Nvidia Backs Simplismart in $14M Round

Surging AI demand continues to strain global data centers. In response, Nvidia has backed Bengaluru startup Simplismart in a $14 million Series B funding round.

The startup builds software that lets teams run AI models faster and cheaper. In many cases, the tools allow companies to use fewer graphics processing units (GPUs).

Dallas Venture Capital led the all-primary round. Accel, Shastra VC, Micromax, and Nvidia also joined. The deal values the four-year-old company at about $100 million.

Companies spend heavily on GPUs to run AI apps. Yet running models reliably and at low cost remains difficult. Every workload requires different trade-offs between speed, cost, and throughput. Simplismart automates this tuning work. Its software configures how models run on existing hardware so developers do not have to manage setups manually.

Factsheet: Simplismart Series B and Technology Profile

The Shifting Economics of AI Inference

The early AI boom focused on model training. Tech giants raced to build massive clusters with thousands of linked GPUs. Today, the focus is shifting to inference. Inference means running models in real products for end users.

Training is a one-time batch job. Inference runs all day. Every prompt, image, and API call costs money right away.

“Nobody actually needs the fastest or the cheapest inference,” said Amritanshu Jain, co-founder and chief executive of Simplismart. He noted that businesses need speeds and costs that fit their specific apps.

A voice bot needs quick responses to feel natural. A document summary tool can wait several seconds if that cuts server costs. Finding the right balance used to take a team of engineers weeks of manual testing. Simplismart does this automatically. Its software sits between the model and the chip to adjust settings on the fly.

The Nvidia Paradox: Why Chip Makers Back Efficiency Tools

Nvidia dominates the market for AI chips. It earns record profits selling top-tier GPUs like the H100 and Blackwell. Backing software that helps customers use fewer chips might seem odd at first glance.

Yet high inference costs remain a major barrier to widespread AI adoption. When cloud bills climb too high, companies pause new rollouts or cut back features. In broader market trends, investors are already exploring how to hedge computing risk through platforms like Hyperbolic tradable GPU compute futures as chip expenses mount.

By making inference cheaper, Simplismart helps more businesses launch AI products. As more companies deploy viable tools, overall demand for reliable GPU hardware grows.

“Where we tend to win is where people have to be more stringent on their SLAs and TCO,” Jain said. He pointed to strict service-level agreements and total cost of ownership. Unlike competing tools that offer rigid presets for speed or price, Simplismart tailors setups to each client.

The deal goes beyond venture funding. Simplismart engineers work closely with Nvidia solutions architects. Together, they optimize workloads around Nvidia Dynamo inference software. Nvidia also introduces the startup to potential enterprise clients and partners.

Inside Simplismart’s Middleware Stack

Amritanshu Jain and Devansh Ghatak founded Simplismart in 2022. Jain worked as an engineer at Oracle, and Ghatak worked at Google. The two met while studying at BITS Pilani. They launched Simplismart after seeing firsthand how hard it was to run production machine learning systems.

The platform provides a complete set of serving tools:

• Dynamic Scaling: The software provisions, scales, and shuts down GPU resources as incoming traffic shifts.
• Fine-Tuning Workflows: Teams can adapt open-weight models to private datasets without rebuilding serving setups.
• Flexible Deployment: The software runs on Simplismart’s managed cloud, private enterprise clouds, or bare-metal servers.
• Cloud Provider Tools: Regional data centers can use the software to launch their own serverless inference products.

Indian cloud operator Yotta Data Services is a key partner. Yotta uses Simplismart’s engine inside its Shakti Studio platform. This setup gives developers turnkey serverless inference on sovereign GPU clusters.

Real-World Benchmarks and Customer Adoption

Generative video is a tough test for inference software. Video models demand heavy memory bandwidth and long processing cycles for every frame.

Simplismart proved its value with InVideo, a generative video creation platform. By reordering batch schedules and tuning system memory, Simplismart helped InVideo cut GPU serving costs by 56%. The platform achieved these savings while keeping render times fast.

The startup also serves other major video and automation teams. Its customer roster includes HeyGen, Higgsfield, and enterprise AI assistant builder Ema.

These deployments have fueled rapid revenue growth. Over the past six months, Simplismart’s annualized revenue run rate surged 14-fold into eight figures. The figure reflects direct platform usage and excludes pass-through infrastructure expenses.

India accounts for the majority of Simplismart’s user accounts. Yet more than half of its total revenue now comes from United States enterprises. US firms facing high cloud bills have moved quickly to adopt inference tools.

Addressing Valuation Rumors and Looking Ahead

The Series B round follows media reports in India. The Economic Times recently reported that Simplismart was in talks to raise $40 million to $50 million at a valuation above $300 million.

Jain addressed the report directly. He confirmed that the startup is not actively fundraising. The $14 million Series B round is closed and will not be extended.

Simplismart employs 58 people across engineering, product, and sales. It plans to use the fresh funds to expand its sales and support footprint in North America. Winning US enterprise accounts is a core priority.

The team will also invest in new inference research. Key focus areas include speculative decoding, multi-tenant model sharing, and automated kernel compilation for new chips.

Frequently Asked Questions

What is Simplismart, and what does its software do?

Simplismart is an AI infrastructure startup that optimizes machine learning inference. Its middleware sits between AI models and GPUs, adjusting settings to reduce latency, boost throughput, and cut computing costs.

Why is Nvidia backing software that reduces GPU usage?

High inference costs hold back enterprise AI adoption. By making inference more economical, Simplismart broadens the market for AI tools, which lifts long-term demand for Nvidia data center hardware.

Which companies use Simplismart today?

Simplismart powers inference for InVideo, HeyGen, Higgsfield, Ema, and Indian cloud provider Yotta Data Services. InVideo cut its GPU serving costs by 56% using the platform.

How much funding has Simplismart raised in this round?

Simplismart raised a $14 million Series B round led by Dallas Venture Capital, with backing from Nvidia, Accel, Shastra VC, and Micromax. The round values the company at about $100 million.

For deeper strategies on selling AI software to major corporations, read our breakdown of the Databricks playbook on how early-stage AI startups land Fortune 500 contracts.