Normal Computing Secures $50 Million Funding from Samsung Catalyst to Innovate Silicon Design
On March 25, 2026, Normal Computing announced a significant achievement: it secured $50 million in strategic funding led by the Samsung Catalyst Fund. With this latest round, the total funding received by Normal has now exceeded $85 million. Joining Samsung Catalyst in this funding effort are new investors such as Galvanize, Brevan Howard Macro Venture Fund, and ArcTern Ventures, along with existing investors including Celesta Capital, Drive Capital, Eric Schmidt's First Spark Ventures, and Micron Ventures.
Founded to confront two pressing issues—the growing complexity of silicon design and the escalating energy requirements associated with AI infrastructure—Normal Computing aims to revolutionize the semiconductor industry. The company specializes in developing AI for semiconductor design, creating a new class of computing hardware that leverages its proprietary software to design its own hardware intellectual property (IP).
Faris Sbahi, CEO of Normal Computing, expressed the urgency surrounding the energy crisis in data centers, which are projected to hit an 'energy wall' by 2030. He emphasized that merely seeking more energy isn't a sustainable solution. 'To meet increasing demands for intelligence per dollar and per watt, we need innovative architectural designs and a unified methodology in design practices. Our objective with Normal EDA is to expedite the path to market for custom silicon by doubling the speed today, with an ambitious goal of achieving 1000x efficiency gains through our platform in the future,' he stated.
The company's growing influence is notable, as it has partnered with over half of the top ten revenue-generating semiconductor firms. Normal Computing distinguishes itself by gaining trust within the industry to facilitate tangible production silicon. Their specialized AI platform for semiconductors, dubbed Normal EDA, aims to both drastically speed up chip design processes and develop architectures that can alter the economic landscape of AI scaling.
Dede Goldschmidt, Senior Vice President and Managing Director of Samsung Catalyst Fund, commented on the profound transformation occurring in the semiconductor industry influenced by AI. He remarked on the increased complexity surrounding leading-edge semiconductor chip design and commended Normal for assembling a skilled team with diverse experience across AI and semiconductors. 'The platform they have created has the potential to deliver faster time-to-market for those demanding the most from chip design,' Goldschmidt added, expressing enthusiasm for leading this funding round.
The growing integration of AI technologies into various domains, including software engineering, data analysis, and graphic design, has begun to reshape how industries function. However, the semiconductor sector has experienced minimal innovation in engineering tools, which continues to be a significant constraint. Given the increasing shortage of talent, low first-silicon success rates, high respin costs, and the pressing need for energy-efficient AI scaling, Normal Computing promises to bridge this gap, shrinking design timelines significantly without compromising the collaborative nature that hardware projects demand.
Normal EDA employs an innovative approach called auto-formalization, an emerging area of AI research that merges large language models (LLMs) with formal logic. This technique allows both AI and human engineers to tackle the toughest aspects of silicon design. The AI platform learns by integrating with operational specifications and workflows, designing and optimizing silicon to ensure correctness, while continuously learning from the process.
Today, Normal's capabilities extend to a range of automation techniques in design verification (DV) and register-transfer level (RTL) automation, remaining committed to enhancing unconventional silicon design workflows that will evolve alongside upcoming software and algorithms. The company actively supports open EDA standards and is a founding member of the Silicon Integration Initiative (Si2) LLM Benchmarking Coalition.
In addition to its software innovations, Normal is making strides in developing its hardware. The Carnot hardware program is focused on advancing its own silicon IP. Recently, Normal announced the successful tape-out of CN101, the world’s first thermodynamic computing chip intended for multi-modal generation AI model inference. This milestone is seen as a step toward a roadmap strategizing for up to 1000x energy efficiency gains and reduced latency in processing.
Unlike standard GPUs that consume energy by minimizing the inherent randomness in physical systems, Normal’s physics-based application-specific integrated circuits (ASICs) utilize these dynamics for more efficient computing. This discovery unlocks new architectural possibilities for demanding AI tasks, such as image and video generation.
The hardware project is partly funded by the Advanced Research + Invention Agency (ARIA), as a contribution to its Scaling Compute Programme. Dr. Suraj Bramhavar, Programme Director of ARIA’s Scaling Compute programme, remarked on the high technical risks associated with innovative ideas like those pursued by Normal. 'The fact that Normal has successfully delivered working silicon in CN101 is an exceptional accomplishment in such ambitious endeavors,' he noted.
Normal Computing's mission is clear: to harness mathematical and physical intelligence to expedite hardware innovation in collaboration with top semiconductor companies. Established in 2022, the firm seeks to employ cutting-edge AI methodologies and physics-based computing advancements to create a step change in chip design and production. The organization boasts a team of leading experts and engineers from notable tech companies like Meta, Palantir, Graphcore, Apple, Broadcom, NVIDIA, Google Brain, and Google X, in addition to graduates from prestigious national laboratories. With its headquarters in New York, and offices in San Francisco, London, and Copenhagen, Normal Computing is gearing up for further expansion into South Korea. For more information, visit normalcomputing.com.
