# The Much-Awaited RTX 5000 Series: Specifications, Leaks, and the Elusive Price Tag
The technology community is alight with anticipation as Nvidia’s forthcoming RTX 5000 series is projected to make its spectacular unveiling at CES 2025. With Nvidia’s CEO Jensen Huang poised to deliver the keynote speech, excitement is reaching new heights. The RTX 5090, RTX 5080, and RTX 5070 are poised to take center stage, and although numerous technical specifications have been leaked, one vital detail remains notably missing: **the price**.
## A Peek into the Future: RTX 5090, 5080, and 5070
As per a report from [wccftech](https://wccftech.com/nvidia-geforce-rtx-5090-32-gb-rtx-5080-16-gb-rtx-5070-12-gb-ces-2025-launch/), the RTX 5000 series will be built on Nvidia’s cutting-edge Blackwell architecture, representing a major advancement in both performance and efficiency. Below is a summary of the leaked specifications for these three models:
### **RTX 5090: The Ultimate Powerhouse**
The RTX 5090 is anticipated to be a formidable performer, boasting the following rumored specs:
– **Graphics Chip**: GB202-300, showcasing 21,760 CUDA cores.
– **Memory**: 32 GB of GDDR7, operating at 28 Gbps.
– **Memory Interface**: 512-bit, delivering an impressive bandwidth of 1,792 GB/s.
– **Power Consumption**: 600 watts.
This card is crafted for those seeking the pinnacle of gaming and computational capabilities. The 5090 is forecasted to excel in 4K gaming and may even redefine the limits of 8K gaming.
### **RTX 5080: The Balanced Performance**
The RTX 5080 is projected to provide a more budget-friendly alternative while still delivering remarkable performance:
– **Graphics Chip**: GB203-400, equipped with 10,752 CUDA cores.
– **Memory**: 16 GB of GDDR7, functioning at an enhanced 32 Gbps.
– **Memory Interface**: 256-bit, offering a bandwidth of 1,024 GB/s.
– **Power Consumption**: 400 watts.
Interestingly, the 5080 will incorporate faster GDDR7 memory compared to the flagship 5090, despite having half the memory and a narrower memory interface.
### **RTX 5070: The Budget-Friendly Option**
The RTX 5070 is expected to emerge as the most economical choice in the range, although details are limited:
– **Graphics Chip**: Currently unknown.
– **Memory**: 12 GB of GDDR7, running at 28 Gbps.
– **Memory Interface**: 192-bit, yielding a bandwidth of 672 GB/s.
– **Power Consumption**: 250 watts.
Though the RTX 5070 may lack the raw might of its larger counterparts, it is likely to be a favored option for gamers seeking dependable performance at a more accessible price.
### **Overview of Key Specifications**
| | **RTX 5090** | **RTX 5080** | **RTX 5070** |
|———————-|—————————–|—————————–|—————————–|
| **Graphics Chip** | GB202-300, 21,760 CUDA cores| GB203-400, 10,752 CUDA cores | N/A |
| **Memory Size & Type**| 32 GB GDDR7 | 16 GB GDDR7 | 12 GB GDDR7 |
| **Memory Interface** | 512-bit, 28 Gbps | 256-bit, 32 Gbps | 192-bit, 28 Gbps |
| **Bandwidth** | 1,792 GB/s | 1,024 GB/s | 672 GB/s |
| **Power Consumption** | 600 watts | 400 watts | 250 watts |
## New Features and Innovations
### **12V-2×6 Power Connector**
It is anticipated that all three cards will utilize the 12V-2×6 power connector, which Nvidia has standardized for its in-house products, commencing with the RTX 40 Super series. This connector is engineered to meet the heightened power requirements of high-performance GPUs while preserving a compact design.
### **DisplayPort 2.1a**
Another thrilling development is the integration of DisplayPort 2.1a, delivering a complete 80 Gbps bandwidth. This will facilitate support for elevated resolutions and refresh rates, establishing the RTX 5000 series as ideal for gamers and content creators who seek top-tier performance.