Nvidia CEO Jensen Huang recently shared a nostalgic reflection on the company’s origins during a conversation at Microsoft’s Windows and Surface event in San Francisco on October 7. He emphasized that Nvidia’s very foundation is intertwined with Microsoft’s operating system, stating, “Nvidia was founded because of Windows.” This connection traces back to the introduction of Windows 3.1 in 1992, which laid the groundwork for the GPUs that now drive much of the AI landscape.
Sitting alongside him, Microsoft CEO Satya Nadella interjected humorously, suggesting, “If you could just do a little transfer of market cap back to me,” a light-hearted acknowledgment of Nvidia’s soaring valuation. Huang, quick on his feet, responded, “I do that every day.” The dialogue, moderated by Sriram Krishnan, a former senior policy adviser for AI at the White House, followed Microsoft’s unveiling of the Surface Laptop Ultra and a suite of laptops powered by Nvidia’s new RTX Spark chip.
Windows 95 and DirectX gave Nvidia’s GPUs a way into the PC
Huang reminisced about the late 1992 era when Windows 3.1 was revolutionizing personal computing. He candidly admitted that he had not encountered a PC before, having primarily worked with workstations, file servers, and supercomputers designed for chip development. Alongside his co-founders, he envisioned a future where PCs could deliver 3D graphics for gaming, serve as workstations, and facilitate design work.
The launch of Windows 95 marked a pivotal moment, enabling the integration of GPUs into PCs. The introduction of DirectX, particularly its Direct3D API, allowed applications to harness the power of GPUs. Huang noted that with DirectX 8, Nvidia and Microsoft co-created the first programmable GPU featuring programmable shaders. This innovation eventually paved the way for CUDA, Nvidia’s parallel computing platform.
“If not for Windows, there would be no GeForce,” Huang asserted, adding that without GeForce, the emergence of CUDA would have been unlikely, and AI researchers like Andrew Ng might not have had the computational resources necessary for deep learning.
A four-year-old conversation between Huang and Nadella led to RTX Spark
Nadella characterized this narrative as “the history of computing in some sense,” commending Huang for being a steadfast advocate for the potential of accelerated computing. He recounted an early Azure collaboration aimed at scaling high-performance computing in the cloud, which ultimately led to the development of a supercomputer that Microsoft later provided to OpenAI. When Sam Altman inquired about Microsoft’s computational capabilities, Nadella’s response was a tentative, “yeah, maybe.”
This partnership eventually steered both companies back to the PC realm. Huang explained that discussions revolved around integrating major tools, from Blender to Adobe and Autodesk applications, with AI agents on a single machine. “By the way, took four years,” Nadella added, highlighting the lengthy journey to this point.
The culmination of these efforts is RTX Spark, a project Huang noted required over 4,000 engineering years and extensive testing across more than 1,200 applications. Laptops featuring RTX Spark from manufacturers such as Asus, Dell, HP, Lenovo, MSI, and Microsoft’s own Surface are now available for preorder, with shipments commencing on October 16.
As the session drew to a close, Huang reflected on the enduring relationship between Nvidia and Windows, stating, “Windows has been with me for 34 years, it looks like it’s going to be with me for 34 more.”
