EnCharge AI, a startup specializing in semiconductor technology, has successfully secured over $100 million in a Series B funding round spearheaded by Tiger Global to accelerate its growth in developing analog memory chips for AI applications.
This funding is particularly noteworthy given the unprecedented interest in AI, even as the costs associated with developing and maintaining AI services remain a concerning issue. EnCharge, which originated from Princeton University, is confident that its analog memory chips—intended for integration into devices such as laptops, desktops, smartphones, and wearable tech—will not only enhance the speed of AI processing but also reduce overall expenses.
Based in Santa Clara, EnCharge asserts that its AI accelerators consume 20 times less energy for handling workloads compared to existing chips in the market, with expectations to launch its first products by the end of this year.
The timing of EnCharge’s funding is significant, as the U.S. government has identified hardware and infrastructure, including chip technology, as key sectors for fostering domestic innovation. A successful execution could position EnCharge as a vital player in this initiative.
This round, categorized as Series B, has been confirmed by the company. It is important to note that a funding tranche we reported in December 2023 was not included within this Series B round. There were indications of this funding back in May when Bloomberg reported that EnCharge was looking to raise at least an additional $70 million for business expansion.
In a conversation with TechCrunch, Naveen Verma, the CEO and co-founder of EnCharge, refrained from revealing the company’s valuation. Contrary to PitchBook’s reports, EnCharge clarified to TechCrunch that they did not raise funds in October at a $438 million post-money valuation.
While Verma did not disclose the identities of their clients, he indicated that the funding had attracted a diverse mix of strategic and financial investors, hinting at potential partnerships.
Besides Tiger Global, the funding round included participation from Maverick Silicon, Capital TEN (from Taiwan), SIP Global Partners, Zero Infinity Partners, CTBC VC, Vanderbilt University, and Morgan Creek Digital, along with returning investors such as RTX Ventures (the venture capital branch of an aerospace and defense contractor), Anzu Partners, Scout Ventures, AlleyCorp, ACVC, and S5V.
Corporate investors in this funding round include Samsung Ventures and HH-CTBC, a collaboration between Hon Hai Technology Group (Foxconn) and CTBC VC. The VentureTech Alliance had previously supported EnCharge, alongside notable investors like In-Q-Tel, the CIA-affiliated investment firm, and Constellation Technology, a clean energy manufacturer. The startup has also received grants from U.S. organizations, including DARPA and the Department of Defense.
Verma mentioned that EnCharge is collaborating closely with TSMC, which he had previously stated would be the manufacturer of its initial chips.
“TSMC has been tracking my research for many years,” he shared during an interview, indicating their involvement traced back to the formative stages of EnCharge’s research and development. “They’ve provided us access to highly advanced silicon, which is quite exceptional for them.”
Emphasis on Analog Technology
Focusing on analog technology, EnCharge is differentiating itself from its rivals. While attention has primarily been on the processing chips for training and AI inference at the server level—resulting in significant demand for GPU manufacturers like Nvidia and AMD—EnCharge presents a novel alternative.
The distinction in EnCharge’s strategy is elaborated upon in a recently published paper on analog chips by IBM’s research team. As noted in their research, there exists “no division between compute and memory, rendering these processors remarkably cost-effective compared to conventional designs.”
IBM, similar to EnCharge, concludes that while the physical properties of these chips enable them to be effective for inference, they are less suited for training applications. EnCharge’s chips are specifically designed to execute existing AI models “at the edge,” though both the startup and IBM continue to explore new algorithms to broaden their use cases.
EnCharge is not the only company pursuing analog technologies, but Verma emphasizes that one of their key advancements is the design of their chips, particularly in their noise resilience.
“When you have 100 billion transistors on a chip, each can produce noise, and they all need to function properly. Therefore, achieving signal separation is crucial, but you also miss out on efficiency by not capturing the intermediate signals between analog representations,” Verma explained. “Our significant breakthrough was how to design analog components that are resistant to noise.”
The company utilizes “a very precise component available through standard supply chains,” which Verma describes as a set of geometry-dependent metal wires that can be manipulated with great precision.
Verma also states that EnCharge operates with a full-stack model, having developed software to complement its hardware solutions.
Verma and his co-founders, COO Echere Iroaga and CTO Kailash Gopalakrishnan—former employees at semiconductor firms Macom and IBM—bring a wealth of expertise to EnCharge. However, it remains to be seen whether this foundation will be sufficient for EnCharge to maintain competitiveness in a saturated market. Other emerging competitors in the analog chip space include Mythic and Sagence.
“At Anzu, we have evaluated over 50 companies in the semiconductor field—potentially more than 50 since 2021,” stated Jimmy Kan, an investment partner at Anzu Partners, who previously worked with chips at Qualcomm.
“Among those, one in five had innovative architectures, such as analog or spiking neural network computation chips. Our goal is to identify an AI computing technology that is genuinely transformative rather than merely incremental or something Nvidia may release in the near future,” he added. “We are genuinely enthusiastic about the progress EnCharge has made.”
The emergence of EnCharge contrasts with the trends observed among many deep tech startups in recent years.
The technology boom of the past quarter-century has spurred significant venture capital, supporting startups aiming to replicate the success of major tech giants like Google, Microsoft, Apple, Meta, or Amazon. Consequently, this has led to a burgeoning number of startups entering the market.
This surge has manifested in a growing number of deep tech initiatives: entrepreneurs securing funding not for fully-fledged products but for compelling concepts not yet ready for market, with quantum computing standing as a prime example of such deep-tech endeavors.
EnCharge could easily have been grouped with these deep tech startups had it launched sooner from Princeton and taken a more discreet route toward establishing partnerships while refining its innovative chip technology.
However, the startup opted to delay its market entry. It wasn’t until 2022, nearly a decade after Verma and his team initiated their research at Princeton, that EnCharge emerged from stealth and began forging commercial partnerships alongside ongoing technological development.
“Certain innovations lend themselves to quick venture backing, but developing a fundamentally new technology encompasses many complexities that must be understood to mitigate risks since many of these ventures ultimately fail,” Verma remarked. “Once you accept venture funding, your priorities shift… It’s no longer about mastering the technology; it becomes predominantly customer-centric.”
TechCrunch offers an AI-focused newsletter. Subscribe here to receive it in your inbox every Wednesday.
Compiled by Techarena.au.
Fanpage: TechArena.au
Watch more about AI – Artificial Intelligence
