Nvidia backs Indian startup helping companies run AI models with fewer GPUs

Bengaluru-based Simplismart raised a $14 million Series B led by Dallas Venture Capital, with Nvidia, Accel, Shastra VC and Micromax participating. The round values the four-year-old company at about $100 million. Its software sits between AI models and the hardware they run on, tuning response speed, throughput and compute use so companies get more out of the GPUs they already have.

Simplismart says it helped AI video platform InVideo cut GPU serving costs by 56 percent, and it counts HeyGen, Higgsfield and Ema among its customers. The CEO says annualized revenue run rate grew 14-fold in six months to an eight-figure sum, based on recent usage rather than contracts. More than half of revenue comes from the U.S.

We think this is a smart bet, and the apparent paradox isn't really one. Nvidia backing software that lowers GPU needs makes sense if cheaper inference brings more AI applications into existence. Those applications still run on Nvidia hardware. It's the old efficiency story: make something cheaper to use, and people tend to use more of it.

The product idea is also right. CEO Amritanshu Jain's line that nobody actually needs the fastest or the cheapest inference gets at something true. Most businesses need a specific balance of speed and cost for a specific job, and tuning that by hand is hard, slow and expensive. Software that automates it makes AI more affordable for companies without big infrastructure teams, and that's useful.

We'd treat the numbers with some care. A 14-fold run-rate jump based on recent usage isn't the same as contracted recurring revenue, and the exact figure wasn't shared. The 56 percent savings comes from a single customer case. The market is also crowded: Nvidia itself, AWS, Google Cloud, Azure and many startups are all chasing the same inference optimization work. A small company in that field needs to keep proving it beats tools its own investor sells.

There's a broader point too. Gains like this can ease pressure on GPU supply and power. They don't reduce total compute demand if efficiency simply fuels more usage, and the record of efficiency gains in tech suggests that's the likelier outcome. Cheaper inference is good for builders. It isn't an answer to AI's energy footprint.

Inference is the work of running a trained model to answer requests, and as AI products scale, it has become a major share of compute spending. Simplismart was founded in 2022 by former Oracle engineer Amritanshu Jain and former Google engineer Devansh Ghatak, and has 58 employees. Nvidia has been working with the startup on technical integrations, including its Dynamo inference software.