Skip to main content

EdgeCortix

Category: AI Infrastructure

EdgeCortix develops energy-efficient AI semiconductors and accelerators for edge computing, focusing on low-power AI inference and machine learning workloads. EdgeCortix was founded in 2019. The company is led by Sakyasingha Dasgupta. Based in Tokyo, Japan. Team size: 50-200. Total funding raised: $110M+. Latest round: Series B (oversubscribed, closed November 2025). Key investors include ["TDK Ventures","CDIB Capital Innovation Advisors","Jane Street Global Trading","Macquarie Asia Conference","BittWare","SBI Investment","GHOVC","Renesas Electronics","Yanmar Holdings","NTT Finance Corporation","Axiro Semiconductor","MPower Partners"].

Founded
2019
Headquarters
Tokyo, Japan
Team size
50-200
Total funding
$110M+

Value proposition

Delivers high-performance, energy-efficient AI processing for edge devices with near cloud-level performance while reducing power consumption and enabling real-time AI inference in resource-constrained environments

Products and solutions

["SAKURA-II AI accelerators","MERA compiler and framework for AI inference","Dynamic Neural Accelerator technology","AI accelerator modules and cards","Low-power AI chiplet platforms"]

Unique value

Software-first approach to AI processor design using patented 'hardware & software co-exploration' technique, combining proprietary compilers with hardware optimization for maximum energy efficiency

Target customer

Industries requiring edge AI solutions including industrial automation, defense, aerospace, smart cities, autonomous vehicles, robotics, and semiconductor manufacturing

Industries served

["Defense and aerospace","Industrial automation (Industry 4.0)","Smart cities","Autonomous vehicles","Robotics","Semiconductor manufacturing","Space technology"]

Technology advantage

Patented low-power AI inference architectures, runtime reconfigurable processors, and collaboration with Japanese NEDO for next-gen chiplet platforms. First Japanese semiconductor company awarded U.S. Defense Innovation Unit contract

How they differentiate

EdgeCortix differentiates through energy-efficient AI processors using a software-first design approach, enabling rapid deployment on existing hardware (FPGAs/ASICs). Focus on low-power edge inference with proprietary compiler technology and runtime reconfigurable architecture

Main competitors

["Preferred Infrastructure","Netrasemi","Celestial AI","Rebellions","Literal Labs","Mythic"]

Key partnerships

["Renesas (compiler integration for RZ/V MPU series)","TDK Ventures (global expansion and AI platform development)","Saudi Arabia's National Semiconductor Hub (Riyadh office)","BittWare (FPGA-based AI solutions)","NEDO 3 billion yen project for advanced AI chiplet development","ispace (space technology partnership)","Arm, TSMC, Cadence, Synopsys (technology partnerships)","Axiro Semiconductor (strategic investment and partnership for RF/mmWave/connectivity IC)","MPower Partners (strategic investment)","NASA (heavy ion radiation testing validation for SAKURA-II)"]

Notable customers

["U.S. Defense Innovation Unit (contract award)","Renesas Electronics (partnership)","ispace (space technology)","NASA (SAKURA-II radiation validation for space missions)","Various industrial automation clients in Japan"]

Major milestones

["Series B funding exceeding $110M (2025)","Strategic partnership with Renesas Electronics (2025)","Expansion into MENA region through Ebttikar Technology (2025)","U.S. Department of Defense contract award","Product launch of SAKURA-II AI accelerator","3 billion yen NEDO project award for AI chiplet development","Forbes Technology Council induction (2025)","NASA validates SAKURA-II radiation resilience for space missions (Jan 2026)","New strategic investment from Axiro Semiconductor and MPower Partners (Apr 2026)"]

Growth metrics

N/A (Private company)

Market positioning

Niche player in edge AI acceleration, targeting industrial automation, defense, and IoT sectors with hardware-software co-optimized solutions. Positions itself as a challenger to GPU-centric AI acceleration models

Geographic focus

Primary focus on Asia-Pacific (Japan as headquarters), with expansion into North America (US office in Arlington, VA) and Europe, plus MENA region partnerships

Patents and IP

Hold over 20 patents worldwide for energy-efficient AI chiplets and inference acceleration, with 5 key patents granted and others pending

About Sakyasingha Dasgupta

Founder and CEO with over two decades of experience in AI research, semiconductor design, and entrepreneurship. Previously led AI research at Max Planck Society and developed energy-efficient AI processors. Holds PhD in Physics of Complex Systems from Max Planck Institute and Masters in AI from University of Edinburgh.

Official website: