, and enhanced AI processing.
Key Features of RTX 50 Series:
- CUDA Cores: The RTX 5090 has 21,760 cores, offering 32% more than the RTX 4090.
- Tensor Cores: Enhanced to 3,352 TOPS in the RTX 5090, providing 154% higher AI performance than the RTX 4090.
- Ray Tracing: The RTX 5090 delivers 318 TFLOPS, 66% higher than its predecessor.
- Memory: All models feature faster GDDR7 memory, with the RTX 5090 offering 32GB on a 512-bit bus.
- Power Consumption: The RTX 5090 requires 575W, a significant increase from the RTX 4090.
- Price: The RTX 5090 is priced at $1999, while the RTX 5070 is priced at $549.
DLSS 4 Enhancements:
Multi-Frame Generation: DLSS 4 introduces the ability to generate and insert three AI-generated frames between each traditionally rendered frame, improving image quality and precision.
Improved DLSS Experience: Older RTX cards will also benefit from new features like enhanced single-frame generation and a switch to a transformer-based model for better image detail and stability.
Game Support: DLSS 4 will be available in 75 apps and games, including Unreal Engine 5, on launch day.
The new RTX 50 series will be available starting January 30 for the RTX 5090 and RTX 5080, while the RTX 5070 Ti and 5070 will launch in February.
Leave a comment