Nvidia backs Simplismart's $14M round for AI inference software that cuts GPU needs

What happened

Nvidia has invested in Simplismart, a four-year-old Bengaluru startup that builds software for running AI models more efficiently on existing hardware. According to TechCrunch, the company raised $14 million in an all-primary Series B led by Dallas Venture Capital, with Accel, Shastra VC, Micromax and Nvidia also participating. The deal values the startup at roughly $100 million including the new funding.

The round lands as companies spend heavily on GPUs to power AI applications, then struggle to run those models reliably and economically at scale, where different workloads demand distinct trade-offs between speed, cost and performance.

What the software does

Simplismart's platform sits between AI models and the computing infrastructure serving them, automatically tuning how models are deployed and managed to match customer requirements. It optimizes factors such as how quickly a model responds, how many requests it handles and how much compute it consumes, letting companies get more from the GPU fleets they already own.

"Nobody actually needs the fastest or the cheapest inference," co-founder and CEO Amritanshu Jain told TechCrunch, arguing that companies instead need models running at speeds and costs suited to their specific applications.

Customers can deploy and run models through Simplismart's own platform or inside their existing cloud and on-premises environments. The product also covers fine-tuning, automatic scaling of compute as demand changes, and performance monitoring. Cloud providers can additionally use the software to build and operate their own AI inference services on GPU infrastructure.

Customers and growth

TechCrunch reports that Simplismart helped AI video-generation platform InVideo cut its GPU serving costs by 56%. Other customers include AI video companies HeyGen and Higgsfield and enterprise AI startup Ema. Indian cloud provider Yotta uses the technology to power the serverless inference service in its Shakti Studio platform.

Jain told TechCrunch that Simplismart's annualized revenue run rate has grown 14-fold over the past six months to an eight-figure sum. He did not disclose the exact figure, and the company noted the number reflects recent customer usage rather than contracted recurring revenue and excludes revenue passed through to infrastructure providers. More than half of revenue comes from the U.S., even though the startup has more customers in India.

Nvidia's paradoxical bet

The investment carries an apparent tension: Nvidia, the dominant supplier of AI chips, is backing software that in some cases lets companies run workloads with fewer GPUs. Jain argued the opposite effect could prevail, saying cheaper and more efficient inference makes Nvidia's hardware more attractive to customers and encourages broader AI adoption, ultimately benefiting the chipmaker.

The relationship goes beyond capital. Jain told TechCrunch the two companies have been working on technical integrations, including around Nvidia's Dynamo inference software, with Simplismart's engineers collaborating with Nvidia's solutions architects. Nvidia has also introduced the startup to potential customers and partners.

A crowded market

Simplismart operates in an increasingly crowded field for AI inference software, competing with offerings from Nvidia itself as well as Amazon Web Services, Google Cloud and Microsoft Azure, alongside a wave of startups building tools to make models faster and cheaper to run. Jain said Simplismart tends to win where customers face stringent service-level agreements and tight total-cost-of-ownership constraints, rather than offering preset options for faster or cheaper inference.

The company was founded in 2022 by Jain, a former Oracle engineer, and Devansh Ghatak, a former Google engineer, who met while studying at BITS Pilani. It now employs 58 people.

Simplismart plans to use the funding to expand its U.S. presence, grow its team and further develop its inference software. The announcement follows a report in Indian daily The Economic Times that the startup was in talks to raise $40 million to $50 million at a valuation of $300 million to $400 million; Jain declined to comment on that report, telling TechCrunch the company is not actively fundraising and that the Series B is closed and will not be extended.

Why it matters

Inference is becoming the dominant recurring cost of AI, and GPU capacity remains a bottleneck for everything from video generation to enterprise agents. Software that squeezes more requests out of each GPU changes the unit economics of every AI product built on top, which is why hyperscalers and a long tail of startups are converging on the same problem. Nvidia's participation is the notable signal here: the chipmaker is effectively hedging that better efficiency per GPU will expand total AI usage rather than shrink chip demand, and it is funding a startup that builds on its own software stack, including Dynamo. The gap between the previously reported raise and the final, smaller round is also a reminder that early funding reports and closed terms often diverge sharply.