AI Startup Infinity Secures $15M to Challenge Nvidia's Software Dominance

Global Tech Market — Ekhbary News Agency

AI infrastructure startup Infinity recently announced a successful $15 million funding round, valuing the company at $100 million. The investment, revealed on Monday, drew capital from notable firms like Touring Capital and Principal VC, alongside key researchers from AI giants OpenAI and Anthropic. This substantial backing underscores the growing interest in Infinity's mission to simplify AI model execution across diverse hardware.

Software Innovation to Rival Dominance

Infinity's core endeavor involves developing software that enables AI chips to run complex models more efficiently. The company aims to create a universal inference library, effectively an alternative to Nvidia's proprietary CUDA software, which has long solidified its market position. This new kernel software, designed to be compatible with various chip architectures—from SRAM and GPUs to phone chips and Systolic Arrays—seeks to democratize access to high-performance AI processing.

Automated Invention and Strategic Growth

Founded last year by Jeremy Nixon, a former Google Brain researcher and creator of AGI House, Infinity emerged from Nixon's fascination with "automated invention." He posits that "AI systems can actually be a meta technology," a vision he previously explored with his Omega algorithm, which autonomously generated and evaluated new machine learning algorithms. Infinity’s AI research agent, Ignition, embodies this philosophy by writing, testing, and optimizing low-level code for non-Nvidia chips, adapting to proprietary designs for improved performance. The startup, currently employing 26 staff, counts AI chip maker D-Matrix among its clients and is actively discussing partnerships with other major chip and cloud companies. For what it's worth, their model charges based on performance gains, not upfront fees.