NVIDIA has taken the stage at CES 2025, making waves with the launch of its highly anticipated GeForce RTX 50 Series Desktop and Laptop GPUs. At the forefront of this new lineup is the GeForce RTX 5090, touted as the fastest consumer GPU yet, boasting cutting-edge advancements powered by the NVIDIA Blackwell architecture.
During his keynote address, NVIDIA’s CEO, Jensen Huang, declared, “Blackwell, the engine of AI, has arrived for PC gamers, developers, and creatives.” This new architecture features fifth-generation Tensor Cores and fourth-generation RT Cores, enabling advancements like neural shaders and AI-driven rendering technologies to significantly boost graphic fidelity.
The GeForce RTX 5090 is set to retail at $1,999 and is packed with 92 billion transistors, which allows it to handle over 3,352 trillion AI operations per second (TOPS) of computing power. This marks up to double the performance of its predecessor, the RTX 4090, allowing gamers to experience cutting-edge graphics at unprecedented levels.
Alongside the RTX 5090, NVIDIA also unveiled the RTX 5080, priced at $999, the RTX 5070 Ti at $749, and the RTX 5070 at $549. This new pricing strategy has resulted in some price reductions compared to prior models, which may entice gamers attuned to budget constraints. For example, the lower-tier RTX 5070 displays performance similar to the RTX 4090, representing compelling affordability for gamers on tight budgets.
Building on the theme of affordability without compromising power, the new DLSS 4 technology promises to revolutionize graphics performance. This system introduces Multi Frame Generation, allowing AI to generate up to three frames for each rendered frame, leading to performance increases of up to eight times more than traditional rendering methods. DLSS 4 will debut with over 75 games and applications, offering gamers immediate and expansive benefits.
One of the standout developments showcased during the event was the RTX Neural Shaders. This latest advancement incorporates small AI networks directly within programmable shaders, unlocking cinema-quality materials and lighting effects for real-time gaming experiences. Huang noted, “There’s no getting around it; the company is now hyper-focused on AI, but it also plans on using the tech to completely revolutionize GeForce.”
NVIDIA also highlighted its focus on gaming realism by introducing RTX Neural Faces, enabling dynamic rendering of game characters using generative AI technology. This new feature aims to mitigate the common challenges associated with rendering lifelike human characters, setting the stage for more immersive gaming.
The enhancements do not stop there. The RTX 50 Series also marks the introduction of NVIDIA Reflex 2, which aims to cut down game latency significantly by synchronizing frame updates with the latest mouse input. Reflex 2 can potentially reduce latency by as much as 75%, providing players with the speed and accuracy needed for intense multiplayer competition.
While the RTX 5090 is undoubtedly the flagship model, discussions surrounding the entire RTX 50 Series inevitably lead to the performance thresholds set by the RTX 5090 and its lower-tier models. The RTX 5070, for example, is pitched as delivering RTX 4090-level performance at $549, offering substantial value as gamers choose to upgrade their systems.
According to Huang, NVIDIA is committed to delivering AI as the cornerstone of all its advancements: “AI is coming home to GeForce.” Continuing the AI theme, NVIDIA is rolling out its new NVIDIA ACE technologies, which reimagine autonomous game characters capable of perceiving, planning, and acting within the game similarly to human players. This framework aims to enrich gameplay experiences by adding layers of strategy and realism to game interactions.
When it came to the physical design of the new GPUs, attendees noted the sleek profile of the RTX 5090 compared to the bulkier design of the RTX 4090. The new model is reportedly thinner, measuring around two slots, which could be pivotal for those building compact gaming rigs.
The anticipated launch date for these GPUs is January 30, with NVIDIA already fielding interest for the RTX 5090 and its siblings, alongside laptop variants expected to hit the market shortly thereafter. This echoes NVIDIA's assertive position as they strive to maintain dominance amid the graphics card competition.
Despite some concerns raised about increased pricing — with the RTX 5090 costing $400 more than its RTX 4090 counterpart — NVIDIA's commitment to delivering beyond expectations seems unwavering, as it continues to lead the charge with innovative technologies.
Consumers and enthusiasts alike are eager to see how these technological advancements will influence the future of gaming, particularly with high-profile titles expected to integrate these developments. Whether the enhanced realism via AI rendering can sustain the hype remains to be seen, but one thing is clear: NVIDIA's GeForce RTX 50 Series has set the stage for innovation at CES 2025.