Startup EnCharge AI raised more than $100 million in a Series B funding round led by Tiger Global to bring more efficient and less expensive AI chips to the market, the company said on Thursday.
The company did not disclose details on valuation.
EnCharge AI develops analog chips that are integrated into semiconductors used for storage. These in-memory chips are designed for inference, a phase where AI models are utilized rather than trained.
While most AI inference chips are typically housed in vast server clusters within data centers, EnCharge AI's chips are designed for edge computing, being utilized in user-facing devices like laptops.
Their approach to embedding analog processing in memory chips allows their accelerators to perform AI tasks with up to 20 times less energy consumption compared to some of the leading AI chips, the company said.
Battery-powered devices like laptops and smartphones need efficient chips to process AI and analog chips situated inside semiconductors for memory are a viable solution, CEO Naveen Verma told Reuters.
"It turns out that these platforms can now really overcome many of the barriers in terms of cost and sustainability, but also in terms of privacy and security, which the enterprise and also a lot of consumer applications care very much about."
Groq, founded by a former Alphabet (GOOGL.O) chip engineer and Cerebras, are among the companies developing specialized chips for AI inference.
Other investors in the round included Samsung Electronics' (005930.KS) VC arm and HH-CTBC, a partnership between Taiwan's Foxconn (2317.TW) and CTBC Venture Capital.
