"AI Industry's Spring Gala" is Coming! A Complete Guide: What Are the Highlights of Nvidia GTC This Year?

NVIDIA’s annual developer conference (GTC 2026), known as the “AI Spring Festival,” will kick off on Monday in San Jose, California, with CEO Jensen Huang delivering a keynote speech. The event will run from March 16 to 19.

According to the official website, Jensen Huang will give his keynote from 11:00 to 13:00 Pacific Time on March 16 (02:00 to 04:00 Beijing Time on March 17). He will appear at the SAP Center in San Jose to introduce NVIDIA’s plans for the coming year to developers, analysts, and the media.

The CEO, often seen wearing a leather jacket, is known for densely announcing new products and technological updates during his keynote speeches, and this event is expected to be no different.

Over the past few months, NVIDIA has engaged in a series of acquisitions and collaborations, covering chip companies, software firms, and more. It is almost certain that this conference will reveal how this AI chip giant plans to integrate these technologies and capabilities into its product ecosystem.

One of the most talked-about deals in the past year was NVIDIA’s non-exclusive agreement in December with AI chip startup Groq, granting them inference technology licensing. Under the agreement, Groq founders Jonathan Ross, President Sunny Madra, and other core team members will join NVIDIA to advance the licensed technology and scale its application.

Groq’s chips are called Language Processing Units (LPUs), designed specifically for AI inference tasks. The company claims that LPUs can run large language models and other AI models with efficiency up to 10 times higher than GPUs.

As the industry shifts focus from training AI models to inference, more companies are seeking chips that can run AI software at lower costs. NVIDIA has emphasized the efficiency of its GPUs, but integrating Groq’s technology into its own processors or launching dedicated inference chips will help address market concerns that customers might switch to more specialized processors instead of continuing to use NVIDIA GPUs.

On the other hand, NVIDIA may also unveil its long-rumored PC processors at GTC. There have been reports that the company is preparing to launch two chips, N1 and N1X, which will power Windows laptops.

These processors will be based on the ARM architecture, similar to Qualcomm’s products, but will be more focused on gaming scenarios. NVIDIA’s GPUs are highly popular in gaming, and these CPUs are expected to be well-received as well.

NVIDIA chips already power Nintendo Switch and Switch 2 consoles and have been used in other computers, so launching laptop CPUs is not surprising. Amid the company’s large-scale shift toward data center business, this move can also help maintain support from gamers.

However, these chips are unlikely to generate the same massive revenue as NVIDIA’s GPUs and networking products. In 2025, NVIDIA’s gaming revenue totaled $22.5 billion, while data center revenue reached $19.35 billion.

Additionally, Jensen Huang is expected to share more details about the upcoming Vera Rubin AI platform and the planned release of Vera Ultra in late 2027. He may also provide updates on the Feynman GPU, scheduled for 2028.

On the software front, reports suggest that NVIDIA may launch an AI agent platform. Named NemoClaw, this service will allow enterprises to deploy intelligent agents within their systems.

NVIDIA often showcases its software strength by releasing various open-source AI models, used in fields such as autonomous vehicle development. Given the current industry-wide focus on intelligent agents, expanding into software and launching AI agent services is a logical step.

The conference will also feature many other software-related announcements, especially related to physical AI and robotics, which are particularly noteworthy. Huang has previously stated that this could represent a market opportunity worth trillions of dollars.

CITIC Securities notes that with GTC 2026 approaching, NVIDIA’s chip product lineup is expected to expand further. In addition to the full set of six core chips for the Vera Rubin AI platform, the company may also reveal more details about the Rubin Ultra chip and cabinet designs, bringing innovations in data connectivity and power supply architecture. The deployment prospects for new products like orthogonal backplanes and CPO are also expected to improve.

The report also suggests that NVIDIA may announce LPUs for inference, expanding its inference portfolio alongside the CPX chips. Furthermore, NVIDIA might outline the future direction of the Feynman architecture upgrade, sharing insights and judgments on future computing infrastructure and the AI industry. The outlook is optimistic that GTC 2026 will further strengthen market confidence in the sustained growth and incremental value of the AI industry.

(Source: Cailian Press)