NY Gadget
PhonesComputingAIGamingTVsAudioStreamingEV
Live
01EVFerrari's First-Ever Electric Car Sold Out Its Entire First-Year Run Despite the Backlash02EVWhy the Prius Is Suddenly Cool Again — The Numbers Behind Its Comeback03PhonesThe 'Flock Sock': How a 3D-Printed Camera Cover Became a Privacy Protest Symbol04PhonesiPhone 17 Pro Camera Deep Dive: Is It Still Worth Buying With the 18 Pro Coming Soon?05ComputingWhy Falcon 9 Launches Are Happening Almost Weekly Now — Inside SpaceX's Pace in 202606EVTesla's 6-Seat Model Y L Finally Launches in the US at $61,990 — What You Get07ComputingCopilot+ PCs One Year Later: Are AI Laptops Actually Worth the Premium08ComputingFoldable Laptops Are Having a Moment — Are They Actually Practical Yet09ComputingAI Mini PCs Are Getting Cheaper Even as Laptops Get Pricier — Here's Why10ComputingBuilding a PC in 2026: Why Every Component Costs More Than Last Year01EVFerrari's First-Ever Electric Car Sold Out Its Entire First-Year Run Despite the Backlash02EVWhy the Prius Is Suddenly Cool Again — The Numbers Behind Its Comeback03PhonesThe 'Flock Sock': How a 3D-Printed Camera Cover Became a Privacy Protest Symbol04PhonesiPhone 17 Pro Camera Deep Dive: Is It Still Worth Buying With the 18 Pro Coming Soon?05ComputingWhy Falcon 9 Launches Are Happening Almost Weekly Now — Inside SpaceX's Pace in 202606EVTesla's 6-Seat Model Y L Finally Launches in the US at $61,990 — What You Get07ComputingCopilot+ PCs One Year Later: Are AI Laptops Actually Worth the Premium08ComputingFoldable Laptops Are Having a Moment — Are They Actually Practical Yet09ComputingAI Mini PCs Are Getting Cheaper Even as Laptops Get Pricier — Here's Why10ComputingBuilding a PC in 2026: Why Every Component Costs More Than Last Year

Computing/Components/Nvidia

AI Coding Agents Recreate Nvidia's CUDA Software in 10 Hours

Nvidia's competitive advantage is being challenged as AI startup Infinity recreates CUDA-like software in a record time

By Jordan Blake · Updated August 16, 2026 at 5:01 PM

AI Coding Agents Recreate Nvidia's CUDA Software in 10 Hours
AI Coding Agents Recreate Nvidia's CUDA Software in 10 Hours

The recent development of AI coding agents has led to a significant breakthrough in the tech industry. An AI startup, Infinity, has successfully recreated Nvidia's CUDA-like software in just 10 hours. This achievement has raised questions about the chipmaker's AI moat and its competitive advantage in the market.

CUDA, short for Compute Unified Device Architecture, is a software platform developed by Nvidia that allows AI developers to build their models with Nvidia's hardware. It took years to develop and includes ready-made code for common AI tasks, tools to find bugs, and software that lets thousands of chips work together to train models.

The ability of AI coding agents to generate code quickly has sparked a debate about the future of software development and the role of human developers. While some experts argue that the ability to generate code quickly is not the entire solution, it indicates that the software layer protecting Nvidia may no longer be as secure as it once was.

According to Jeremy Nixon, a former Google Brain researcher and the founder of Infinity, the recreation of CUDA-like software in 10 hours is evidence that one of Nvidia's biggest moats is being crossed. This development has significant implications for the tech industry, as it could potentially disrupt the dominance of Nvidia in the AI hardware market.

The use of AI coding agents to recreate CUDA-like software has also raised questions about the potential applications of this technology. For instance, it could be used to develop new software platforms that are compatible with different types of hardware, potentially reducing the dependence on Nvidia's proprietary technology.

The industry is at an important threshold, says Jeremy Nixon, a former Google Brain researcher and the founder of AI software startup Infinity. He told Business Insider his startup used AI coding agents to recreate CUDA-like software for the chip startup D-Matrix in 10 hours — evidence, he said, that one of Nvidia's biggest moats is being crossed.

The pressure on Nvidia does not only come from startups. Cloud giants like Google, Amazon, and Microsoft are also investing heavily in the development of AI technology, including the creation of their own software platforms. This could potentially challenge Nvidia's dominance in the AI hardware market and force the company to adapt to a changing landscape.

In conclusion, the recreation of Nvidia's CUDA-like software by AI coding agents in 10 hours is a significant development that has far-reaching implications for the tech industry. As the use of AI coding agents becomes more widespread, it is likely that we will see a shift in the way software is developed and the role of human developers in the process.

Introduction to CUDA

CUDA is a software platform developed by Nvidia that allows AI developers to build their models with Nvidia's hardware. It includes ready-made code for common AI tasks, tools to find bugs, and software that lets thousands of chips work together to train models.

The Role of AI Coding Agents

AI coding agents are software programs that use artificial intelligence to generate code. They can be used to develop new software platforms, recreate existing ones, and even debug code. The use of AI coding agents has the potential to revolutionize the way software is developed and could potentially disrupt the dominance of Nvidia in the AI hardware market.

Implications for the Tech Industry

The recreation of Nvidia's CUDA-like software by AI coding agents has significant implications for the tech industry. It could potentially challenge Nvidia's dominance in the AI hardware market and force the company to adapt to a changing landscape. Additionally, it could lead to the development of new software platforms that are compatible with different types of hardware, potentially reducing the dependence on Nvidia's proprietary technology.

Conclusion

In conclusion, the recreation of Nvidia's CUDA-like software by AI coding agents in 10 hours is a significant development that has far-reaching implications for the tech industry. As the use of AI coding agents becomes more widespread, it is likely that we will see a shift in the way software is developed and the role of human developers in the process.