NVIDIA is set to showcase the highly anticipated GeForce RTX 50 Series at CES 2025, powered by the innovative “Blackwell” architecture. The series will feature three key models: RTX 5090, RTX 5080, and RTX 5070, poised to revolutionize PC gaming and graphics processing.
GeForce RTX 5090: The Ultimate High-End Model
The RTX 5090 is the flagship of the series, boasting incredible specs that push the boundaries of modern gaming and AI processing.
- GPU: GB202-300-A1 core
- CUDA Cores: 21,760 cores
- VRAM: 32GB GDDR7
- Memory Bus Width: 512-bit
- Memory Speed: 28 Gbps
- Memory Bandwidth: 1,792 GB/s
- TBP (Total Board Power): 600W
With its 32GB of VRAM and an astounding 1,792 GB/s memory bandwidth, the RTX 5090 is designed to handle ultra-high-resolution gaming beyond 4K, as well as large-scale AI model processing with ease.
GeForce RTX 5080: Balanced High-End Performance
The RTX 5080 strikes a balance between high performance and cost efficiency, making it a top choice for gamers and creators.
- GPU: GB203-400-A1 core (full die)
- CUDA Cores: 10,752 cores
- VRAM: 16GB GDDR7
- Memory Bus Width: 256-bit
- Memory Speed: 32 Gbps (fastest in the industry)
- Memory Bandwidth: 1,024 GB/s
- TBP: 400W
The 16GB VRAM may be half that of the RTX 5090, but the RTX 5080 uses industry-leading 32Gbps memory modules, delivering a memory bandwidth of 1TB/s.
GeForce RTX 5070: The Mid-Range Powerhouse
Aimed at a wider audience, the RTX 5070 offers impressive performance for mid-range gaming.
- GPU: GB205 core
- VRAM: 12GB GDDR7
- Memory Bus Width: 192-bit
- Memory Speed: 28 Gbps
- Memory Bandwidth: 672 GB/s
- TBP: 250W
While the CUDA core count remains undisclosed, the RTX 5070 boasts a 33% higher memory bandwidth compared to its predecessor, the RTX 4070, ensuring a significant performance boost.
Key Technical Innovations
The GeForce RTX 50 Series brings several groundbreaking innovations, including:
- Enhanced PCB “back-drill” process for improved signal integrity between the GPU and GDDR7 VRAM.
- Introduction of 14-layer PCBs for more complex circuitry.
- Single 12V-2×6 power connectors for most models, simplifying power delivery.
- Full support for PCIe 5.0.
- DisplayPort 2.1a with up to 80 Gbps bandwidth for seamless 8K output.
These innovations ensure not only a performance leap but also enhanced reliability and expandability for future applications.
Conclusion
The NVIDIA GeForce RTX 50 Series represents a monumental leap forward in both gaming and graphics processing. With models like the RTX 5090, featuring an enormous 32GB of VRAM, and the RTX 5080, with its industry-leading memory speeds, this lineup promises something for every kind of user.
As NVIDIA integrates new AI technologies, the possibilities for gaming, rendering, and AI computation are limitless. Stay tuned for the official unveiling at CES 2025, where the future of PC gaming will be revealed.
NVIDIA GeForce RTX 5090 & 5080 GPU Specs Table
| Graphics Card | NVIDIA GeForce RTX 5090 | NVIDIA GeForce RTX 5080 | NVIDIA GeForce RTX 5070 |
|---|---|---|---|
| GPU Core | Blackwell GB202-300 | Blackwell GB203-400 | Blackwell GB205 |
| CUDA Cores | 21,760 (+33%) | 10,752 (+11%) | TBD |
| VRAM | 32 GB GDDR7 (+33%) | 16 GB GDDR7 (0%) | 12 GB GDDR7 |
| Memory Bus | 512-bit (+33%) | 256-bit (0%) | 192-bit (0%) |
| Memory Speed | 28 Gbps | 32 Gbps | 28 Gbps |
| Memory Bandwidth | 1,792 GB/s | 1,024 GB/s | 672 GB/s |
| Total Board Power | 600W (+33%) | 400W (+25%) | 250W (+14%) |
