At CES, NVIDIA's CEO Jensen Huang introduced the company's next-generation consumer gaming chip utilizing Blackwell technology. NVIDIA launched the new GeForce RTX 50 series GPUs, built on the Blackwell architecture. These GPUs boast 92 billion transistors, 4000 TOPS, and four AI processing units, delivering performance three times greater than the previous Ada generation.
The GeForce RTX 5090 graphics card is priced at $1999, while the RTX 5080 is priced at $999. Micron Technology provides memory for these new GPUs. The RTX 5090's performance is projected to be twice that of the 4090, marking a significant leap in the RTX 5070 to 5090 series capabilities. These cutting-edge GeForce GPUs will be available starting January.
During the presentation, it was highlighted that the amount of data humans will generate in the upcoming years is expected to surpass all data created throughout human history. The data is becoming multimodal, encompassing video, images, and sound. NVIDIA is leveraging AI feedback and synthetic data generation, enabling self-practice and iterative learning until the correct answers are achieved.
