Infinity Secures $15M Funding to Challenge Nvidia in Chip Software Market

Infinity raised $15 million in fresh capital, aiming to disrupt the chip ecosystem with a universal kernel platform that could rival Nvidia’s dominance.

A New Kernel Frontier: The Birth of a CUDA Alternative

Infinity is building a software layer that mirrors Nvidia’s CUDA, but is designed to run on any processor architecture—SRAM, GPU, mobile SoCs, and Systolic Arrays alike. The goal is a single code base that developers can deploy across heterogeneous hardware.

■One code base, multiple chip families

■Continued use of popular languages like Python

■Automated re‑compilation driven by performance metrics

Funding Source and Investor Profile

The round attracted Touring Capital, Principal VC, and researchers from OpenAI and Anthropic, signaling not only capital but deep technical expertise and network access.

■$15 million raised

■Company valuation set at $100 million

■Investors bring strategic AI‑hardware insight

Customer Base and Strategic Partnerships

Infinity already counts D‑Matrix, a challenger to Nvidia, among its early adopters and is in talks with other major chip and cloud providers. The firm’s revenue model eschews upfront licensing fees, instead taking a share of the performance gains and cost savings realized by customers.

■D‑Matrix: emerging Nvidia competitor

■Performance‑based revenue share

■Development cycle cut from months to hours

Potential Market Ripple Effects

If Infinity’s universal kernel delivers the promised efficiency, it could erode Nvidia’s roughly 70 % market share in high‑performance compute and introduce a new competitive tier in the chip supply chain. Cloud providers and large‑scale data centers may benefit from lower costs and faster integration times.

■Nvidia’s dominance at risk

■Increased chip variety spurring price competition

■Faster integration thanks to standardized software stack

By decoupling software from specific hardware, Infinity could accelerate innovation cycles across the compute stack. Should the platform achieve its performance targets, both chip makers and cloud operators might see a 5‑10 % reduction in overall hardware spend, reshaping cost structures in high‑intensity sectors such as finance, scientific research, and media production.