Infinity AI Raises US$15 Million to Accelerate AI Inference Software for Next-Generation Semiconductor Chips

Infinity.inc (Infinity Artificial Intelligence Institute), an early-stage AI infrastructure research company building the software layer that makes any AI chip inference-ready, announced it has raised $15 million in seed funding at a $100 million post-money valuation. The round included significant participation from Touring Capital, along with Principal VC, executives at major chip companies, researchers from OpenAI and Anthropic, and other prominent angel investors. The funding will be used to scale Infinity's automated AI research platform and its autonomous AI agent, Ignition, which writes inference code for any new AI chip; expand Infinity's engineering team; and accelerate its work with chip partners, including d-Matrix. Infinity is already generating millions of dollars in ARR from its chip design partnerships, as it solves a key challenge for established and new silicon companies alike: creating the software that runs AI inference on their chips.

“The AI industry has operated under an artificial constraint that only a handful of chips could run AI well, because only NVIDIA spent decades building the software to make them work optimally. Ignition eliminates that constraint,” said Jeremy Nixon, founder and CEO of Infinity. “We believe the next era of AI will be defined not just by who makes the best chip, but by who can make any chip run state-of-the-art models at blazing speeds. For hardware providers, the difference between having that optimized software stack and not is the difference between mere potential and true performance. Our role is to ensure every partner reaches that potential.”

NVIDIA’s CUDA ecosystem has long anchored the AI accelerator market, with NVIDIA holding an estimated ~80% share of data-centre AI accelerators. While hardware performance from providers like AMD, Qualcomm, and AWS often rivals or exceeds industry standards, the challenge has historically been the time-intensive development of production-grade inference stacks. Infinity bridges this gap, allowing these providers to rapidly deploy their hardware at scale. Addressing this development cycle is a top priority across the semiconductor industry, as inference is on track to represent two-thirds of all AI compute spending in 2026.

Ignition: Any Chip. Any Model. In Days, Not Years.

Infinity's answer is Ignition, an AI research agent that automatically generates, tests and optimizes the low-level compute kernels that determine how efficiently a chip runs AI models. Ignition empowers human engineers to focus on high-level architecture rather than manual tuning by utilizing critical human steering and collaborative input throughout the optimization process. The system iterates continuously, with performance data feeding each successive generation of code. Infinity works shoulder to shoulder with partner engineering teams, treating their hardware and software stack as a strategic foundation that Ignition supercharges. Ignition's core features and benefits include:

• Autonomous Kernel Generation: Generates, writes, tests and optimizes compute kernels with human-led architectural steering, removing the core bottleneck to new chip deployment.
• Recursive Self-Improvement (RSI): A feedback loop in which real-world performance data trains the next generation of generated kernels, improving efficiency over time without manual tuning.
• Proven Performance:
  • Infinity improved inference throughput on a Qwen3-8B model by 34% more tokens/second in a single day of automated, large-scale library writing, producing an inference framework that outperformed the widely used vLLM framework.
  • In a design partnership with d-Matrix, Infinity’s AI agents reached up to 92% of a new chip's theoretical peak performance by running tensor-parallel matrix multiplications across all 32 of the Corsair's compute units, and only 10 hours after first accessing its hardware. Within 10 days, Infinity had three frontier models, Qwen3, Qwen3.5 and Gemma4, running end-to-end on the chip, with every layer written from scratch.
• Partnering for Cross-Architecture Reach: Successfully generalizes across proprietary instruction sets and memory architectures, empowering partners to unlock hardware potential even without prior inference ecosystems or representation in AI training data.
• Collaborative Incentive Alignment: Infinity’s business model is built on mutual success, capturing a share of the performance gains and cost savings it generates alongside chip partners, rather than charging upfront licensing fees.

“The approach to AI-driven model enablement has the potential to significantly shorten the time required to bring new AI compute architectures into production, helping accelerate deployment and time-to-first-revenue,” said Sid Sheth, founder and CEO of d-Matrix. “We’re excited about Infinity’s mission to help unlock the full potential of the next generation of AI compute.”

“Every major technology platform ultimately creates the most value when it becomes broadly accessible. Infinity is tackling one of the defining challenges of the AI era: making powerful AI affordable and available at scale,” said Songyee Yoon, founder and Managing Partner, Principal Venture Partners. “We are proud to back a team building the infrastructure needed to extend the benefits of this revolution to many more people.”

"As neural network architectures evolve at a rapid pace, AI hardware companies are caught in a constant cycle of keeping their software stacks current with state-of-the-art models and approaches," said Samir Kumar, General Partner at Touring Capital. "Infinity solves this fundamental software bottleneck. Its AI agents automate kernel development and build optimized inference libraries tailored to each target hardware platform, dramatically shortening the time it takes new silicon to achieve its intended peak performance. We're excited to support Jeremy and his team as they change the game in how chip companies maximize the real-world performance of their silicon."

Infinity was founded by Nixon, a former Google Brain researcher who co-founded AGI House, a San Francisco-based Artificial General Intelligence community and hacker network that has launched hundreds of startups and projects. Nixon is a widely recognized voice on the future of AI, with his commentary covered in The New York Times and Forbes.