Edgecortix
- Company typePrivate
- Founded2019
- HeadquartersGinza, Japan
- Headcount101–250
- GTM typeB2B
- OfferingHardware or Manufacturing
Edgecortix firmographics
Firmographics- Name
- Edgecortix
- Legal name
- EdgeCortix Pte. Ltd.
- Website
- https://edgecortix.com
- Company type
- Private
- Founded year
- 2019
- Operating status
- Operating
- Headcount range
- 101–250 employees
- Ownership category
- akta.pro rank
Edgecortix industry classification
Industry- Product category
- Edge AI Semiconductor Accelerators
- akta.pro primary industry
- Edge AI Hardware & NPUs (SoCs, modules, accelerators) (HDAAAJAC)
- akta.pro secondary industries
- AI Inference Hardware & Edge Accelerators (HDAAAAAC), Advanced SoCs & Application Processors (HDAHAAAB), AI Hardware IP & Chiplet/Packaging for Accelerators (EDA/IP, advanced packaging) (HDAAAAAL)
Keywords
Where Edgecortix is headquartered
LocationHeadquarters
- HQ city
- Ginza
- HQ country
- Japan
- HQ region
- Asia
Offices5 records
Markets served
Edgecortix business model
Business model- GTM type
- B2B
- Offering type
- Hardware or Manufacturing
- Cost components
- Technology or R&D, Personnel, Operations, Marketing or Sales, Supply Chain
Revenue model
- AI Accelerator Hardware Sales: Sale of SAKURA-II AI accelerator modules, cards, and systems. Products include M.2 modules and larger accelerator cards delivering up to 240 TOPS in system configurations. Hardware sales to defense, aerospace, space, and industrial customers.
- Software Licensing (MERA): Licensing of MERA compiler and software framework for AI inference deployment. Software is licensed separately under MERA Software EULA. Provides toolchain for model optimization and deployment across heterogeneous systems.
- IP Licensing: Licensing of DNA (Dynamic Neural Accelerator) processor IP to semiconductor companies for integration into their products. Collaboration with Renesas to deliver enhanced compiler for DRP-AI accelerator.
Go-to-market motion2 records
Distribution channels5 records
Marketing channels8 records
Edgecortix product offering
Product offeringCore offering
EdgeCortix is a fabless semiconductor company that designs and sells energy-efficient AI accelerator chips (SAKURA-I, SAKURA-II and forthcoming SAKURA-X chiplet platform) and the companion MERA compiler/software framework, all underpinned by its proprietary Dynamic Neural Accelerator (DNA) runtime-reconfigurable processor architecture. The hardware is delivered as bare chips, M.2 modules and accelerator cards (up to 240 TOPS in systems), while the DNA IP and MERA software are also licensed to semiconductor partners such as Renesas.
Product overview
EdgeCortix is a fabless semiconductor company offering a portfolio of AI-specific processors and companion software. The core offering consists of the SAKURA-II edge AI accelerator (flagship, delivering up to 60 TOPS at 8W with M.2 module variants delivering up to 240 TOPS in systems), the SAKURA-I first-generation co-processor, and the upcoming SAKURA-X chiplet-based next-generation platform. These hardware products are underpinned by the proprietary Dynamic Neural Accelerator (DNA) architecture — a flexible, runtime-reconfigurable processor design. The software stack is anchored by the MERA Compiler and Framework, which enables AI inference across heterogeneous systems. Together, the DNA-powered accelerators and MERA software form an integrated hardware-software edge AI platform for energy-efficient generative AI at the edge.
Differentiator
Problem solved
Functional benefit
Brands
- SAKURA-II: World's most flexible and energy-efficient AI accelerator platform, handling multi-billion parameter models for generative AI at the edge.
- SAKURA-X
- MERA
- Dynamic Neural Accelerator (DNA)
Products and services
- SAKURA-II AI Accelerator EdgeCortix's flagship edge AI accelerator chip delivering up to 60 TOPS in an 8W power envelope, designed for high-performance, low-power generative AI inference at the edge including LLMs and large vision models. Built on the second-generation Dynamic Neural Accelerator (DNA) architecture.
- SAKURA-II M.2 Module and Accelerator Cards M.2 modules and accelerator cards powered by SAKURA-II AI accelerators, delivering up to 240 TOPS in systems for embedded and edge deployment scenarios.
- SAKURA-I AI Co-processor First-generation edge AI co-processor from EdgeCortix, delivering industry-leading low latency and high energy efficiency, with validated radiation resilience for space and orbital missions. Selected by BittWare as the edge-focused AI acceleration solution for Intel Agilex FPGA-based products.
