Strong computing power, a Sydney, Australia-based startup helping developers remove bottlenecks in their machine learning training pipelines, announced today that it has raised a $7.8 million seed round. The round includes a total of 30 funds and angels, including Sequoia Capital India, Blackbird, Folklore and Skip Capital, as well as Y Combinator, Starburst Ventures and founders and engineers from companies such as Cruise, Waymo, Open AI, SpaceX and Virgo Galactic.
The company, which was part of Y Combinator’s Winter ’22 batch, promises that its optimizations can speed up the training process by 10x to 1000x, depending on the model, pipeline and framework. As Strong Compute founder Ben Sands, who previously co-founded AR company metatold me that the team recently made some breakthroughs where it could use Nvidia’s reference implementation, which its customer LowJot used to run 20 times faster.
“That was a big win,” Sands said. “It really made us feel like there’s nothing that can’t be improved.” He didn’t want to reveal all the details about how the team’s optimizations worked, but noted that the company now hires mathematicians and builds tools that give it a more granular view of how their user’s code interacts with the CPUs and GPUs on a much wider scale. deeper level than previously possible.
As Sands emphasized, the current focus for the company is on automating much of the current work to optimize the training process — and that’s something the company can now address, thanks to this funding round. “Our goal now is to have serious development partners in the areas of self-driving, medical and aerial imaging, to look at what’s really going to generalize well,” he explained. “We now have the resources to have an R&D team that doesn’t have to deliver something in a two-week sprint, but can actually look at what is some real core technology that can take a year to actually achieve a win.” but that can really help with that automated analysis of the problem.”
The company currently has six full-time engineers, but Sands plans to double that number in the coming months. In part, that’s also because the company is now getting inbound interest from large companies that often spend $50 million or more on their computing resources (and Sands noted that the market is essentially bimodal, with customers either paying less than $1 million or $500 million). 10 to $100 million, with just a few players in the middle).
However, every company trying to build ML models faces the same problem: training models and conducting experiments to improve them still takes a lot of time. That means the highly paid data scientists working on these issues spend a lot of time in a waiting pattern, waiting for the results.”Powerful Calculate solves the problem of the basketball court,” said SteadyMD CFO Nikhil Abraham. “Long training times had our best developers firing hoops all day, waiting for machines.”
And while some of that incoming interest has come from the financial industry and companies looking to optimize their natural language processing models, Strong Compute’s focus remains on computer vision for now.
“We’ve only just scratched the surface of what machine learning and AI can do.” said Folklore partner Tanisha Banaszcyk. “We enjoy working with founders with long-term ambition and visions that will last for generations. Because we’ve invested in autonomous driving, we know the importance of speed in the marketplace — and see the impact Strong Compute can have in this market with its purpose-built platform, deep understanding of the $500 billion market, and world-class team. ”