EdgeCortix Receives 4 Billion Yen Subsidy from Japan’s NEDO to Advance Energy-Efficient AI Chiplets for Post-5G Communication Systems
Substantial subsidy to support cutting-edge AI semiconductor development, enhancing energy efficiency and performance in AI inference and Radio Access Network (RAN) acceleration.
Tokyo, Japan -- November 7, 2024 — EdgeCortix® Inc., a fabless semiconductor company specializing in energy-efficient AI processing, announced today it has been awarded a 4 Billion Yen subsidy from Japan’s New Energy and Industrial Technology Development Organization (NEDO), a national research and development agency under the Ministry of Economy, Trade and Industry (METI). This funding will propel the development of next-generation, energy-efficient AI chiplets designed to revolutionize multi-tenancy systems by powering both AI processing and advanced RAN communications with unprecedented energy efficiency.
As digital transformation accelerates, advanced networks are essential for enabling ultra-fast communication, low latency, and broad device connectivity. Yet, edge computing and cloud infrastructure face deployment challenges, especially as generative AI applications place new demands on future networks. Current systems relying on single-purpose servers for either AI deployment or RAN processing are costly, energy-intensive, and difficult to scale. AI-Integrated RAN (AI-RAN) presents a promising alternative but continues to struggle with energy consumption and cost barriers.
This subsidy enables EdgeCortix to address these challenges by developing SAKURA-X, a pioneering chiplet-based platform that seamlessly integrates high-performance AI processing with RAN acceleration (focusing on distributed units). SAKURA-X is expected to reduce energy consumption and improve computational efficiency by more than five times (5x) compared to current solutions. The platform will be developed using TSMC’s technology node, targeting mass production at Japan Advanced Semiconductor Manufacturing facility in Kumamoto, in the future.
The project builds upon EdgeCortix’s proven Dynamic Neural Accelerator® (DNA) next-generation, DNA-X, AI processor technology, known for its proprietary programmable architecture that excels in both AI and non-AI computations with high speed, real-time processing, and low power requirements. Achieving up to 90% utilization, far exceeding conventional GPUs, DNA-X combines with EdgeCortix’s novel RAN accelerator IP within SAKURA-X, delivering optimized performance and minimized power consumption through advanced power management and co-design strategies. When deployed in next-generation Open RAN or AI-RAN systems, SAKURA-X will support both generative AI applications and low-power distributed unit (DU) acceleration on a unified hardware platform, achieving unparalleled energy efficiency and infrastructure cost savings.
“With the rapid evolution of 5G and the advent of 6G, demand for high-performance, energy-efficient computing has never been more critical,” said Dr. Sakyasingha Dasgupta, Founder & CEO of EdgeCortix Inc. “We are honored to receive strong support from Japan’s NEDO as we develop SAKURA-X, a transformative chiplet-based platform built on our legacy in AI acceleration. SAKURA-X surpasses existing solutions with over five times the energy efficiency and performance, meeting the surging data demands of generative AI and next-gen RAN, while dramatically reducing energy consumption. With its advanced AI-RAN capabilities, SAKURA-X will empower Japan’s telecommunications sector to expand AI-enhanced networking globally, supporting growth across smart factories, autonomous driving, data centers, and beyond.”
The SAKURA-X chiplet architecture introduces scalable solutions that redefine next-generation communications. By offering high-speed, low-power AI inference alongside integrated AI processing and RAN acceleration, these innovations promise unmatched efficiency and cost-effectiveness. Post validation by key partners in Japan’s telecommunications industry, they are set to advance the future of energy-efficient AI and communication networks.
About EdgeCortix Inc.
Pioneering the future of the connected intelligent edge, EdgeCortix is a fabless semiconductor company focused on energy-efficient AI processing of Generative-AI workloads at the edge. Founded in 2019 with R&D headquarters in Tokyo, Japan, EdgeCortix delivers a software-first approach with its patented “hardware and software co-exploration” system to design an AI specific runtime reconfigurable accelerator processor from the ground up. EdgeCortix’s products disrupt the rapidly growing edge AI hardware markets including telecommunications, defense, aerospace, smart cities, industry 4.0, autonomous vehicles and robotics.
For more information about EdgeCortix and its solutions, please visit: https://www.edgecortix.com/en/
Related Chiplet
- Direct Chiplet Interface
- HBM3e Advanced-packaging chiplet for all workloads
- UCIe AP based 8-bit 170-Gsps Chiplet Transceiver
- UCIe based 8-bit 48-Gsps Transceiver
- UCIe based 12-bit 12-Gsps Transceiver
Related News
- Tenstorrent Partners with LG to Build AI and RISC-V Chiplets for Smart TVs of the Future
- Axelera plans AI chiplets for automotive
- How do UCIe and BoW interconnects support generative AI on chiplets?
- ASE’s VIPack™ Enables Innovational AI Devices Through Advanced Interconnect Technology for Chiplets
Latest News
- Eliyan Ports Industry’s Highest Performing PHY to Samsung Foundry SF4X Process Node, Achieving up to 40 Gbps Bandwidth at Unprecedented Power Levels with UCIe-Compliant Chiplet Interconnect Technology
- GlobalFoundries and U.S. Department of Commerce Announce Award Agreement on CHIPS Act Funding for Essential Chip Manufacturing
- One Chip Vs. Many Chiplets
- Cadence Unveils Arm-Based System Chiplet
- Chiplets: Revolutionizing Semiconductor Design and Manufacturing