Highlights –

  • The firm is on the lookout for mathematicians and is developing tools that provide a detailed view of how their user’s code stands out, interacting with the CPUs and GPUs at a much deeper level.

Strong Compute, a Machine Learning (ML) training platform that focuses on ultra-high-performance and efficiency, announced that it had raised USD7.8 million in seed round funding. The round was funded by the likes of Sequoia Capital India, Blackbird, Folklore and Skip Capital, Y Combinator, Starburst Ventures and founders and engineers from companies like Cruise, Waymo, Open AI, SpaceX and Virgin Galactic.

The Australia-based company that was part of Y Combinator’s Winter ’22 batch guarantees that its optimizations could spur the training process by 10x to 1,000x, subject to the model, pipeline, and framework. Strong Compute founder Ben Sand said that the team has recently made some breakthroughs where it was able to make Nvidia’s reference implementation, which its customer LayerJot used, run 20 times faster.

“That was a big win,” Sand said. “It gave us the sense that there is nothing that can’t be improved.” Though not wanting to reveal all the details about the team’s working on the optimizations, he said that the firm is on the lookout for mathematicians and is developing tools that provide a detailed view of how their user’s code stands out, interacting with the CPUs and GPUs at a much deeper level.

Sand stressed that the company’s current focus is to automate much of its current work to optimize the training process — this is something that can be tackled by the company easily, thanks to this funding round. “Our goal now is to have serious development partners in self-driving, medical and aerial, looking at what is going to generalize well,” he explained. “We’ve now got the resources to have an R&D team that doesn’t have to deliver something in a two-week sprint, but that can look at what’s some real core tech that could take a year to get a win out of, but that can help with that automated analysis of the problem.”

With a current strength of six full-time engineers, Sands said that he plans to double that in the next few months. In part, that’s also possible because the firm is now receiving inbound interest from big companies that often spend USD50 million or more on their compute resources (Sands noted that the market is primarily bimodal, with users either spending less than USD1 million or USD10 million to USD100 million, with only a few players in the middle).

Every company building ML models faces the same problem: Training models and running experiments to improve them still takes time. This means that data scientists trying to solve these issues spend maximum time in a holding pattern, waiting for the results. “Strong Compute is solving the basketball court problem,” said SteadyMD CFO Nikhil Abraham. “Long training times had our best devs shooting hoops all day, waiting on machines.”

Though some of that inbound interest is being generated from the financial industry and firms who want to optimize their natural language processing models, Strong Compute is focusing itself on computer vision for the time being.

“We’ve only just scratched the surface of what machine learning and AI can do.” said Folklore partner Tanisha Banaszczyk. “We love working with founders who have long-range ambition and visions that will endure across generations. Having invested in autonomous driving, we know how important speed to market is — and see the impact Strong Compute can have on this market with its purpose-built platform, deep understanding of the usd500 billion markets, and a world-class team.”