Infinity Ignition

Infinity Ignition is an autonomous AI research agent that generates, tests, debugs and optimizes the low-level compute kernels determining how efficiently a chip runs AI models, compressing inference software development from years into days. It is aimed at silicon vendors, cloud providers and infrastructure teams deploying models on non-NVIDIA accelerators.

Infinity Inc., founded in 2025 by former Google Brain researcher and AGI House creator Jeremy Nixon, builds the software layer that makes any AI chip inference-ready. Its product, Ignition, is an autonomous AI research agent that writes the low-level code required to run models on NVIDIA-alternative silicon — generating debuggers, profilers, compilers, kernels and the orchestration of kernels in inference — then tests, debugs and benchmarks the hardware against that code and automatically rewrites it to improve throughput. The system builds decompilers to translate executable code back into manipulable source, constructs bit-for-bit accuracy checks and memory detection tooling, and self-improves by recording successes and building reusable problem representations. Reported results include a 34% inference throughput increase for Qwen3-8B versus the vLLM framework achieved in a single day, 92% of theoretical peak performance on d-Matrix's Corsair accelerator reached in 10 hours, and full on-chip deployment of Qwen3, Qwen3.5 and Gemma4 within 10 days. Infinity charges no upfront license fees; it takes a percentage cut of the performance gains and cost savings it delivers, measured in tokens per second — roughly 20% of the customer's compute purchase savings from throughput improvement. d-Matrix Corp is a confirmed customer, with discussions underway with other major chip and cloud companies, and the company says it is already generating millions of dollars in annual recurring revenue from chip design partnerships. On 20 July 2026 Infinity announced a $15 million seed round at a $100 million post-money valuation led by Touring Capital with Principal Venture Partners, plus angel participation from executives and researchers at OpenAI, Anthropic and chipmakers. The company had 26 employees at the time of the raise and will use the funding to scale Ignition, expand engineering and accelerate chip partnerships.

Category

Infrastructure & Cloud

Pricing

Usage-based, Contact for pricing

Target Market

CTOs, Heads of AI Infrastructure, Silicon Vendors, Platform Engineering Leaders, ML Systems Engineers

Founded

2025

• ✓
  Autonomous kernel generation

  Ignition writes the low-level compute kernels that determine how efficiently a chip runs AI models, without manual engineering.

• ✓
  Automated test, debug and rewrite loop

  The agent measures hardware performance against generated code and automatically rewrites it when results fall short.

• ✓
  Full toolchain synthesis

  Generates debuggers, profilers, compilers, kernels and the orchestration of kernels used in inference.

• ✓
  Bit-for-bit correctness verification

  Builds decompilers and bit-for-bit accuracy checking plus memory detection tooling to validate generated code.

• ✓
  Self-improving research loop

  Records successful optimizations and constructs reusable problem representations that improve subsequent runs.

• ✓
  Outcome-based commercial model

  No upfront license fee — Infinity takes roughly 20% of the compute cost savings delivered, measured in tokens per second.

• •
  New accelerator bring-up

  Reached 92% of theoretical peak performance on d-Matrix's Corsair accelerator within 10 hours.

• •
  Throughput optimization on existing hardware

  Delivered a 34% inference throughput increase for Qwen3-8B versus the vLLM framework in a single day.

• •
  Rapid model porting to new silicon

  Deployed Qwen3, Qwen3.5 and Gemma4 fully on chip within 10 days.

Infinity charges no upfront license fees and instead takes a percentage cut of the performance gains and cost savings it delivers, reported at roughly 20% of the customer's compute purchase savings, measured in tokens per second.

This page was written from 2 sources, 2 on domains other than infinity.inc.

1. 1.techcrunch.com — inference startup infinity raises 15m from touring capital o
2. 2.siliconangle.com — infinity raises 15m run ai inference chipset