News
Mar 3, 2026
Hyundai Motor Group Invests $6.7 Billion in Global Robotics, AI, and Hydrogen Innovation Hub
Success Stories
NewDecoded
4 min read
Turiyam, a Bengaluru-based deep-tech startup, has successfully raised $4 million in Pre-Seed funding to accelerate the development of its high-efficiency AI inference platform. Led by Ankur Capital with participation from Micelio Fund, the investment marks a significant step for India's semiconductor ambitions. The company aims to provide a radically lower Total Cost of Ownership for enterprises deploying generative AI models. Unlike traditional manufacturers focused on model training, Turiyam targets the rapidly expanding inference market. The startup is building a full-stack solution that includes custom RISC-V hardware and a proprietary, CUDA-free software layer. This approach allows developers to bypass expensive, supply-constrained GPU ecosystems while maintaining high performance across various workflows.
At the heart of the Turiyam technology is a hybrid memory architecture designed to solve the memory wall bottleneck in large language models. By combining SRAM and High Bandwidth Memory, their silicon targets a tenfold improvement in performance per dollar compared to current industry standards. This efficiency is critical for high-intensity workloads like real-time video analytics and large-scale audio processing. The company was founded by semiconductor veterans including CEO Sanchayan Sinha, formerly of Groq, and CPO Parag Jain. Their combined experience at industry giants like Nvidia, AMD, and Broadcom provides a strong foundation for building complex silicon from the ground up. The team is currently headquartered in Bengaluru with an additional presence in the San Francisco Bay Area.
Looking ahead, Turiyam plans to use the new capital to expand its engineering team and move toward early pilot deployments. Following the successful showcase of its prototype hardware at the India AI Impact Expo 2026, the company is now testing its solution in real-world data center environments. Enterprises can expect the first commercial-grade accelerators to emerge as the company scales its production pipeline in the near future. Turiyam's mission draws inspiration from India's successful digital public infrastructure like UPI and Aadhaar. By building an open platform, they intend to make high-performance intelligence accessible to billions of people rather than a select few. This vision positions them as a key player in the global movement toward sovereign AI capabilities that reduce reliance on proprietary stacks.
The rise of Turiyam signals a critical shift in the global AI landscape from training-centric dominance to inference-led efficiency. As the industry moves from building massive models to deploying them at scale, the reliance on expensive, proprietary hardware becomes a primary barrier to growth. By leveraging RISC-V and avoiding the CUDA ecosystem, Turiyam is not just competing on price but is providing a roadmap for hardware sovereignty that bypasses traditional bottlenecks. Their success would validate the potential to export deep-tech infrastructure and could fundamentally alter the economics of the token economy by lowering entry barriers for enterprises globally.
Related Articles