NVIDIA has once again set the stage for anticipation with its latest leak concerning the RTX 5090 graphics card. Today, a marketing image showcasing the unannounced next-gen GPU from NVIDIA has surfaced, courtesy of VideoCardz, confirming the buzz about the RTX 5090 sporting a hefty 32GB of GDDR7 memory.

The leak stems from an Inno3D RTX 5090 iChill X3 model’s box shot, which, while keeping most specs under wraps, hints strongly at the GPU coming with a substantial 3.5-slot cooler. This design choice points towards NVIDIA tackling the thermal challenges posed by the anticipated power and performance of this flagship GPU.

Delving into the specs, the RTX 5090 is rumored to outshine its predecessor, the RTX 4090, in several key areas. It’s said to boast 21,760 CUDA cores, promising a significant leap in computational power. The memory bandwidth is expected to hover around an impressive 1.8TB/s, facilitated by the new GDDR7 memory technology. Perhaps one of the most striking figures is its TDP, reportedly set at 575 watts, a 125-watt jump from the RTX 4090, underlining NVIDIA’s push for performance at potentially higher power costs.

This leak comes just in time as NVIDIA gears up for CES 2025, where they are slated to take the stage with a keynote tomorrow night.

According to rumors, there’s a timeline floating around where the RTX 5080 might make its debut on January 21st, suggesting a staggered release for NVIDIA’s RTX 50-series. This strategy could be NVIDIA’s attempt to manage market saturation and maintain consumer excitement over a longer period.

Related /

• NVIDIA teases RTX 50 series at GeForce LAN 50
• CES 2025 NVIDIA spotlight: RTX 5000 series, AI, and how to watch
• NVIDIA’s Jetson Orin Nano Super is here to democratize AI development

The RTX 5080, while not the flagship, is no slouch either, rumored to come with 16GB of GDDR7 memory, making it a formidable choice for gamers and professionals not requiring absolute top-tier performance but still wanting cutting-edge capabilities.

The shift to GDDR7 memory is significant, promising faster data transfer rates compared to previous generations, which could translate to smoother gameplay at higher resolutions and frame rates, alongside improved performance in professional applications like 3D rendering or video editing.

NVIDIA’s decision to increase the VRAM on the RTX 5090 to 32GB reflects the growing demands of modern games and applications which are increasingly memory-intensive. This move could also position NVIDIA to cater to the VR and AI sectors, where memory capacity plays a crucial role in performance.

NVIDIA has historically been a pace-setter in GPU technology, and with the RTX 5090, they seem poised to continue this tradition. However, with great power comes great responsibility, particularly in ensuring that these GPUs do not become energy hogs or too expensive for the average enthusiast.

While we await official confirmation and detailed specs at CES, these leaks paint a picture of what could be one of the most powerful consumer GPUs to date, potentially setting new benchmarks in graphics performance. The question now is how NVIDIA will balance this power with accessibility, both in terms of pricing and power consumption.

Stay tuned for more updates from CES, where NVIDIA will undoubtedly shed more light on the RTX 50-series, offering gamers and tech enthusiasts alike a glimpse into the future of graphical computing.