EnCharge AI recently joined the National Semiconductor Technology Center (NSTC) as one of the first 100 members since its establishment in early 2024. NSTC is a public-private consortium established by the CHIPS Act to support and extend U.S. leadership in semiconductor research, design, engineering, and advanced manufacturing.
At last year’s National Semiconductor Technology Center (NSTC) Symposium in Washington, D.C., our CEO, Naveen Verma, delivered a compelling keynote on America’s innovation capacity by connecting fundamental research with industrial R&D. A summary of his talk is below and can be viewed in its entirety here.
America’s innovation ecosystem
America is an innovation juggernaut with groundbreaking discoveries stemming from multiple critical corners of the country’s ecosystem. Semiconductors are just one of many areas of innovation strength in the United States, recently bolstered by investments and leadership established in the CHIPS Act. However, over the past decade, there has been an unprecedented explosion in the computational needs for AI. This 10,000x increase in both compute operations and data movement requirements highlights the need for transformative and differentiated technological solutions that can readily transition into existing innovation ecosystems.
Fundamental research, innovation ecosystems, and industrial leadership constitute the three key spheres of innovation in the United States. Each of these spheres has a different yet complementary objective and risk tolerance. Fundamental research is a critical source of breakthrough ideas and is highly acceptant of failures as an essential component of exploring technological limits. Innovation ecosystems, including startups and research labs, are tasked with transitioning transformative ideas into marketable products, while industrial leadership must consistently deliver platforms that reliably provide the base for leveraging technological capabilities.
The synergistic approach to technological innovation
Fundamental research can provide insights into technological trade-offs. For example, research conducted at Princeton prior to the creation of EnCharge was critical to understanding signal-to-noise ratio limitations in different approaches to in-memory computing. This type of research is particularly valuable because it is non-competitive and allows for collaborative exploration of technological boundaries. In this pre-commercial sphere, federal funding has played a longstanding role in providing critical resources to explore ideas that have led to breakthroughs that maintain U.S. innovation leadership. However, the transition from fundamental research to innovation ecosystems is not straightforward.
Fundamental research innovations are often not silver-bullet solutions due to the complexity of existing technological systems. Incumbent technologies benefit from system design tradeoffs made over many product generations that optimize for existing use cases. In contrast, introducing new innovations into highly complex systems can alter the tradeoff space, requiring additional work to manage the ripple effects of change throughout a complex system and ultimately approach a similar level of maturity and on-going optimization. For example, in-memory compute approaches can require clarity of fundamentals to separate unviable from viable solutions and, from there, rethink the traditional way compute and memory resources are coupled in complex system architecture. As researchers look to hand off innovations to the industrial sphere, it becomes important to focus on emerging use cases rather than trying to immediately replace mature technologies given the ensuing complex system management that integration requires. Innovations can struggle to prove value with existing use cases given the optimization advantage incumbent solutions enjoy. Positioning innovations to take on emerging use cases can level that playing field, allowing a clear demonstration of the benefits of new innovations.
A significant challenge in this ecosystem is the disconnect between fundamental research and venture capital. Investors may lack a deep understanding of the technological groundwork that indicates which innovations are truly promising. Meanwhile, insights gained through fundamental research can reveal clear limits and dead ends of certain potential solutions that investors may not be aware of. Therefore, fundamental researchers need to take a more active role in advising investment communities, helping to direct resources toward technologies with genuine potential and steering away from those with potentially problematic technological approaches. Such communication can lead to better, more efficient outcomes.
As an example of this fundamental-research and innovation-ecosystem hand-off, emerging use cases like AI-powered edge devices are promising pathways to bring innovative compute technologies out of the lab and into the competitive arena. These new markets provide opportunities for startups and researchers to introduce breakthrough compute architectures capable of delivering dramatically increased computing speed and efficiency without competing directly with established platforms of daunting complexity, such as the highly optimized data center space.
At this inflection point, industrial leadership can play an important role in supporting these innovation ecosystems. The fact remains that while research can derisk many aspects of semiconductor innovation, it remains limited in reducing the capital costs of chip design. Collaboration between venture capital partners who understand the capital requirements of the semiconductor space and industrial partners who have the scale and resources to realize new products is essential.
Looking forward, several strategies can facilitate better innovation transitions. First and foremost, it is critical to continue public sector support for fundamental research. History has proven time and again, relatively small public investments have been critical to incubating world-changing innovations that cemented U.S. global technology leadership over the last century, from semiconductors to internet technologies and beyond. The National Semiconductor Technology Center is a clear continuation of this beneficial relationship between government and fundamental researchers. In addition, creating pathways for researchers to engage with commercial partners, leveraging trusted industrial insiders as innovation champions, and focusing on emerging use cases are all important elements of continuing America’s innovation machine. The United States' strength in fundamental research and industrial innovation is unparalleled. By understanding the complex innovation ecosystem and creating intentional, collaborative pathways between research, startups, venture capital, and industrial leadership, the United States can continue to drive technological advancement and maintain its position as a global innovation leader at a critical time.
Listen to Naveen’s full keynote on how academia and industry can interlock to build a more cohesive innovation ecosystem and help the U.S. achieve sustained technological leadership.
Join our mailing list for exclusive updates, releases, and exciting news from EnCharge AI.