Ainekko, a startup focused on open, software-defined AI hardware, has announced a merger with Veevx, a company known for its embedded AI solutions and innovative memory tech. This collaboration aims to create more flexible, community-driven silicon platforms that empower developers and researchers. By combining their strengths, they hope to reshape the future of edge AI devices and hardware design.

Combining Open-Source Hardware with Advanced Memory Technology

The merger brings together Ainekko’s open-source approach to silicon design with Veevx’s expertise in embedded AI and memory innovations. Veevx specializes in MRAM-based memory technology called iRAM, which offers high-density, non-volatile storage with SRAM-like performance. This breakthrough addresses common issues like memory bandwidth and power consumption that often slow down AI workloads at the edge.

With this partnership, Ainekko plans to develop a full-stack, open silicon platform optimized for rapid AI innovation. This includes hardware that integrates flexible memory, processing power, and development tools. The goal is to give creators the ability to design custom chips that are both efficient and easy to modify, much like how Linux revolutionized software.

Leadership and Strategic Moves Drive Innovation

Douglas Smith, co-founder of Veevx and a veteran engineer who previously managed intellectual property at Broadcom, will join Ainekko’s leadership team. His experience will help guide the company’s next steps in developing open, community-driven silicon solutions. Additionally, Ainekko recently acquired intellectual property from Esperanto Technologies, including a manycore RISC-V chip architecture, further expanding its technological foundation.

According to Tanya Dadasheva, CEO of Ainekko, this merger is a significant step toward creating a new kind of hardware platform. She compares their vision to what Linux did for operating systems and Kubernetes did for cloud infrastructure. The combined effort aims to build open, modular hardware that accelerates AI development at the edge.

Douglas Smith emphasizes the importance of combining their memory expertise with Ainekko’s open approach. He describes their joint effort as a leap forward in AI hardware innovation, enabling smarter, energy-efficient devices that are easier to develop and customize.

Future of Edge AI and Memory Innovation

The merged company plans to develop intelligent memory subsystems, like iRAM, that replace traditional SRAM with more energy-efficient, scalable options. This is vital for AI inference tasks that require fast, reliable memory but are limited by power constraints at the edge.

They also aim to create embedded accelerators optimized for inference, co-packaged with microcontrollers. By offering open-source RTL designs and emulators, they want to make it easier for developers to prototype and build custom AI hardware tailored to specific needs.

This collaboration signals a move toward open, community-powered silicon platforms that can keep pace with the growing demands of AI workloads. By integrating innovative memory tech and flexible hardware design, Ainekko and Veevx are working to make intelligent devices more accessible, efficient, and adaptable at the edge of networks.

Inspired by

• https://ai-techpark.com/ainekko-veevx-combine-to-advance-embedded-ai-and-memory-tech/