At Nvidia's annual developer conference on Monday, CEO Jensen Huang took the stage to a packed house and said he expects purchase orders between Blackwell and Vera Rubin to reach $1 trillion through 2027.
Last year, the company had projections for a $500 billion revenue opportunity between the two chip technologies. Following Nvidia's earnings report last month, Finance chief Colette Kress said the company expects growth this year to exceed what was included in that estimate.
Huang said demand is booming from startups and big companies alike. Nvidia shares rose about 2% on Monday.
"If they could just get more capacity, they could generate more tokens, their revenues would go up," Huang said at GTC in San Jose, California.
Nvidia's graphics processing units for artificial intelligence have turned the brand into a household name and the most valuable public company in the world, worth about $4.5 trillion. As mass AI adoption shifts from chatbots to agentic apps that spawn off other agents to accomplish tasks, the number of tokens being generated has exploded, creating even greater need for running inference at faster speeds.
The chipmaker said in February that year-over-year revenue this quarter will surge about 77% to roughly $78 billion. The company has reported 11 straight quarters of revenue growth above 55%.
Nvidia is scheduled to roll out Vera Rubin later this year. The system, which is made up of 1.3 million components, will deliver 10 times more performance per watt than its predecessor, Grace Blackwell, the company claims. That's a significant development when energy consumption is one of the most critical issues facing the AI build-out.
Also on Monday, Huang unveiled the Nvidia Groq 3 Language Processing Unit, or LPU, the company's first chip from the startup that it mostly acquired through a $20 billion asset purchase in December, its largest deal ever. It's expected to ship in the third quarter.
Groq was founded by the creators of Google's in-house tensor processing unit, which has gained traction in recent years as a competitor to Nvidia's graphics processing units. The Groq 3 LPU is built to enhance its technology, with one core optimized for speeding up the GPU.
Huang introduced a full rack dedicated to housing the new Groq accelerators.
The Groq 3 LPX rack will hold 256 LPUs, and is meant to sit beside the Vera Rubin rack-scale system that's shipping to customers later this year. Huang said the Groq LPX rack can increase the tokens per watt performance of its Rubin GPUs by 35 times.
"We united, unified two processors of extreme differences, one for high throughput, one for low latency. It still doesn't change the fact that we need a lot of memory," Huang said. "And so we're just going to add a whole bunch of Groq chips, which expands the amount of memory it has."
Huang also showed off a prototype of Kyber, Nvidia's next big rack architecture leap after Rubin. It will integrate 144 GPUs in compute trays that sit vertically instead of horizontally in order to boost density and lower latency. The Kyber design will be available in Vera Rubin Ultra, Nvidia's next rack-scale system, expected to ship in 2027.
Roughly two hours into his keynote, Huang turned to the phenomenon of OpenClaw, which was launched in January by Austrian software developer Peter Steinberger. It's surged in popularity, due in part to attention on social media, as consumers and businesses swarm to products that can autonomously complete tasks, make decisions, and take actions on behalf of users without constant human guidance.
Steinberger joined OpenAI last month, and CEO Sam Altman said OpenClaw will "live in a foundation as an open source project that OpenAI will continue to support."
Huang highlighted a new developer toolkit to help people build and experiment with what's possible in new AI realms, using Nvidia hardware. He introduced a so-called reference stack named NemoClaw, specifically for OpenClaw, helping to make it "enterprise ready."
"It finds OpenClaw, it downloads it. It builds you an AI agent," Huang said.
In automotive, Huang gave details on a previously announced partnership with Uber, announcing the ride-hail service will launch a fleet powered by Nvidia's Drive AV software across 28 cities in four continents by 2028, starting with Los Angeles and San Francisco next year.
Huang announced that Nissan, BYD, Geely, Isuzu and Hyundai are building level 4 autonomous vehicles on Nvidia's Drive Hyperion program. Isuzu and China's Tier IV are also building autonomous buses using the platform, with help from Nvidia's AGX Thor robotic system chip.
— CNBC's Jordan Novet contributed to this report.
Facts Only
Nvidia CEO Jensen Huang announced at the company's annual developer conference that purchase orders for Blackwell and Vera Rubin chip technologies could reach $1 trillion by 2027.
Last year, Nvidia projected a $500 billion revenue opportunity for these technologies.
Nvidia reported 11 consecutive quarters of revenue growth above 55%, with this quarter's revenue expected to surge 77% year-over-year to $78 billion.
Nvidia unveiled Vera Rubin, a new system with 1.3 million components, claiming 10 times better performance per watt than its predecessor, Grace Blackwell.
Vera Rubin is scheduled to ship to customers later this year.
Nvidia introduced the Groq 3 Language Processing Unit (LPU), acquired in a $20 billion deal, expected to ship in the third quarter.
The Groq 3 LPU is designed to enhance GPU performance, with one core optimized for speeding up the GPU.
