Nvidia backs Indian startup Simplismart in push for AI efficiency

Simplismart, founded in 2022 by former Oracle and Google engineers Amritanshu Jain and Devansh Ghatak, addresses a critical bottleneck in the current AI boom: the high cost and complexity of running models at scale. Its platform sits between AI models and underlying computing infrastructure, automatically adjusting deployment configurations to balance speed, performance, and cost. By tailoring how models run on existing hardware, the company claims it can significantly lower operational expenses, citing a 56% reduction in GPU costs for video-generation platform InVideo.

The investment creates a strategic alignment between the world’s leading chipmaker and a firm focused on hardware optimization. While the partnership might appear counterintuitive—as Simplismart’s software helps companies utilize fewer GPUs—Jain argues that increasing efficiency ultimately accelerates AI adoption, benefiting the broader hardware ecosystem. Beyond capital, Nvidia is collaborating with the startup on technical integrations for its Dynamo inference software and assisting with market outreach.

With an annualized revenue run rate that has increased 14-fold over the last six months, the 58-person team plans to utilize the fresh capital to scale operations in the United States, which currently accounts for over half of its revenue. Simplismart enters a competitive landscape that includes internal tools from major hyperscalers like Amazon, Google, and Microsoft, but positions itself as a specialized alternative for companies facing stringent service-level agreements and tight cost constraints.