During Nvidia's annual software developer conference, CEO Jensen Huang introduced the company's latest range of chips to affirm its leading position in the fast-changing artificial intelligence sector. The new products unveiled include:

1. Blackwell Ultra: Nvidia's upcoming GPU with expanded memory capacity to support larger AI models, expected to hit the market in the second half of this year. 2. Vera Rubin: Improved speed and data transfer capabilities compared to previous models, specifically designed for efficient operation in AI systems utilizing multiple chips. To be succeeded by Vera Rubin Ultra in 2027.

3. Feynman: The successor to the Vera Rubin system, utilizing the Feynman architecture, set for release in 2028.

4. DGX Personal AI Computers: Nvidia's new DGX AI computers powered by Blackwell Ultra chips, aimed at aiding developers in processing large models on desktops, to be manufactured by companies like Dell, Lenovo, and HP. This poses competition to high-end Macs by Apple.

5. Spectrum-X and Quantum-X Networking Chips: Nvidia's new silicon photonics networking chips to facilitate connection of numerous GPUs across different locations while significantly cutting down on energy consumption. Expected launch of Quantum-X Photonics chips later this year followed by Spectrum-X chips in 2026.

6. Dynamo Software: A software developed by Nvidia to accelerate the reasoning process for AI models by enabling them to think through an answer in multiple steps rather than providing a single response.

7. Nvidia Isaac GR00T N1: a foundational model for humanoid robots featuring a dual system for both quick and deliberate thinking, similar to reasoning models. The framework includes Newton, an open-source physics engine developed in collaboration with Google DeepMind and Disney Research, intended for robot creation.

The information was reported by Zaheer Kachwala and edited by Mohammed Safi Shamsi.