TechFlow News, July 20: According to TechCrunch, AI infrastructure startup Infinity announced the completion of a $15 million funding round, with a valuation of $100 million. Investors include Touring Capital, Principal VC, and researchers from OpenAI and Anthropic. Infinity is committed to building a CUDA alternative, developing a universal inference software library capable of running across various hardware architectures including SRAM, GPU, mobile chips, and systolic arrays, to reduce developers' dependence on the NVIDIA ecosystem.
Its core product, the AI research agent Ignition, can automatically write low-level kernel code for non-NVIDIA chips, and possesses capabilities for self-testing, debugging, and continuous optimization, compressing the original development cycle of several months down to hours or days.
The company was founded by former Google Brain researcher and AGI House founder Jeremy Nixon, currently has 26 employees, has partnered with AI chip manufacturer D-Matrix, and is currently in negotiations with multiple large chip and cloud computing companies. Infinity adopts a business model based on revenue sharing according to performance gains, charging no upfront licensing fees.
