Intel has been doubling down on building chips and related architecture for the next generation of computing, and today it announced an acquisition that will bolster its expertise and work specifically in one area of future technology: artificial intelligence.
The semiconductor giant today announced that it has acquired SigOpt, a startup out of San Francisco that has built an optimization platform that can be used to run modeling and simulations (two key applications of AI tech) in a better way. Anthony described SigOpt as a startup built to “optimize everything” when we covered its Series A, but Intel specifically will be integrating the tech into its AI business, specifically into its AI Analytics Toolkit, a spokesperson tells me.
Terms of the deal were not disclosed, but SigOpt already counted a number of large enterprises — “SigOpt’s customer base includes Fortune 500 companies across industries, as well as leading research institutions, universities and consortiums using its products” — among its customers. The product was still in a closed beta, however. Notably, it had raised money from an interesting group of investors that included In-Q-Tel (the firm associated with the CIA that makes strategic investments) and Andreessen Horowitz, and Y Combinator, among others. It had raised less than $10 million.
The plan will be to continue providing services to existing users, and to continue building out the company’s platform — co-founders Scott Clark (CEO) and Patrick Hayes (CTO) and their team are joining Intel.
“We will continue to work with SigOpt’s existing customers and will also integrate the technology into our product road map,” a spokesperson confirmed.
While Intel is working hard on streamlining its business around next-generation chips to better compete against the likes of Nvidia (which itself is growing substantially with the acquisition of ARM) and smaller players like GraphCore, in part by divesting more legacy operations, it seems a strong opportunity in providing services for its customers alongside those chips, and these services specifically will help customers with the compute loads that they will be running on those chips.
The focus for Intel has been on the next generation of computing to offset declines in its legacy operations. In the last quarter, even as it beat expectations, Intel reported a 3% decline in its revenues, led by a drop in its data center business. It said that it’s projecting the AI silicon market to be bigger than $25 billion by 2024, with AI silicon in the data center to be greater than $10 billion in that period.
In 2019, Intel reported some $3.8 billion in AI-driven revenue, but it hopes that tools like SigOpt’s will help drive more activity in that business, dovetailing with the push for more AI applications in a wider range of businesses.
“In the new intelligence era, AI is driving the compute needs of the future. It is even more important for software to automatically extract the best compute performance while scaling AI models,” said Raja Koduri, Intel’s chief architect and senior vice president of its discrete graphics division. “SigOpt’s AI software platform and data science talent will augment Intel software, architecture, product offerings and teams, and provide us with valuable customer insights. We welcome the SigOpt team and its customers to the Intel family.”
Intel has invested $132M in 11 startups this year, on track for $300M-$500M in total
While there could potentially be a number of applications for SigOpt’s tech, this is a signal of how bigger players will continue to consolidate specific services around their bigger business, giving the small startup a much bigger horizon in terms of potential business (even if it is all tied to customers that only use Intel hardware).
“We are excited to join Intel and supercharge our mission to accelerate and amplify the impact of modelers everywhere. By combining our AI optimization software with Intel’s decades-long leadership in AI computing and machine learning performance, we will be able to unlock entirely new AI capabilities for modelers,” said Clark in a statement.
Comment