- SAKURA-X Chiplet Platform Next-generation chiplet-based AI platform under development for generative and agentic AI workloads, targeting higher performance and energy efficiency than SAKURA-II. Expected to deliver 5x better power efficiency compared to NVIDIA GPUs.
- MERA Compiler and Framework Industry-first compiler and software framework enabling AI inference across heterogeneous systems. Licensed separately under MERA Software EULA and integrated with Renesas' RZ/V MPU series to enable AI inference on Renesas embedded processors.
- Dynamic Neural Accelerator (DNA) Processor IP Proprietary runtime-reconfigurable, highly-parallelized processor architecture licensed to semiconductor companies (such as Renesas) for integration into their own AI accelerator products. Protected by multiple patents.
Quantifiable outcome
- 8W power consumption for 60 TOPS performance (vs. GPU alternatives requiring 100s of watts)
- +4 more outcomes
Companies that use Edgecortix
Customer profileNamed customers7 records
Segments8 records
Ideal customer profiles5 records
Edgecortix technology and API
TechnologyTechnology focussed Yes
API detail
- Has API
- No
- API docs
- API detail
Core technology
AI maturity
App detail
Integration3 records
AI capability5 records
Feature5 records
Edgecortix partnerships and signals
Strategic signalPartnerships
Eleven partnerships are on record, tiered core, flagship and minor.
- Carnegie Mellon University (Software Engineering Institute)coreCarnegie Mellon University's Software Engineering Institute provided independent validation of AI benchmarking performance for SAKURA-II platform as part of DIU project.
- North American, European, and MENA PartnerscoreEdgeCortix announced expansion of global reach with strategic partnerships across North America, Europe, and MENA regions for broader market coverage.
- U.S. Defense Innovation Unit (DIU)flagshipEdgeCortix became the first Japanese and semiconductor company awarded a DIU contract. DIU provides rapid evaluation pathway for EdgeCortix hardware and software solutions to be deployed across US Defense Department advanced technology programs.
- ispacecoreCollaboration on latest AI-enabled lunar missions. ispace to utilize EdgeCortix low-power edge AI processors for cislunar development, advancing autonomous capabilities for lunar exploration.
- Saudi Arabia Ministry of Investment / National Semiconductor HubcoreEdgeCortix selected by Saudi Arabia's Ministry of Investment to expand into Riyadh in partnership with National Semiconductor Hub, marking entry into Middle East market.
- NASAflagshipNASA conducted proton and heavy-ion radiation testing on SAKURA-II accelerator at Texas A&M cyclotron facility. Testing validated high radiation resiliency for space and defense AI applications from LEO through cislunar regimes.
- AI-RAN AlliancecoreEdgeCortix joined AI-RAN Alliance to accelerate integration of AI and next-gen RAN infrastructure. Contributing AI acceleration and wireless acceleration hardware expertise for energy-efficient network infrastructure.
- Renesas ElectronicscoreStrategic collaboration where EdgeCortix developed enhanced compiler for Renesas DRP-AI accelerator. Renesas invested in EdgeCortix and appointed Mohammed Dogar to EdgeCortix board. Collaboration enables customers to rapidly deploy AI-inference solutions on Renesas embedded processors.
- BittWarecoreBittWare selected EdgeCortix SAKURA-I AI processors as edge-focused AI acceleration solution for Intel Agilex FPGA-based product portfolio. BittWare delivers high-performance, low-latency AI acceleration for enterprise customers.
- AetinacoreAetina offers AI computing modules powered by EdgeCortix accelerators for industrial and enterprise applications.
- Global Semiconductor Alliance (GSA)minorEdgeCortix joined GSA to engage with global semiconductor ecosystem and industry peers.
Scale indicators8 records
Recent moves8 records
Expansion highlights8 records
Edgecortix competitors and assessment
Company assessmentDirect peers
- Hailo: Hailo is an Israel-based fabless edge AI processor company shipping the Hailo-8/15 NPU family targeting computer vision and now generative AI at the edge. It is the closest direct peer to EdgeCortix in edge AI silicon targeting embedded, automotive, and industrial customers.
- Axelera AI: Axelera AI is a Netherlands-based fabless edge AI accelerator company whose Metis chip targets computer vision and generative AI inference at low power. It competes head-to-head with SAKURA-II in industrial, robotics, and European enterprise edge segments.
