Intel announced on Thursday its acquisition of a startup company focused on software optimization as it diverts from chip making to artificial intelligence technology.
Intel said it has acquired the San Francisco, California-based startup SigOpt as the semiconductor giant plans to use its optimization platform to run simulations and modeling applications.
Intel senior vice president of graphics division and chief architect Raja Koduri said that AI currently drives the computing needs in the future. "It is even more important for software to automatically extract the best compute performance while scaling AI models," Koduri said in a statement posted on Intel website.
The chief architect also welcomes SigOpt to the Intel family as the startup's AI software and data science capabilities will provide Intel with valuable customer insights as well as boost its architecture, software as well as product offerings and teams.
SigOpt on optimizing everything
SigOpt is a startup designed to "optimize everything," although Intel vows to integrate its technology on the giant's AI business, including its Analytics Toolkit. Despite the buy-out and CEO Scott Clark and CTO Patrick Hayes, co-founders of the startup, would join Intel, SigOpt will continue supporting its existing users and strengthening the platform.
While terms of the SigOpt-Intel deal were not provided, among SigOp's clients are included in Fortune 500 companies as well as leading universities, consortiums, and research institutions. While the product was still in a closed beta, it was able to gather almost $10 million money from a group of investors such as In-Q-Tel, Y Combinator, and Andreessen Horowitz.
"We are excited to join Intel and supercharge our mission to accelerate and amplify the impact of modelers everywhere," Clark noted in an articlepublished in Intel website. The company also gears up to combining their AI optimization software with Intel's decades experience in machine learning and AI computing. "We will be able to unlock entirely new AI capabilities for modelers," Clark added.
From chip-making to artificial intelligence
The acquisition is Intel's response to setbacks it recently faced as it pushed back the launch of its 10-nanometer processor hardware as well as the 7-nanometer manufacturing process. In February, the company ended its Nervana AI model training program after acquiring chipmaker Habana Labs for $2 billion.
While Intel is restructuring its chip-making business to better compete against NVIDIA, which recently acquired ARM as well as smaller players such as GraphCore. This provides a great opportunity for its chip customers as these services will help them with the computing loads that they will run on those chips.
In 2019, Intel reported around $3.8 billion revenue in AI, but it hopes SigOpt and other tools will boost its business activities to create more AI applications for wider range of clients. This is Intel's way of coping up with the recent declines in its legacy operations.
In the 2020's third quarter report, Intel reported a 3% drop in revenues, which is hugely affected by its data center business. The company is now focusing on its AI silicon market projecting, which is expected to reach more than $25 billion in 2024.
This is owned by Tech Times
Written by CJ Robles
