Normal Computing, a New York-based developer of AI-powered electronic design automation (EDA) software and thermodynamic application-specific integrated circuits (ASICs), has raised $50M in strategic funding led by Samsung Catalyst. The platform combines Normal EDA for zero-defect chip verification with physics-based ASICs promising 1000x reductions in AI energy use. The capital will accelerate silicon design and address the AI hardware energy crisis.
AI EDA Funding Heats Up
The raise follows ChipAgents securing $74M Series A1 in February 2026 for agentic AI in chip design. Normal's full-stack approach spans EDA software used by over half of the top 10 semiconductor companies by revenue and its CN101 thermodynamic chip, taped out in June 2025. This positions it amid surging demand for verification tools as AI chips grow more complex.
Chip Design Hits Energy Wall
Data centers face an energy wall around 2030 due to AI power demands, with U.S. power use projected to hit record highs in 2026-2027. AI chip development costs exceed $500M, where verification errors prove costly. Current CMOS-based systems struggle with efficiency for probabilistic AI workloads like diffusion models.
Autoformalizing AI Speeds Verification
Normal EDA uses AI to autoformalize test plans and generate production-grade verification artifacts, integrating with tools like VCS and XCelium for 2x faster time-to-market. Unlike pure software rivals, Normal pairs this with thermodynamic ASICs that harness physics and noise for stochastic computing in AI inference and scientific simulations.
Thermodynamic Chips Embrace Noise
The CN101 prototype validates stateful, asynchronous compute, targeting 1000x energy gains over GPUs for generative AI. As CEO Faris Sbahi noted:
"The mission of the company is to go after this so-called AI energy crisis. Data centers are expected to hit an energy wall around 2030."
This physics+ML approach stems from founders' Google Brain and X backgrounds.
Samsung Validates Hardware Shift
Samsung Catalyst leads with new investors Galvanize, Brevan Howard Macro Venture Fund, and ArcTern Ventures joining Celesta Capital, Drive Capital, Eric Schmidt's First Spark Ventures, and Micron Ventures. Prior raises total $93.5M including $35M in March 2025. Strategic semi investors signal conviction in Normal's dual EDA-ASIC play against GPU dominance.
Neuromorphic Market Powers Ahead
The neuromorphic computing market stands at $7.5B with a 16.5% CAGR. Competitors like Extropic ($14M funded) focus on thermodynamic hardware, while Normal integrates EDA for custom silicon. AI EDA grows from $4.27B in 2026 to $15.85B by 2032 at 24.4% CAGR, driven by verification bottlenecks highlighted at DVCon 2026.
Ex-Google Brain Team Leads
Founders include ex-Google Brain's Faris Sbahi, who led 9-figure Bayesian ML projects, and X moonshot alumni. Head of ML Thomas Dybdahl Ahle co-founded SupWiz, acquired by Puzzel in 2024. This physics-AI expertise powers Normal's innovations, with 71 employees across four offices.
Scaling Post-Tapeout Momentum
Normal eyes production scaling of Normal EDA, already in use at leading semis, and CN101 silicon bring-up results from December 2025. Recent hires like Chief Business Officer Craig Churchill (ex-Google X Tidal) bolster GTM amid partnerships like Mirafra Technologies.