As a pioneer in the AI wave, NVIDIA is achieving an exponential leap in productivity through its own technological closed loop. According to AIbase, NVIDIA has officially deployed generative AI coding tools on a large scale for up to 30,000 engineers within the company. This move means that this chip giant not only provides global AI computing power but has also entered a new phase of "AI-assisted design" itself.

For this initiative, NVIDIA has deeply collaborated with the San Francisco-based startup Anysphere Inc., integrating a customized Cursor integrated development environment (IDE). This tailored tool focuses on AI-assisted code design and optimization, providing precise support for NVIDIA's complex hardware development needs. According to the latest feedback, with the assistance of AI, the code output of NVIDIA engineers has reached three times that of previous development processes.

In fact, NVIDIA's exploration of AI-assisted workflows has been ongoing for a long time. From using supercomputers to optimize DLSS technology, to using internal AI tools to assist in chip design, NVIDIA has already proven the reliability of AI in critical tasks. AIbase learned that although code production has surged, the error rate of new code remains stable due to specialized optimization for enterprise scenarios. Currently, whether it's GPU drivers or large-scale training frameworks, a significant amount of code in NVIDIA's products may have been generated through "human-machine collaboration."