Huang showcased Kyber, a next-generation rack architecture integrating 144 GPUs vertically, expected in Vera Rubin Ultra by 2027.
OpenClaw, an open-source AI agent framework, was highlighted, with Nvidia releasing a developer toolkit called NemoClaw to support its enterprise adoption.
Uber will launch a fleet powered by Nvidia's Drive AV software across 28 cities in four continents by 2028, starting with Los Angeles and San Francisco next year.
Nissan, BYD, Geely, Isuzu, and Hyundai are building level 4 autonomous vehicles using Nvidia's Drive Hyperion program.
Isuzu and China's Tier IV are building autonomous buses using Nvidia's AGX Thor robotic system chip.
Nvidia's market valuation is approximately $4.5 trillion.
Executive Summary
Nvidia CEO Jensen Huang announced at the company's annual developer conference that purchase orders for its Blackwell and Vera Rubin chip technologies could reach $1 trillion by 2027, doubling last year's $500 billion projection. The company reported 11 consecutive quarters of revenue growth exceeding 55%, with this quarter's revenue expected to surge 77% year-over-year to $78 billion. Nvidia unveiled Vera Rubin, a new system with 1.3 million components promising 10 times better performance per watt than its predecessor, Grace Blackwell, addressing energy consumption concerns in AI infrastructure. Additionally, Nvidia introduced the Groq 3 Language Processing Unit (LPU), acquired in a $20 billion deal, which will ship in Q3 and is designed to enhance GPU performance. Huang also showcased Kyber, a next-generation rack architecture expected in 2027, and highlighted partnerships with Uber, Nissan, BYD, Geely, Isuzu, and Hyundai for autonomous vehicle development. The conference also featured OpenClaw, an open-source AI agent framework gaining traction, with Nvidia releasing a developer toolkit called NemoClaw to support its enterprise adoption. Nvidia's market valuation stands at approximately $4.5 trillion, reflecting its dominance in AI hardware.
The announcements underscore Nvidia's aggressive expansion in AI infrastructure, autonomous systems, and developer tools, positioning the company at the forefront of the AI revolution. However, the rapid growth and high demand also raise questions about supply chain constraints, energy efficiency, and the broader implications of AI-driven automation.
Full Take
**STEELMAN:** Nvidia's conference underscores its dominance in AI hardware, with bold projections ($1 trillion in orders by 2027) and rapid innovation (Vera Rubin, Groq 3 LPU, Kyber). The company is addressing critical challenges like energy efficiency and supply constraints while expanding into autonomous vehicles and AI agent frameworks. The partnerships with major automakers and the open-source OpenClaw project signal Nvidia's ambition to shape the next phase of AI adoption—agentic systems that operate autonomously. The financial performance and market valuation reflect real demand, not just hype.
**PATTERN SCAN:** The narrative leans heavily on authority (Nvidia's market position, Huang's statements) and future projections that may be difficult to verify. The framing of "booming demand" and "exploding tokens" could subtly pressure investors and competitors, though no outright distortion is present. The focus on energy efficiency (Vera Rubin's performance per watt) is a strategic counter to criticisms of AI's environmental impact, potentially preempting skepticism.
**ROOT CAUSE:** The paradigm here is the accelerationist push for AI adoption, where hardware advancements are positioned as inevitable and universally beneficial. The unstated assumption is that faster, more efficient AI infrastructure will solve problems like energy consumption and automation risks, without addressing broader societal trade-offs. This echoes historical tech cycles where infrastructure providers drive adoption by framing their solutions as indispensable.
**IMPLICATIONS:** For human agency, the rise of agentic AI (OpenClaw) raises questions about autonomy—who controls these systems, and how? The benefits accrue to Nvidia and its partners (automakers, ride-hail services), while costs (job displacement, energy use) may be externalized. Second-order consequences include further centralization of AI power in a few corporations and potential regulatory backlash if deployment outpaces ethical frameworks.
**BRIDGE QUESTIONS:**
How might the energy efficiency claims of Vera Rubin hold up under real-world deployment at scale?
What safeguards are needed to ensure AI agents like OpenClaw align with human values rather than corporate or algorithmic priorities?
If Nvidia's projections fall short, what ripple effects could that have on the broader AI ecosystem?
**COUNTERSTRIKE SCAN:** A coordinated influence campaign would amplify Nvidia's projections as inevitable, downplay risks, and frame critics as anti-innovation. The actual content avoids outright exaggeration but does emphasize growth and partnerships without deep scrutiny of challenges. No structural alignment with a manipulative playbook is detected—this appears to be standard corporate messaging, albeit with high stakes.
Patterns detected: none
Sentinel — Human
The article shows strong signs of human authorship, with natural sentence variance, specific attributions, and a distinct journalistic voice. No significant indicators of synthetic generation were detected.