doubleAI, the artificial intelligence company founded by Amnon Shashua, has raised $200 million in a Series A round co-led by Lightspeed Venture Partners and Bessemer Venture Partners, with participation from NVIDIA, Dell Technologies Capital, Greenfield, Pitango, and BRM. CTech reported the round values the Tel Aviv-based company at more than $1 billion.
Unlike coding assistants that help developers write software, doubleAI is building what it calls Code Execution Models (CEMs) — AI that operates “between the code and the silicon.” Its first model, Maxwell, continuously analyzes a running program, the hardware it executes on, and the data it processes, then optimizes the underlying machine instructions to fit those exact conditions. No source-code changes are required, and the company plans to charge customers based on the computing cost savings it delivers.
The benchmarks doubleAI cites are striking. Maxwell rewrote NVIDIA’s open-source cuGraph library, speeding up 55 percent of its algorithms by at least two times. It improved the pandas data toolkit’s performance on the PDS-H query benchmark by a factor of 38, and doubled the video generation speed of the LTX-2.5 world model. An early customer reportedly saw computing costs drop 67 percent.
The company plans to sell Maxwell to enterprises looking to cut AI workload costs on both CPUs and GPUs. The Wall Street Journal reported that doubleAI will spend roughly half the round on 1,000 NVIDIA B300 Blackwell GPUs. Compare this to developer tools like Cursor, Replit, and ChatGPT, which optimize how code gets written — Maxwell optimizes how it runs.
Also see our Gemini vs. Copilot comparison.
Source: SiliconANGLE; Ctech