Founded to confront two pressing issues—the growing complexity of silicon design and the escalating energy requirements associated with AI infrastructure—Normal Computing aims to revolutionize the semiconductor industry. The company specializes in developing AI for semiconductor design, creating a new class of computing hardware that leverages its proprietary software to design its own hardware intellectual property (IP).
Faris Sbahi, CEO of Normal Computing, expressed the urgency surrounding the energy crisis in data centers, which are projected to hit an 'energy wall' by 2030. He emphasized that merely seeking more energy isn't a sustainable solution. 'To meet increasing demands for intelligence per dollar and per watt, we need innovative architectural designs and a unified methodology in design practices. Our objective with Normal EDA is to expedite the path to market for custom silicon by doubling the speed today, with an ambitious goal of achieving 1000x efficiency gains through our platform in the future,' he stated.
The company's growing influence is notable, as it has partnered with over half of the top ten revenue-generating semiconductor firms. Normal Computing distinguishes itself by gaining trust within the industry to facilitate tangible production silicon. Their specialized AI platform for semiconductors, dubbed Normal EDA, aims to both drastically speed up chip design processes and develop architectures that can alter the economic landscape of AI scaling.
Dede Goldschmidt, Senior Vice President and Managing Director of Samsung Catalyst Fund, commented on the profound transformation occurring in the semiconductor industry influenced by AI. He remarked on the increased complexity surrounding leading-edge semiconductor chip design and commended Normal for assembling a skilled team with diverse experience across AI and semiconductors. 'The platform they have created has the potential to deliver faster time-to-market for those demanding the most from chip design,' Goldschmidt added, expressing enthusiasm for leading this funding round.
AI-Powered Silicon Design
The growing integration of AI technologies into various domains, including software engineering, data analysis, and graphic design, has begun to reshape how industries function. However, the semiconductor sector has experienced minimal innovation in engineering tools, which continues to be a significant constraint. Given the increasing shortage of talent, low first-silicon success rates, high respin costs, and the pressing need for energy-efficient AI scaling, Normal Computing promises to bridge this gap, shrinking design timelines significantly without compromising the collaborative nature that hardware projects demand.
Normal EDA employs an innovative approach called auto-formalization, an emerging area of AI research that merges large language models (LLMs) with formal logic. This technique allows both AI and human engineers to tackle the toughest aspects of silicon design. The AI platform learns by integrating with operational specifications and workflows, designing and optimizing silicon to ensure correctness, while continuously learning from the process.
Today, Normal's capabilities extend to a range of automation techniques in design verification (DV) and register-transfer level (RTL) automation, remaining committed to enhancing unconventional silicon design workflows that will evolve alongside upcoming software and algorithms. The company actively supports open EDA standards and is a founding member of the Silicon Integration Initiative (Si2) LLM Benchmarking Coalition.
Advancements in Hardware Technology
In addition to its software innovations, Normal is making strides in developing its hardware. The Carnot hardware program is focused on advancing its own silicon IP. Recently, Normal announced the successful tape-out of CN101, the world’s first thermodynamic computing chip intended for multi-modal generation AI model inference. This milestone is seen as a step toward a roadmap strategizing for up to 1000x energy efficiency gains and reduced latency in processing.
Unlike standard GPUs that consume energy by minimizing the inherent randomness in physical systems, Normal’s physics-based application-specific integrated circuits (ASICs) utilize these dynamics for more efficient computing. This discovery unlocks new architectural possibilities for demanding AI tasks, such as image and video generation.
The hardware project is partly funded by the Advanced Research + Invention Agency (ARIA), as a contribution to its Scaling Compute Programme. Dr. Suraj Bramhavar, Programme Director of ARIA’s Scaling Compute programme, remarked on the high technical risks associated with innovative ideas like those pursued by Normal. 'The fact that Normal has successfully delivered working silicon in CN101 is an exceptional accomplishment in such ambitious endeavors,' he noted.
Normal Computing's mission is clear: to harness mathematical and physical intelligence to expedite hardware innovation in collaboration with top semiconductor companies. Established in 2022, the firm seeks to employ cutting-edge AI methodologies and physics-based computing advancements to create a step change in chip design and production. The organization boasts a team of leading experts and engineers from notable tech companies like Meta, Palantir, Graphcore, Apple, Broadcom, NVIDIA, Google Brain, and Google X, in addition to graduates from prestigious national laboratories. With its headquarters in New York, and offices in San Francisco, London, and Copenhagen, Normal Computing is gearing up for further expansion into South Korea. For more information, visit normalcomputing.com.
Topics Consumer Technology）
【About Using Articles】
You can freely use the title and article content by linking to the page where the article is posted.
※ Images cannot be used.
【About Links】
Links are free to use.