AI infrastructure company Infinity on Monday announced a $15 million raise at a $100 million valuation from investors including Touring Capital, Principal VC, and researchers at companies like OpenAI and Anthropic.

The startup is creating software to make it easier for AI chips to run AI models. A big reason why Nvidia became the top player is not only its high-performance chips, but also its CUDA (Compute Unified Device Architecture) software, which allows its GPUs (originally designed to run graphics) to act as general-purpose processing CPUs. The largest AI development frameworks, PyTorch and TensorFlow, have been built on top of CUDA. This allows developers to write their apps in popular languages ​​like Python, use leading AI frameworks, and their apps will, by default, run on Nvidia chips.

Most of these application-level startups would not have the resources or knowledge to write their own cores (the low-level software that operates the chips) and port their applications to other AI chips. Therefore, Infinity is trying to create a CUDA alternative kernel software that works with any type of chip, such as SRAM, GPU, phone chips, and Systolic Arrays. Infinity is part of a new wave of startups that are trying, product by product, to weaken Nvidia’s market dominance.

Infinity is attempting to build a universal inference library that runs on all chips, allowing these chips to automate the replication of next-generation research results.

Infinity was launched last year by Jeremy Nixon, once a Google Brain researcher and creator of the hacker networking community AGI House. Nixon told TechCrunch that he decided to launch this company because he was obsessed with the idea of ​​”automated invention”: the belief that “AI systems may actually be a metatechnology.” He himself had invented a machine learning algorithm called Omega, he said, which essentially created new machine learning algorithms and automatically evaluated them in a feedback loop.

That success got him thinking about other cases where this approach could work, and he turned to hardware, believing that automated systems could also generate the low-level code, like cores and such, needed to help run chips more effectively.

Ignition, Infinity’s AI research agent, is intended to write the low-level code needed for AI inference on alternative chips to Nvidia. Test, debug and measure how fast the hardware runs on the code and automatically rewrite the code if necessary to improve performance. The system is automatically optimized, meaning it continually learns and improves itself. It also accommodates different chip architectures, regardless of proprietary designs, Nixon says. The result is what Infinity claims is a CUDA-level software stack.

Its clients include AI chip maker (and would-be Nvidia challenger) D-Matrix, and Infinity is in talks with other big chip and cloud companies, Nixon said.

However, humans are in the loop, providing high-level direction while the agent does more of the tedious hard work. In one case study, the startup found that the agent works much faster than a single human, reducing what could have been a years or months process to hours or days. Infinity does not charge an upfront license fee; instead, a portion of performance gains and cost savings are needed by measuring changes in tokens per second.

Currently, Infinity has 26 employees, including design, operations and engineering.