Normal Computing, a NYC-based company which specializes in probabilistic AI and thermodynamic computing, has raised $50 million in a funding round led by Samsung Catalyst Fund.
The round also saw participation from Galvanize, Brevan Howard Macro Venture Fund, and ArcTern Ventures, alongside existing investors Celesta Capital, Drive Capital, Eric Schmidt’s First Spark Ventures, and Micron Ventures.
The funding increased the total to $85 million.
The company plans to use the funds to grow its operations and improve development.
Normal was created to solve two growing problems: the increasing complexity of chip design and the rising energy needs of AI systems. The company builds AI tools for the semiconductor industry and is also developing new computing hardware, designed with its own software.
Normal works with more than half of the top ten semiconductor companies by revenue and has built strong trust in the industry by helping create real production chips. Its platform, Normal EDA, is designed to accelerate chip design and develop new architectures that can greatly improve the cost and efficiency of AI scaling.
Read More:PDW Raises Over $110M in Series B Funding
AI is changing many types of knowledge work, like software development, data analysis, and graphic design. However, tools in industries like semiconductors have not improved much over the years. Now, the industry faces challenges such as talent shortages, low success rates in first-chip designs, high redesign costs, and growing energy and performance demands.
Normal aims to solve these issues by speeding up chip development, reducing timelines from years to months while still maintaining the high accuracy and teamwork needed in hardware design.
Normal EDA uses advanced AI that combines language models with formal logic to help engineers solve complex chip design problems. It understands a team’s goals by connecting with their workflow and uses that to design, improve, and verify chips while continuously learning.
The platform currently supports design and testing automation and is expanding to full end-to-end chip design. Normal also supports open industry standards and is a founding member of the Silicon Integration Initiative (Si2) LLM Benchmarking Coalition.
At the same time, Normal is working on its Carnot hardware program, using its own Normal EDA platform to design chips. In August 2025, the company completed CN101, the world’s first thermodynamic computing chip, designed for advanced AI tasks such as image and video generation. This is the first step toward improving energy efficiency by up to 1000x, and reducing delays.
Unlike traditional GPUs that consume extra energy to control system randomness, Normal’s chips leverage natural physical processes to compute more efficiently, creating a new approach to handling AI workloads.
“Datacenters may face an energy limit by 2030, and simply adding more power is not a long-term solution,” said Faris Sbahi, CEO of Normal Computing. “To meet growing demands, we need a new type of technology and better design methods. Normal EDA helps bring custom chips to market twice as fast today and aims to improve efficiency by up to 1000x in the future.”
“Our industry is changing rapidly because of AI, and advanced chip design has become much more complex,” said Dede Goldschmidt of Samsung Catalyst Fund. “Normal has a strong team with experience in both AI and semiconductors. Their platform can help customers design chips faster, and we are excited to support them and lead this funding round.”
“Most of the advanced computing landscape is oriented toward the next incremental improvement. ARIA exists to fund ideas with the potential to have an impact that is not marginal but transformational, even when the technical risk is high. Normal’s team has taken a fundamentally unconventional approach and delivered working silicon in CN101. That is an exceptionally rare outcome for work this ambitious, and we are excited to witness this next phase of the journey,” said Dr. Suraj Bramhavar, Programme Director of ARIA’s Scaling Compute programme.
About Normal Computing
Founded in 2022, by Faris Sbahi, Matthias Tan, and Antonio Martinez, Normal builds advanced AI and physics-based technologies to accelerate hardware innovation with top semiconductor companies. The company is working on new ways to design and build chips more efficiently. Its team includes experts from companies such as Meta, Apple, NVIDIA, and Google, as well as experienced founders and researchers. Normal is based in New York, with offices in San Francisco, London, and Copenhagen, and plans to expand to Korea.
Read More:Thesis Care Raises $45M in Series A Funding