- Syntiant: Syntiant designs ultra-low-power neural decision processors for always-on edge AI in IoT, hearables, and battery-powered devices. It overlaps with EdgeCortix on low-power edge inference but is more focused on tinyML/voice than generative AI workloads.
- Mythic AI: Mythic develops analog compute-in-memory edge AI processors for vision and sensor analytics. Although smaller in scale, it is a direct competitor in low-power edge AI silicon serving industrial and security camera markets overlapping with EdgeCortix's industrial segment.
Emerging players
- Sima.ai: Sima.ai is an emerging U.S. startup building edge AI MLSoC devices focused on computer vision and embedded AI. It is a niche, emerging-edge-AI silicon player with partial overlap with EdgeCortix's industrial and embedded vision use cases.
- Tenstorrent: Tenstorrent develops RISC-V-based AI accelerators spanning edge to data center, with a comparable software-first philosophy and growing focus on chiplet architectures. It is an emerging player with broader scope but overlapping AI silicon IP and tooling ambitions.
- Groq: Groq builds LPU inference accelerators primarily for data center, but its low-latency, energy-efficient inference design philosophy overlaps with EdgeCortix's DNA architecture. It represents an adjacent competitive threat as inference moves toward the edge.
- Esperanto Technologies: Esperanto Technologies develops RISC-V-based AI accelerators for inference, with emphasis on energy efficiency and chiplet-style integration. It is an emerging AI silicon peer with overlapping philosophy on software-defined hardware and low-power inference.
Broad incumbents
- NVIDIA (Jetson): NVIDIA's Jetson edge AI platform dominates the edge AI silicon market with the Orin and upcoming Thor SoCs. It is the broad incumbent EdgeCortix benchmarks against on power efficiency and developer ecosystem, even though NVIDIA does not target radiation-hardened or ultra-low-power niches.
Others
- Renesas Electronics: Renesas is both an investor and partner licensing EdgeCortix's DNA IP for the DRP-AI accelerator and integrating MERA. As a global microcontroller and embedded processor incumbent, it serves as both a complementary channel and a partial substitute for AI inference on embedded platforms.
Market position
Strengths5 records
Weaknesses5 records
Competitive moat6 records
Key risks6 records
Key highlights7 records
Customer concentration
Edgecortix social profiles
Digital presenceEdgecortix financial estimates
Financial estimateRevenue estimate
Valuation estimate
Edgecortix leadership team
Management profileNumber of profiles
Profiles12 records
Edgecortix funding detail
Funding detailFunding overview
Funding rounds7 records
Investors18 records
Funding detail is available on the Subscription and Enterprise plan.Contact sales →
Edgecortix M&A and investment
M&A and investmentM&A
Investments
M&A and investment is available on the Subscription and Enterprise plan.Contact sales →
Frequently asked questions about Edgecortix
What does Edgecortix do?
EdgeCortix is a fabless semiconductor company that designs and sells energy-efficient AI accelerator chips (SAKURA-I, SAKURA-II and forthcoming SAKURA-X chiplet platform) and the companion MERA compiler/software framework, all underpinned by its proprietary Dynamic Neural Accelerator (DNA) runtime-reconfigurable processor architecture. The hardware is delivered as bare chips, M.2 modules and accelerator cards (up to 240 TOPS in systems), while the DNA IP and MERA software are also licensed to semiconductor partners such as Renesas.
Is Edgecortix a public or private company?
Edgecortix is a private company. It is classified as venture growth investor backed and is currently operating.
When was Edgecortix founded?
Edgecortix was founded in 2019. It employs 101 to 250 people.
Where is Edgecortix based?
Edgecortix is headquartered in Ginza, Japan, in the Asia region.
How does Edgecortix make money?
Three revenue lines are on record. AI Accelerator Hardware Sales are the primary driver. The others are software Licensing (MERA) and IP Licensing.
Who are Edgecortix's main competitors?
Direct peers on record are Hailo, Axelera AI, Syntiant and Mythic AI. Emerging players are Sima.ai, Tenstorrent, Groq and Esperanto Technologies. NVIDIA (Jetson) is listed as a broad incumbent. Renesas Electronics is listed as an others.
Does Edgecortix have an API?
No public API is recorded for Edgecortix.
What industry is Edgecortix in?
Edgecortix's product category is Edge AI Semiconductor Accelerators. Its primary akta.pro industry code is HDAAAJAC, Edge AI Hardware & NPUs (SoCs, modules, accelerators), with a secondary code of HDAAAAAC, AI Inference Hardware & Edge Accelerators.