NVIDIA CEO Jensen Huang delivered a sweeping two-and-a-half-hour keynote early Tuesday Beijing time, offering a comprehensive overview of the AI industry's hardware and software landscape. For capital markets, the event proved highly rewarding—anticipated investment themes largely materialized, and investors were also treated to Huang's striking new financial forecast for computing chip revenue. Key Buzzword: $1 Trillion During his presentation, Jensen Huang confirmed that NVIDIA's flagship chips are expected to help the company generate $1 trillion in revenue by 2027. The significance of this projection depends on individual investor interpretation. Previously, Huang had stated that data center equipment would bring in $500 billion in sales by the end of 2026. The latest forecast extends this outlook by one year and doubles the cumulative amount. This announcement marked the most exciting moment of the keynote for shareholders. NVIDIA's stock surged as much as 4% intraday before settling with a 1.6% gain at the close. GPU(X) AI Factory Platform(✓) NVIDIA emphasized that Vera Rubin is not a single chip but a complete AI supercomputing platform composed of 7 types of chips and 5 rack systems. Beyond the Rubin GPU and Vera CPU combination (Vera Rubin NVL72 GPU rack) familiar to investors, the keynote introduced two significant CPU products. The Vera CPU rack integrates 256 Vera CPUs per rack, delivering twice the computational efficiency and a 50% increase in operating speed compared to traditional CPUs. The Groq 3 LPX rack, equipped with 256 LPU processors, offers 128GB of on-chip SRAM and 640TB/s of expanded bandwidth. When combined with the Vera Rubin platform, the LPX rack is expected to improve inference throughput/power efficiency by 35 times. Huang noted that the LPU chips will be manufactured by Samsung, with rack shipments anticipated to begin in the second half of this year. All three rack systems utilize liquid cooling architecture. As widely anticipated by investors, the Spectrum-6 SPX adopted co-packaged optics (CPO) technology, delivering 5 times higher optical power efficiency and 10 times greater network reliability. For future products, Rubin Ultra will use a vertical insertion arrangement in the Kyber rack, enabling connections of up to 144 GPUs within a single NVLink domain. Next-generation Feynman architecture GPUs will incorporate stacked chips and custom HBM technology. Space Data Chip NVIDIA also introduced the Space-1 Vera Rubin module, designed to deploy data center-level AI computing capabilities on satellites and orbital data centers (ODCs), with a focus on in-orbit inference, real-time geospatial intelligence, and autonomous space missions. The company highlighted its comprehensive product portfolio—including Jetson Orin, IGX Thor, RTX PRO 6000 Blackwell GPU, and the upcoming Space-1 module—forming a complete computing architecture spanning orbital edge computing, ground-based AI data centers, and cloud analytics. One-Click "Shrimp Farming" By venturing into the "lobster industry," NVIDIA is positioning its AI agent infrastructure as a new growth vector. NemoClaw serves as the infrastructure layer for the OpenClaw agent platform, enabling deployment of AI agents via a "single command" and integrating Nemotron models with the OpenShell runtime environment, while enhancing security, privacy, and sandbox capabilities. The goal is not only simplified deployment but also "secure shrimp farming." NVIDIA emphasized that NemoClaw can operate on RTX PCs, RTX PRO workstations, and devices like DGX Station and DGX Spark, advancing the need for dedicated computing hardware for "always-on AI assistants." The company also announced an expansion of its "open model ecosystem," covering three major AI domains: Agentic AI, Physical AI, and Medical AI. DLSS 5: A "GPT Moment" for Graphics At GTC, NVIDIA unveiled DLSS 5, describing it as the most significant breakthrough in computer graphics since the introduction of real-time ray tracing in 2018. Huang stated, "Twenty-five years after NVIDIA invented the programmable shader, we are once again redefining computer graphics. DLSS 5 is the 'GPT moment' for the graphics field." The new DLSS 5 system integrates traditional 3D graphics data with generative AI models, which can predict and complete portions of an image, allowing NVIDIA GPUs to generate richly detailed scenes and highly realistic characters without rendering every element from scratch.
Comments