Nvidia has broadly deployed generative artificial intelligence tools across its engineering organization, rolling them out to an estimated 30,000 engineers. Through a partnership with San Francisco based Anysphere Inc., Nvidia is using a customized version of the Cursor integrated developer environment that focuses on artificial intelligence assisted code design. According to internal results, Nvidia’s engineers are now reportedly producing as much as three times the code compared to the previous development pipeline, meaning a growing share of the company’s products and services is being created with artificial intelligence guidance under human oversight.
The company’s adoption of generative artificial intelligence is occurring in the context of highly sensitive, mission critical products that cannot tolerate the typical error rates associated with artificial intelligence generated code. These products include GPU drivers that support everything from basic gaming to large scale artificial intelligence training and inference operations. To mitigate risk, Nvidia is likely enforcing strict internal guidelines for newly generated code and requiring an extensive range of tests before any changes reach production systems, reflecting the need for reliability in graphics and data center workloads.
Nvidia’s latest rollout builds on several years of artificial intelligence assisted workflows inside the company. Nvidia has already implemented a dedicated supercomputer that has been continuously enhancing DLSS (Deep Learning Super Sampling) for several years, using machine learning to improve image quality and performance in supported applications. In addition, some chip designs have been optimized using the company’s internal artificial intelligence tools, demonstrating that automated and assisted methods are influencing both software and hardware development across Nvidia’s product stack.
