Listen to this article
MemryX has raised $44 million in Series B funding to scale production of its MX3 AI accelerator and develop the next-generation MX4 chip. The company’s energy-efficient, software-first architecture supports a range of edge applications across industries such as IoT, transportation, and industrial systems. With growing adoption and expanded global operations, MemryX is strengthening its position in the edge AI semiconductor market.
Big Money Moves: MemryX Lands $44M to Fuel Edge AI Growth
MemryX secured $44 million in Series B funding to advance its edge AI semiconductor development. The round received participation from both new and existing investors. Key contributors include HarbourVest, M Ventures, eLab Ventures, Motus Ventures, and Arm IoT Fund. The announcement follows MemryX’s production-ready release of its MX3 Accelerator chip, which has gained adoption across multiple enterprise use cases.
The funding supports MemryX’s goal of expanding its existing hardware production and completing the design of its next-generation MX4 accelerator. According to CEO Keith Kressin, the company’s software-first approach has allowed it to remain capital efficient, while serving dozens of customers and maintaining a scalable development pipeline.
The MX3 Accelerator: What Makes MemryX Chips Stand Out
The MX3 is designed to deliver high performance for AI inference at the edge using significantly less power than conventional solutions. Compared to mainstream GPUs, the MX3 operates at under 10% of the power requirement while maintaining a high level of performance. The chip is integrated by a range of companies in production environments, including AI-driven Video Management Systems, Industrial PCs, and Edge Servers.
MemryX provides open access to performance metrics and benchmarking data through its Developer Hub, supporting evaluation across hundreds of AI models. This transparency allows developers to assess integration viability in various deployment scenarios and supports customer-specific development processes.
Why Everyone’s Talking About Software-First Architecture
MemryX uses a compute-at-memory technology backed by a proprietary dataflow architecture. This design supports efficient execution by reducing data movement and improving overall throughput. The architecture’s software-first nature allows compatibility with a wide variety of AI frameworks and models.
The company’s strategy centers on enabling faster deployment and lower total cost of ownership for customers. Its software stack is built to streamline integration, which supports quicker transition from development to production. This focus on software infrastructure has helped MemryX maintain a lean operation while developing advanced silicon products.
Recommended: An Interview With Alex Tay, Founder And CEO Of ZeusX
What’s Next: The MX4 Accelerator and Expansion Strategy
MemryX is actively developing the MX4 chip, which is positioned as the successor to the current MX3 platform. Series B funding will accelerate design and production timelines. While technical specifications of the MX4 remain under wraps, leadership confirmed that the architecture will continue to evolve along its current energy-efficient, high-performance trajectory.
MemryX operates globally, with headquarters in Ann Arbor, Michigan, and offices in Bangalore, Taipei, and Hsinchu. These locations serve engineering, customer support, and business development functions. Continued growth across regions reflects the company’s expanding reach and ongoing customer engagement.
Industry Momentum: MemryX’s Position in the Edge AI Race
MemryX’s chips are built for industries requiring low-latency, high-efficiency AI at the network edge. Use cases span transportation, industrial automation, IoT, and consumer-facing systems. The company reports adoption across more than a dozen enterprise customers currently moving into production phases.
The company plans to showcase its products and solutions at ISC West 2025 in Las Vegas. It will demonstrate use cases in collaboration with hardware partners. The event serves as a platform to highlight performance benchmarks and allow prospective clients to interact with operational hardware.
MemryX’s recent partnership with Variscite, a system-on-module provider, was previously announced to further enhance edge AI capabilities. This collaboration is one of several that extend MemryX’s integration with third-party platforms and hardware ecosystems.
Why This Funding Round Signals a Turning Point for Edge AI
MemryX continues to expand its presence in the edge AI market through targeted hardware innovation and a robust software ecosystem. The successful completion of the Series B round validates its direction and enables delivery of the MX4, while scaling adoption of the MX3 across industrial and enterprise applications.
By focusing on compute efficiency, integration readiness, and global support infrastructure, MemryX aligns its growth strategy with the increasing demands for energy-efficient edge AI. The combination of funding, product momentum, and developer accessibility positions the company to influence ongoing advancements in edge computing.
Please email us your feedback and news tips at hello(at)techcompanynews.com
