Bud Ecosystem, a generative AI research lab based in Technopark, Thiruvananthapuram, has launched Bud Runtime, a breakthrough product aimed at dramatically reducing the cost and complexity of deploying Generative AI (GenAI) applications. Targeted at startups, developers, enterprises, and research institutions, Bud Runtime enables GenAI deployment for as low as $200 per month, making the technology significantly more accessible.
Designed to address the dual challenge of high GPU costs and infrastructure scarcity, Bud Runtime enables efficient deployment on CPU-based systems, offering a sustainable and scalable alternative without compromising on performance. It supports a wide array of hardware including CPUs, GPUs, HPUs, TPUs, and NPUs from major vendors like Intel, AMD, Nvidia, and Huawei.
A standout feature of Bud Runtime is its heterogeneous cluster parallelism, allowing organizations to leverage mixed hardware environments. This enables seamless scaling of GenAI workloads using available compute resources, helping teams overcome GPU shortages and high infrastructure costs.
Bud Ecosystem’s CEO Jithin VG explained that the product evolved from early internal efforts to deploy GenAI on existing infrastructure and has matured into a robust, production-ready platform. Chief Scientific Officer Linson Joseph emphasized Bud’s mission to democratize GenAI by enabling its use on commodity hardware, supported by ongoing research in efficient transformer architectures and AI inference optimization.
Backed by collaborations with Intel, Microsoft, Infosys, and LTIMindtree, Bud Ecosystem has also been active in open-source AI research, releasing over 20 models and achieving top rankings on the Hugging Face LLM leaderboard.
Bud Ecosystem plans to further its mission with the release of a new open-source project in the coming weeks, reinforcing its commitment to accessible and affordable GenAI.