NVIDIA’s GeForce RTX 50-series desktop lineup brings Blackwell architecture with four models led by the RTX 5090 and its 32 GB of GDDR7 memory.
NVIDIA’s Blackwell generation has arrived in desktop form, and the Nvidia 5000 series specs confirm a serious generational jump. The four-model lineup — RTX 5090, RTX 5080, RTX 5070 Ti, and RTX 5070 — launched in early 2025 with GDDR7 memory, DLSS 4, and higher core counts across the board. All cards support PCIe 5.0 and use the 12V-2×6 power connector.
RTX 5090, 5080, 5070 Ti, and 5070: Complete Specifications
The four GeForce RTX 50-series desktop cards share Blackwell architecture and GDDR7 memory but differ in core counts, memory capacity, and power draw. The table below lays out the key specs across the stack; specs for the workstation-class RTX PRO 5000 are included separately where relevant.
| Model | CUDA Cores | Memory | Memory Bandwidth | TGP |
|---|---|---|---|---|
| RTX 5090 | 21,760 | 32 GB GDDR7 | 1,792 GB/s | 575 W |
| RTX 5080 | 10,752 | 16 GB GDDR7 | 960 GB/s | 360 W |
| RTX 5070 Ti | 8,960 | 16 GB GDDR7 | 896 GB/s | 300 W |
| RTX 5070 | 6,144 | 12 GB GDDR7 | 672 GB/s | 250 W |
For shoppers comparing the full lineup, our roundup of the best 5000 series GPUs breaks down real-world performance and value across the stack.
What Makes The Blackwell Architecture Different
All four cards use NVIDIA’s 4th-gen ray tracing cores and 5th-gen Tensor cores, enabling DLSS 4 with Multi-Frame Generation — which generates up to three additional frames per traditionally rendered frame. The updated NVENC and NVDEC encoder blocks improve video encoding and decoding throughput. These are desktop-only GPUs; no notebook variants are officially confirmed. The RTX PRO 5000 workstation card uses a different die and targets professional visualization, not gaming.
Power, Compatibility, and What You Need To Know
All cards use the 12V-2×6 power connector and support PCIe 5.0 x16, which is backward-compatible with PCIe 4.0 and 3.0 slots. GDDR7 memory delivers significantly higher bandwidth than GDDR6X.
As noted on the GeForce RTX 50 series Wikipedia page, the line marks the second consumer generation to use a multi-chip module design on the flagship card, alongside the architectural improvements in ray tracing and Tensor performance.
FAQs
When did the RTX 50-series launch?
The RTX 5090 and 5080 launched on January 30, 2025, followed by the RTX 5070 Ti and 5070 in February 2025. All four models are available in the U.S. market through NVIDIA’s partners.
Is the RTX 5090 worth the price jump?
At a reported $1,999, the RTX 5090 targets enthusiasts and professionals needing 32 GB of GDDR7 and 21,760 CUDA cores. For most gamers, the RTX 5080 or 5070 Ti deliver strong performance at roughly half the price.
Do I need a new power supply for an RTX 50-series card?
The RTX 5090 requires a 575 W TGP, so a high-wattage PSU with the 12V-2×6 connector is recommended. The RTX 5070 at 250 W is more accommodating, but all cards benefit from a quality unit with sufficient headroom.
References & Sources
- Wikipedia. “GeForce RTX 50 series.” Comprehensive specs table for RTX 5090, 5080, 5070 Ti, and 5070.
- NVIDIA. “NVIDIA RTX PRO 5000 Blackwell.” Official workstation card specs with 48 GB GDDR7 ECC.
Mo Maruf
I created WellFizz to bridge the gap between vague wellness advice and actionable solutions. My mission is simple: to decode the research and give you practical tools you can actually use.
Beyond the data, I am a passionate traveler. I believe that stepping away from the screen to explore new environments is essential for mental clarity and physical vitality.