The conference call discusses NVIDIA's financial results for the first quarter of fiscal 2026, featuring remarks from the CEO and CFO. It highlights the company's performance, addresses forward-looking statements, and outlines the risks and uncertainties associated with future financial results. The call also mentions the availability of the webcast replay and provides guidance on non-GAAP financial measures.

Despite challenging operating conditions, NVIDIA delivered a strong quarter with a 69% year-over-year revenue increase to $44 billion. Data center revenue, particularly driven by AI factory build outs, surged 73% to $39 billion. The company faced a $4.5 billion charge due to new US export controls on its H20 data center GPU, impacting sales to the China market. NVIDIA's Blackwell GPU ramp contributed significantly to data center revenue growth, with major hyperscalers deploying thousands of units weekly. The company anticipates further growth in AI demand, driven by exponential increases in token generation by companies like OpenAI, Microsoft, and Google. NVIDIA is also witnessing a sharp increase in inference demand and developer engagements across various industries.

Introduced is a family of open reasoning models, Lama Nemo Tra, designed to boost AI platforms for enterprises, offering significant improvements in accuracy and inference speed. These models, available as Nims or Nvidia inference microservices, are being leveraged by major companies including Accenture, Deloitte, and Microsoft to optimize and scale AI applications. Notable enhancements include a 40% increase in model accuracy and a 10x improvement in response time. Partnerships with Yum Brands and cybersecurity firms further demonstrate the models' versatility and impact across industries. Additionally, advancements in networking, such as the NV Link, are enabling more efficient scaling of AI workloads.

NVIDIA has experienced significant growth in revenue, particularly in AI and gaming sectors, with strong adoption across major CSPs and consumer internet companies. The introduction of advanced silicon photonics switches and continued innovation in PC, creator, and gaming technologies highlight the company's commitment to pushing the boundaries of AI, robotics, and automotive advancements. Additionally, partnerships with major automotive and technology companies underscore the widespread integration and future potential of NVIDIA's technologies.

The company reported Q1 gross margins of 60.5% and 61% excluding a 4.5 billion charge, with non-GAAP gross margins at 71.3%. Operating expenses rose due to increased compensation and employee growth, alongside investments in infrastructure and AI solutions. Shareholder returns reached a record 14.3 billion in Q1. For Q2, total revenue is forecasted at 45 billion with modest growth across all platforms, expecting gross margins of 71.8% and 72% respectively. The company aims for mid-70s gross margins by the end of the year. Upcoming financial events include conferences in San Francisco, a virtual AI summit, and GTC Paris.

Export restrictions have significantly impacted the AI market, particularly in China, leading to a multibillion-dollar inventory write-off and driving innovation and talent towards rivals. The dialogue highlights the unintended consequences of export controls on American platforms, emphasizing the importance of considering global AI talent and market dynamics in policy-making.

The introduction of reasoning AI, like Deep Seqr, significantly increases inference demand due to its compute-intensive nature. This underscores the importance of open-source AI and reinforces American leadership in AI technology. Strategic investments in on-shore manufacturing, including partnerships for chip production and AI supercomputer construction, highlight a commitment to strengthening national security and creating jobs. Additionally, rescinding the AI diffusion rule promotes US AI tech with trusted partners, leading to historic investments in AI infrastructure projects globally. Countries are racing to build national AI platforms, recognizing AI as core to the next industrial revolution, driving a new growth engine for Nvidia.

The dialogue discusses the increasing demand for scaling up inference around reasoning models, highlighting the progress made over the past year and addressing the capacity to serve this demand. It also touches on the potential future needs for large-scale solutions in reasoning inference.

The Grace Blackwell NV Link 72 is highlighted as the optimal engine for reasoning AI due to its capability to generate significantly more tokens for complex problem-solving. It excels in tasks requiring deep thought, such as planning multiple paths to an answer, utilizing tools, and processing extensive data, thereby providing higher-quality responses at a much faster speed than previous models. Compared to Hopper, it boasts a 40 times higher speed and throughput, significantly reducing costs while enhancing response quality and service. This advancement necessitated a complete redesign of supercomputers, marking a major leap in AI technology.

