Bengaluru: Semiconductor start-up SandLogic is in early talks to lift $30-40 million in Collection A funding at a valuation of $200 million, two folks conscious of the matter mentioned.
Bengaluru-based SandLogic is amongst India’s few chip builders focusing on commercial-grade synthetic intelligence (AI) functions. The corporate has already appointed an funding banker, and home and overseas enterprise capitalists have been approached, one of many two folks mentioned on the situation of anonymity.
Based in 2018 by Kamalakar Devaki, Jesudas Fernandes, Radhika Kanigiri and Ravi Kumar Rayana, the Texas- and Bengaluru-based firm began as a low-code/no-code platform, earlier than getting into enterprise AI and designing its personal chip prototypes. Its low-power Krsna chip goals to facilitate on-device AI and has already achieved 22 trillion of operations per second (TOPS).
Devaki declined to touch upon the funding spherical.
AI chip start-ups have seen rising investor curiosity this 12 months. Whereas Kerala-based Netrasemi raised about ₹107 crore in Collection A funding led by Zoho and Unicorn India Ventures, Bengaluru’s Maieutic Semiconductors secured roughly $4.15 million seed cash from Endiya Companions and Exfinity. SandLogic has beforehand raised round $3.5 million from high-net-worth people and some angel traders, together with Google and AT&T executives.
ExSLerate V2 is SandLogic’s underlying chip design that may be reused and scaled. It’s meant to run AI duties sooner whereas utilizing lower than 2W of energy, and the corporate says it will likely be prepared for industrial use by December 2026. Krsna chip is constructed on that design and is now in prototyping, with check chips deliberate for the third quarter of FY27, Devaki mentioned. It’s geared toward AI functions like picture, language and audio processing, and is designed to be used in small units with out heating up.
“These chips can provide extra battery life for small units after which nonetheless run AI algorithms on these units resembling good watches, infotainment units within the automobiles, routers at dwelling, TVs, and telephones,” Devaki mentioned.
Semiconductor bets rise
Investor curiosity spans deep-tech specialists resembling Speciale Make investments, which has closed a ₹600 crore fund to again early-stage deep-tech start-ups together with AI and semiconductors, and world traders like Celesta Capital which have been energetic throughout semiconductor and AI infrastructure bets.
Different traders embrace enterprise capital (VC) funds like Endiya Companions, Unicorn India Ventures and 3one4 Capital which can be actively writing seed-to-Collection A cheques in AI-led start-ups. Amongst overseas VCs, Celesta Capital stays energetic throughout semiconductor and AI infrastructure, whereas bigger multi-stage VCs resembling Peak XV Companions, Accel, Lightspeed and Elevation Capital have been main or collaborating in new AI and agentic AI rounds this 12 months.
Shakti is SandLogic’s in-house giant language mannequin (LLM) constructed for enterprise use, with variations launched until 4 billion parameters and eight billion parameters beneath progress, in keeping with Devaki. In easy phrases, parameters are the dials inside an AI mannequin that it learns to tune; extra parameters usually permit the mannequin to deal with extra complicated duties.
SandLogic has been promoting an “agent AI” layer on prime of those fashions to companies. “We began giving it to banks and healthcare corporations and manufacturing corporations,” Devaki mentioned.
