Nvidia backs Simplismart, software that runs AI on fewer GPUs

Nvidia has put money into an Indian startup whose software helps companies run AI models with fewer GPUs. Simplismart, based in Bengaluru, raised $14 million in an all-primary Series B round that values the four-year-old company at roughly $100 million, including the new cash.

The round, led by Dallas Venture Capital, also included Accel, Shastra VC, and Indian electronics maker Micromax alongside Nvidia. The investment lands at a moment when companies are spending heavily on GPUs to power AI applications — and struggling to run those models reliably and economically at scale.

That is the market Simplismart targets. Its software sits between AI models and the computing infrastructure they run on, adjusting how models are deployed and managed to meet each customer's requirements. The company says it optimizes factors such as how quickly a model responds, how many requests it handles, and how much computing power it consumes — helping companies extract more from their existing GPU stacks.

What does the software actually do?

Deploying and operating AI models to meet business requirements often demands complex adjustments, and different workloads force distinct trade-offs between speed, cost, and performance. Simplismart says its platform automates much of that work, tailoring how models run on existing computing infrastructure instead of requiring customers to manage those configurations themselves.

The startup offers several deployment paths:

• Deploy and run AI models through Simplismart's own platform
• Run them within existing cloud or on-premises infrastructure
• Fine-tune models with built-in tools
• Automatically scale computing resources as demand changes
• Monitor model performance in operation

Simplismart also sells to cloud providers, letting them build and operate their own AI inference services (the stage where a trained model actually processes requests) on GPU infrastructure using its software. In short, the company spans both end users and the infrastructure layer beneath them.

Who uses it, and what are the numbers?

The customer list includes named companies with a headline figure attached. Simplismart says it helped AI video-generation platform InVideo cut its GPU serving costs by 56%. It also works with AI video companies HeyGen and Higgsfield, plus enterprise AI startup Ema, among others.

On the infrastructure side, Indian cloud provider Yotta uses Simplismart's technology to power the serverless AI inference service within its Shakti Studio platform.

The financial trajectory is the kind investors cite first. Simplismart's annualized revenue run rate (a projection based on recent usage, not contracted recurring revenue) has grown 14-fold over the past six months to an eight-figure sum, co-founder and CEO Amritanshu Jain told TechCrunch. The figure excludes revenue passed through to infrastructure providers, and Jain did not disclose the exact number — a caveat worth holding onto when weighing the growth claim.

More than half of Simplismart's revenue comes from the U.S., even though the startup has more customers in India, Jain said.

Why would Nvidia fund software that sells fewer GPUs?

The investment contains an apparent paradox: Nvidia, the dominant supplier of AI chips, is backing software that can help companies run some workloads with fewer GPUs. Jain argued the bet makes sense for the chipmaker anyway — making AI inference more efficient could ultimately benefit Nvidia by making its hardware more attractive to customers and encouraging wider adoption of AI applications.

The relationship goes beyond capital. Nvidia has been working with Simplismart on technical integrations, including around its Dynamo inference software, Jain said. The collaboration involves Simplismart's engineers and Nvidia's solutions architects, and Nvidia has also helped introduce the startup to potential customers and partners.

How does Simplismart survive a crowded market?

The startup operates in an increasingly crowded field for AI inference software. It competes with offerings from Nvidia itself, with major cloud providers and hyperscalers including Amazon Web Services, Google Cloud, and Microsoft Azure, and with a growing number of startups building tools to make AI models faster and cheaper to run.

Jain framed the differentiation around cost discipline and contractual performance guarantees. "Where we tend to win is where people have to be more stringent on their SLAs [service-level agreements] and TCO [total cost of ownership]," he said, referring to demanding performance requirements and tight cost constraints. Unlike rivals that offer preset options for faster or cheaper inference, Simplismart aims to find a balance tailored to each customer's needs.

The philosophy extends to his view of the market as a whole. "Nobody actually needs the fastest or the cheapest inference," Jain said in an interview. Companies instead need to run AI models at speeds and costs that make sense for their specific applications.

That positioning matters because the stakes here are straightforward: rising demand for AI applications is colliding with the need to make better use of expensive GPU resources, and whoever reliably lowers that cost of serving wins the enterprise budget.

Who built it, and what comes next?

Simplismart was founded in 2022 by Jain, a former Oracle engineer, and Devansh Ghatak, a former Google engineer. The two met while studying at BITS Pilani, and the startup grew out of their experience building and deploying machine-learning systems. The company now employs 58 people.

The new funding is earmarked for three purposes: expanding Simplismart's presence in the U.S., growing its team, and investing further in developing its inference software. Jain told TechCrunch that the U.S. remains a key market as the startup looks to build on its existing customer base there.

One loose end remains. The announcement follows a report by Indian English daily The Economic Times that Simplismart was in talks to raise $40 million to $50 million at a valuation of $300 million to $400 million. Asked about the report, Jain declined to comment, saying the startup was not actively fundraising. He added that the Series B was closed and would not be extended.

If that larger round ever materializes, it would signal that inference efficiency has moved from a cost-saving niche to a category investors price in the hundreds of millions. For now, a $100 million valuation on a usage-based eight-figure run rate suggests the market is rewarding traction — but still pricing in the risk.