Nvidia Backs doubleAI to Introduce First Code Execution Model
Nvidia is funding doubleAI, a startup launching a code execution model called Maxwell to optimize software performance and reduce enterprise compute costs.
Nvidia Backs doubleAI to Introduce First Code Execution Model
Nvidia is funding doubleAI, a startup launching a code execution model called Maxwell to optimize software performance and reduce enterprise compute costs.
Introduction to the Code Execution Model
Nvidia has taken the unusual step of bankrolling a startup designed to make its own hardware less necessary, participating in a $220 million funding round for Tel Aviv-based doubleAI according to forkast.news. The firm is introducing what it calls the first code execution model, a development intended to tackle the rising costs of infrastructure in enterprise artificial intelligence adoption.
As infrastructure bottlenecks persist and hardware rental prices climb, the industry pressure to optimize software performance has shifted from a luxury to a survival requirement. doubleAI is positioning itself as a tax on inefficiency within the artificial intelligence stack by charging customers a percentage of the compute costs they successfully save.
How Maxwell Operates
Named Maxwell, the new system acts as an agent of the machine rather than a traditional large language model that simply generates text or code. It optimizes software performance across graphics processing units and central processing units by rewriting running code for specific hardware and data environments in real time.
The company has deployed roughly half of its capital to secure approximately 1,000 Nvidia B300 chips to power this engine. By sitting between the orchestration layer and the inference layer, Maxwell functions as a dynamic compiler that attempts to resolve the bloat inherent in modern development workflows.
Performance and Business Model
Vendor-reported metrics suggest aggressive performance gains, though these figures remain unverified by third parties. Examples provided by the vendor include reducing the runtime of specific video clips and delivering a reported reduction in compute costs for early customers.
The business model aligns doubleAI's revenue directly with the bottom line of the customer by taking a share of the compute savings delivered. Investors such as Lightspeed, Bessemer Venture Partners, and Nvidia are betting that software layers managing this type of operational efficiency will become increasingly critical as the market scales.
About The Author
