Nvidia unveiled Vera CPU at GTC conference, its first processor designed for Agent AI and reinforcement learning, featuring 88 proprietary cores with multi-task parallel processing support. Its high-bandwidth memory reduces power consumption, and multiple cloud service providers and hardware manufacturers have already planned integration. It is expected to enhance AI programming and streaming data processing performance, with shipments anticipated in the second half of the year.

Nvidia announced at its GTC conference the launch of the Space-1 space computing module, which integrates 2 Rubin GPUs and 1 Vera CPU, delivering AI inference computing power 25 times that of the H100, and supports on-orbit data processing. Founder Jensen Huang pointed out that thermal dissipation challenges in space remain to be solved. Initial partners include multiple space enterprises, but the launch timeline has not been determined.

Nvidia released the Groq 3 LPU chip on March 17, with shipments expected in Q3. The chip can be used in conjunction with Vera Rubin NVL72 to significantly enhance inference throughput. The company also showcased the next-generation architecture Kyber, expected to launch in 2027.

Nvidia CEO Jensen Huang announced at GTC 2026 that the Vera Rubin platform will enter full-scale production, integrating seven new chips designed to build a supercomputer specifically for agentic AI. Compared to the previous generation, the GPU count has been reduced to one-quarter while inference capabilities have improved significantly. Combined orders are projected to reach $1 trillion between 2025 and 2027. Multiple cloud service providers and system vendors will launch related products in the second half of the year.

Industry analyst Patrick Moorhead released an analysis report focusing on Nvidia's GTC 2026 keynote, examining whether it can showcase a collaboration roadmap for GPUs, accelerators, and CPUs. A successful presentation would signify a transformation, while a failure could benefit cloud providers in developing their own chips. The report confirms several facts and predicts that the keynote will introduce the NemoClaw platform and the Feynman architecture roadmap, while also highlighting risks such as Groq integration, energy constraints, and market share challenges.

The decentralized trading platform of the Sui ecosystem, Bluefin, launched Bluefin Liquidity Network (BLN) and Bluefin Whitelabel on March 9. The former supports other protocols to use its trading and liquidity infrastructure, while the latter allows partners to customize branded trading platforms. The first batch of partners includes Vera and t2000ai.