Bengaluru-based semiconductor start-up SandLogic is currently engaged in early discussions to secure $30-40 million in Series A funding, targeting a valuation of $200 million. This strategic move highlights the growing investor appetite for Indian deep-tech companies focused on commercial-grade artificial intelligence (AI) applications. SandLogic, recognized as one of India's few indigenous chip builders in this specialized domain, has already appointed an investment banker and has initiated outreach to both domestic and foreign venture capitalists.

Founded in 2018 by Kamalakar Devaki, Jesudas Fernandes, Radhika Kanigiri, and Ravi Kumar Rayana, SandLogic operates from both Texas and Bengaluru. The company embarked on its journey as a low-code/no-code platform before strategically pivoting into enterprise AI solutions and the sophisticated design of its own chip prototypes. Prior to this Series A round, SandLogic successfully raised approximately $3.5 million from high-net-worth individuals and a consortium of angel investors, including prominent executives from Google and AT&T.

At the core of SandLogic's hardware innovation is the Krsna chip, a low-power semiconductor designed specifically to facilitate on-device AI processing. This chip has already demonstrated remarkable capabilities, achieving 22 trillion operations per second (TOPS), positioning it as a significant contender in the edge AI market. The Krsna chip is built upon SandLogic’s foundational ExSLerate V2 chip design, which is engineered for reusability and scalability. ExSLerate V2 aims to accelerate AI tasks while consuming less than 2 watts of power, with the company projecting its commercial readiness by December 2026. Test chips for the Krsna are slated for the third quarter of FY27, with applications spanning image, language, and audio processing in compact devices without generating excessive heat. According to Kamalakar Devaki, these chips are poised to extend battery life for small devices and enable the execution of complex AI algorithms on platforms such as smartwatches, automotive infotainment systems, home routers, televisions, and mobile phones.

Beyond its hardware ventures, SandLogic also develops software solutions, notably its in-house large language model (LLM) named Shakti. Developed for enterprise applications, Shakti has seen versions released with up to 4 billion parameters, and an 8 billion-parameter version is currently under development. These parameters are crucial for an AI model's ability to learn and handle complex tasks, with a higher count generally correlating to greater sophistication. SandLogic further augments its offerings by providing an "agent AI" layer atop these models, which has been successfully deployed to businesses in sectors like banking, healthcare, and manufacturing.

The increasing investor interest in AI chip start-ups is a noticeable trend this year, with significant funding rounds for other Indian players underscoring the sector's vitality. Kerala-based Netrasemi recently secured approximately ₹107 crore in Series A funding, spearheaded by Zoho and Unicorn India Ventures, while Bengaluru's Maieutic Semiconductors raised about $4.15 million in seed capital from Endiya Partners and Exfinity. This burgeoning ecosystem attracts a diverse range of investors, from deep-tech specialists like Speciale Invest, which has established a ₹600 crore fund for early-stage deep-tech ventures including AI and semiconductors, to global players such as Celesta Capital, known for its active involvement in semiconductor and AI infrastructure investments. Additionally, venture capital funds like Endiya Partners, Unicorn India Ventures, and 3one4 Capital are actively issuing seed-to-Series A cheques to AI-led start-ups, while larger multi-stage VCs including Peak XV Partners, Accel, Lightspeed, and Elevation Capital are leading or participating in new AI and agentic AI funding rounds.