EdgeCortix® the AI-focused fabless semiconductor company announced that it has closed over $8 Million USD in series A funding. EdgeCortix’s reconfigurable Dynamic Neural Accelerator® processor architecture and AI acceleration SoC SAKURA™ brings best-in-class power-efficiency with near cloud-level performance across infrastructure and embedded edge devices. EdgeCortix solutions offer lower latencies for real-time applications and greater than 10x improvement in frames per second per watt (FPS/Watt) compared to existing GPU based solutions.
As organizations embrace the rapid growth in AI for edge computing, the challenge is to overcome the mismatch between the current generation AI software or deep neural networks and general-purpose processors like CPUs and GPUs. In addition, there is a growing need and focus on developing environmentally friendly and superior energy-efficient solutions. “Completion of this funding round and growing partnerships, is a testament to our team, technology, and growing product traction. We invented the Dynamic Neural Accelerator processor IP taking a unique software-first approach, forming a tight coupling between today’s neural networks and an underlying low-power domain-specific hardware architecture that is backed with robust software to seamlessly process applications across different machine learning frameworks and devices. In the past year, we have been awarded 7 patents world-wide and have accelerated our filings to more than one new patent each month. We are currently working with over 20 globally recognized customers and partners in the areas of defense, autonomous vehicles, satellite communications, smart cities and semiconductor design and manufacturing, helping them to leverage our IP and deliver greater value to their end customers.” said Sakyasingha Dasgupta PhD., founder and CEO EdgeCortix, Inc.
“EdgeCortix is in a truly unique market position. In addition to taking advantage of the massive business need in leveraging AI acceleration across key industry sectors, their business strategy, with respect to how they have developed their solutions and unique go-to-market, is the true differentiator.” said Daniel Fujii, a member of the EdgeCortix Board of Advisors, President & CEO of Trust Capital Co., Ltd. and member of the Executive Committee of Silicon Valley Japan Platform. “In my experience, semiconductor companies can tend to be a bit myopic, some exclusively focus on delivering code, others on semiconductor design and manufacture. What makes EdgeCortix distinctive is their approach, by co-developing their software and hardware IP and delivering the solutions based on customer’s needs. The software uniquely supports heterogeneous platforms like FPGAs, third party custom silicon and also EdgeCortix’s own SoC. It’s this approach to delivering on a software-centric hardware ecosystem that positions it for growth and truly sets the company apart as unique in the AI acceleration space.”
In addition to the development of its own proprietary processor, EdgeCortix is actively collaborating with leading field-programmable gate array (FPGA) companies to license their AI hardware and software IP for both edge and server acceleration use-cases. Along with this funding announcement, EdgeCortix announced it’s partnering with leading enterprise class FPGA acceleration manufacturer BittWare (a Molex company) to bring to market EdgeCortix’s ultra-low latency AI inference IP and robust software with select BittWare hardware solutions in 2022.