NVIDIA's CEO Jensen Huang has signaled a highly positive outlook. In the early hours this morning Beijing time, the company's annual GTC conference, often referred to as the "AI Spring Gala," commenced. During his keynote address, Huang projected that the company's flagship chips would generate at least $1 trillion in revenue by the end of 2027. This announcement prompted a sharp surge in NVIDIA's stock price, which at one point climbed nearly 5% during the trading session, also contributing to a nearly 2% rise in the Nasdaq index.
Additionally, NVIDIA introduced the Space-1 Vera Rubin module, designed to deploy data-center-level AI computing capabilities to satellites and Orbital Data Centers (ODCs). The company emphasized its focus on in-orbit inference, real-time geospatial intelligence, and autonomous space missions.
Overnight, the three major U.S. stock indices all posted significant gains: the Dow Jones Industrial Average rose 0.83%, the Nasdaq Composite increased by 1.22%, and the S&P 500 climbed 1.01%. Large-cap technology stocks collectively strengthened, and the majority of chip stocks closed higher, with the Philadelphia Semiconductor Index advancing nearly 2%. Major European stock indices also closed broadly higher, with Germany's DAX 30 index finishing up 0.67%. International oil prices fell sharply, with WTI crude futures dropping 5.28% to settle at $93.50 per barrel, and Brent crude futures declining 2.84% to $100.21 per barrel.
Jensen Huang announced the $1 trillion revenue forecast on March 16, Eastern Time, during NVIDIA's annual GTC developer conference. He stated that the company's new-generation AI accelerator chip architecture, Blackwell, along with the next-generation Rubin products, are expected to create at least $1 trillion in revenue by the end of 2027. This figure significantly surpasses the $500 billion sales projection Huang provided in October 2025, once again highlighting the rapid expansion of the AI infrastructure investment wave.
Stimulated by this news, NVIDIA's stock experienced a rapid intraday surge, approaching a 5% gain before paring some of the increase to ultimately close 1.65% higher.
Amid the global AI wave, NVIDIA has become the core hardware supplier in the AI infrastructure surge. The market widely believes that global technology companies' capital expenditure on AI infrastructure could reach hundreds of billions of dollars over the next few years. NVIDIA's CFO, Colette Kress, previously stated at an event hosted by J.P. Morgan that, due to strong demand, the company is more optimistic about its data center business. She indicated that by the end of 2026, the expected revenue from NVIDIA's data center chips would "certainly" exceed the $500 billion forecast given in October last year.
Wall Street analysts suggest that NVIDIA's $1 trillion revenue target not only reflects the company's assessment of its own product demand but also indicates the rapid expansion of the entire AI infrastructure market. For capital markets, this revenue target alleviates investor concerns about a potential slowdown in AI demand and reinforces a key judgment: the AI computing power cycle is still in its early stages, not nearing its end. If this target is achieved, it implies that global tech companies will maintain high-intensity investment in AI servers, GPUs, and related systems for the coming years.
As AI models continue to scale, inference demand surges, and enterprise-level AI applications accelerate their deployment, AI computing power is gradually becoming a long-term capital expenditure direction, similar to cloud computing infrastructure.
On the product front, NVIDIA officially launched DLSS 5 at its annual GTC developer conference, describing it as the company's most significant breakthrough in computer graphics since real-time ray tracing was introduced in 2018. Utilizing a real-time neural rendering model, it injects pixels with "cinematic-grade" lighting and material details, aiming to achieve interactive visuals in games that approach Hollywood-level visual effects. In his speech, Huang compared DLSS 5 to the "GPT moment for graphics," emphasizing the new balance generative AI strikes between visual expression and artistic controllability.
According to NVIDIA, numerous top-tier publishers and game developers will integrate DLSS 5, including Bethesda, CAPCOM, NetEase, Tencent, Ubisoft, and Warner Bros. Games.
Furthermore, significant new developments at the event included two CPU products. The Vera CPU rack integrates 256 Vera CPUs per rack, offering double the computational efficiency and a 50% increase in operating speed compared to traditional CPUs. The Groq 3 LPX rack is equipped with 256 LPU processors, providing 128GB of on-chip SRAM and 640 TB/s of expanded bandwidth. When combined with the Vera Rubin platform, the LPX is projected to improve inference throughput per watt by 35 times. Huang announced that the LPU chips will be manufactured by Samsung, with rack shipments expected to begin in the second half of this year. All three rack types utilize liquid-cooling architecture.
The highly anticipated Spectrum-6 SPX, as expected, incorporated co-packaged optics (CPO) technology, delivering 5 times higher optical power efficiency and 10 times greater network reliability.
NVIDIA also launched the Space-1 Vera Rubin module, deploying data-center-level AI computing capability to satellites and orbital data centers. The company highlighted that its product portfolio—Jetson Orin, IGX Thor, RTX PRO 6000 Blackwell GPU, and the future Space-1 module—forms a complete computing architecture spanning orbital edge computing, ground-based AI data centers, and cloud analysis.
By venturing into new areas, NVIDIA is positioning AI agent infrastructure as a new growth vector. NemoClaw is positioned as the infrastructure layer for the OpenClaw agent platform, enabling AI agent deployment via "a single command" and integrating Nemotron models with the OpenShell runtime environment, while also addressing security, privacy, and sandboxing capabilities. The goal is not only effortless deployment but also secure operation.
NVIDIA emphasized that NemoClaw can run on RTX PCs, RTX PRO workstations, and devices like the DGX Station and DGX Spark, promoting the need for dedicated computing hardware for "always-on AI assistants." The company also announced a further expansion of its "open model ecosystem," covering three major AI domains: Agentic AI, Physical AI, and Medical AI.
Comments