The discussion highlights a significant $15 billion impact attributed to China, with ongoing headwinds affecting future quarters. Further, it explores the vast potential of AI spending, estimated to reach nearly a trillion dollars in the coming years, emphasizing AI's transformative role across industries and its infrastructure-like nature in producing digital intelligence essential for global markets. The conversation underscores the early stages of AI adoption and the milestone reached in reasoning and inference time scaling, positioning AI as a foundational technology for future development.

The dialogue highlights the integration of AI into various sectors, emphasizing the significant role of inference in compute workloads and the shift towards enterprise AI, particularly in the United States. Key points include the announcement of new AI servers and systems for on-premises use, the software-defined future of telco infrastructure, and the advent of AI-driven factories in manufacturing. The speaker also predicts the widespread adoption of AI in robotics and the automotive industry, marking the early stages of a comprehensive AI infrastructure build-out.

There has been a significant increase in large GPU cluster investments, with announcements from regions like Saudi Arabia, the UAE, Oracle, and X AI. The speaker highlights ongoing orders and the expansion of the supply chain to meet the growing demand for AI infrastructure, emphasizing that every country and industry will require this infrastructure. They predict numerous future announcements regarding AI factory developments, likening the current phase to the early stages of essential infrastructure buildouts similar to electricity and the internet.

The company discusses its Q2 guidance, noting an unexpected increase in orders due to AI diffusion rule lifting and strong demand post-export controls. The conversation highlights the growth seen in various customer segments and the increasing supply needs, suggesting a positive outlook for sequential growth throughout the year.

The dialogue highlights four significant positive surprises in the AI industry: the exponential increase in demand for reasoning AI, the rescinding of AI diffusion restrictions, the readiness of enterprise AI agents, and the emergence of industrial AI. These developments indicate a global awakening to AI's importance as infrastructure and the readiness of computing systems to integrate AI into enterprise and industrial settings.

The company is currently unable to ship its latest products to China due to stringent export controls, which significantly impact its revenue in the market. While there's no approval yet for shipping modified versions, the company is considering options to serve the Chinese market within the new limits, emphasizing the need to understand these restrictions fully. No announcements are planned until discussions with the administration occur.

The company discusses its networking platforms, emphasizing the scale-up platform nvlink and the Spectrum X for Ethernet. They highlight the complexity of scaling up and the adoption of Ethernet solutions by CSPs, along with trends in network attach rates.

By integrating features like low latency and congestion control, Spectrum X significantly boosts Ethernet efficiency in AI computing clusters, increasing utilization from 50% to up to 90%, resulting in substantial cost savings. Two major cloud service providers recently adopted Spectrum X, marking a major success.

Blue Field serves as the control plane for networks, enhancing storage, security, and enabling high-performance, multi-tenant clusters. It's ideal for achieving isolation among users while maintaining bare metal performance. Four networking platforms are highlighted, all showing growth and success.

A significant advancement marks the beginning of a powerful growth phase, with multiple growth engines in AI services generating revenue. AI is poised to outgrow previous technological shifts, powered by the capabilities to handle the full AI lifecycle, from model training to complex inference and reasoning at scale. Training demands are on the rise due to breakthroughs, while the need for compute for reasoning AI agents is exponentially increasing, setting the stage for scalable growth platforms.

Sovereign AI nations are significantly investing in AI infrastructure, paralleling the historical expansion of electricity and internet access. Major IT providers are integrating AI systems, such as those from RTX Pro, DGX, Nvidia Omniverse, and Isaac Group, into existing IT infrastructure both on-premises and in the cloud, modernizing the vast IT sector. This advancement marks the dawn of the AI age, encompassing AI infrastructure, enterprise AI, and industrial AI applications.

NVIDIA is set to host a significant event at GTC Paris, featuring a keynote on quantum GPU computing, robotic factories, and robots. The company will celebrate partnerships building AI factories across the region and embark on a tour through France, the UK, Germany, and Belgium. The event follows an earnings call, with the keynote scheduled for June 11 at Viva techcity.