EnCharge AI, which is developing innovative AI-accelerating chips to compete against established giants like Nvidia, AMD, Microsoft, Meta, AWS, and Intel. The vision seemed ambitious then — and remains so today. However, EnCharge AI is thriving, recently securing $22.6 million in a new funding round.
The funding round saw participation from the VentureTech Alliance, a strategic venture capital firm linked to semiconductor leader TSMC, alongside RTX Ventures, ACVC Partners, Anzu Partners, S5V, Alley Corp, Scout, and Silicon Catalyst Angels. With this latest investment, EnCharge AI has raised a total of $45 million, aimed at expanding its team of 50 employees across the U.S., Canada, and Germany, while enhancing the development of its AI chips and “full stack” AI solutions, according to co-founder and CEO Naveen Verma.
“EnCharge’s mission is to democratize AI access for the 99% of organizations that struggle to afford today’s costly and energy-hungry AI chips,” said Verma. “Our goal is to enable new AI applications and hardware configurations that are sustainable both economically and environmentally, to unlock AI's full capabilities."
Verma co-founded EnCharge last year with Echere Iroaga and Kailash Gopalakrishnan. Gopalakrishnan, formerly an IBM fellow, dedicated nearly 18 years to the tech giant. Iroaga previously served as VP and GM of the connectivity division at semiconductor company Macom.
The origins of EnCharge date back to federal grants Verma received in 2017, collaborating with researchers at the University of Illinois at Urbana-Champaign. This initiative stemmed from DARPA’s Electronics Resurgence Initiative, which aims to propel various computer chip technologies forward. Verma led a comprehensive $8.3 million project to explore new types of non-volatile memory devices.
Unlike the “volatile” memory commonly used in today’s computers, non-volatile memory keeps data without needing a continuous power supply, theoretically enhancing energy efficiency. Additionally, DARPA funded Verma's research into in-memory computing — which runs calculations directly in RAM to minimize latency caused by storage devices.
EnCharge was established to bring Verma’s research to market. Utilizing in-memory computing, the startup claims its hardware can accelerate AI applications for servers and edge devices while reducing power consumption compared to conventional processors.
“Currently, AI computation is both expensive and power-intensive; only the most financially robust organizations are pioneering in this space,” Verma explained. “For many, scaling AI in their operations or products is still out of reach. EnCharge is positioned to provide the necessary processing power while overcoming the prohibitive energy costs challenging many companies.”
While ambitious, it is important to note that EnCharge has yet to begin mass production of its hardware, with only a handful of customers secured so far. Moreover, the startup faces intense competition within a crowded AI accelerator hardware market, with players like Axelera and GigaSpaces also focusing on in-memory hardware to enhance AI workloads. NeuroBlade, similarly, has garnered substantial VC backing for its in-memory inference chip designed for data centers and edge devices.
Evaluating EnCharge's performance claims can be challenging, as third parties have not yet had the opportunity to benchmark the startup's chips. Nevertheless, its investors remain optimistic about its potential.
“EnCharge is tackling significant challenges about computing power, accessibility, and costs that hinder current AI capabilities while being insufficient for future demands,” remarked Kai Tsang from the VentureTech Alliance via email. “The company has engineered a technologically unique architecture that surpasses today’s system limitations and integrates seamlessly into the current supply chain.”
Ranking
