NVIDIA GTC Conference Opens Tonight - Industry Chain Opportunities at a Glance

From March 16 to March 19 (Monday to Thursday), 2026, NVIDIA GTC will be grandly held in San Jose, California. NVIDIA CEO Jensen Huang announced that the event will showcase a “world-first” new chip.

Industry speculation suggests that this conference is expected to officially unveil core technical details of Rubin and the next-generation Feynman architecture GPUs, as well as launch a new inference chip integrated with LPU technology. LPU (Language Processing Unit) is a specialized computing architecture optimized for large model inference scenarios, emphasizing low latency, high throughput, and efficiency in sequential computations compared to traditional GPUs.

In recent years, the industry has begun to focus on the rapid growth in inference-side computing power demand, with some manufacturers improving inference efficiency through dedicated chips. NVIDIA’s previous acquisition of Groq is also seen as a strategic move to strengthen inference architecture capabilities. Groq’s LPU architecture, with deterministic execution, simplified scheduling, and high-bandwidth data paths, offers advantages in large model inference latency and efficiency.

If LPU architecture is scaled in AI servers, it will demand higher bandwidth and lower-loss interconnects, potentially leading to upgrades in server motherboard and accelerator card PCB layers, materials, and high-speed interconnect designs, thereby boosting the value of high-end PCBs and related industry chains.

Additionally, industry analysts expect that as GPU single-chip power reaches the kilowatt era, the infrastructure for computing power is undergoing disruptive upgrades. This GTC is expected to focus on technological breakthroughs and commercialization progress in CPO, power supplies, liquid cooling, and other areas.

Guosheng Securities pointed out that as NVIDIA GTC 2026 approaches, the AI computing power industry chain is entering a critical window of technological iteration. In the short term, hardware sectors related to computing power have already begun warming up, while in the long term, the technological pathways and commercialization pace of various infrastructure components will become clearer, providing a clear direction for industry development.

(Article source: Oriental Fortune Research Center)