Intel is ramping up its efforts in the small but fast-growing artificial intelligence space with the acquisition of 48-person startup Nervana Systems, which has been developing software and hardware technologies for machine learning.
Intel officials announced the chip maker’s plans to acquire the San Diego, Calif.-based company Aug. 9, a week before their Intel Developer Forum (IDF) kicks off next week in San Francisco. No financial details were released, though news site Recode cited anonymous sources who put the price tag at more than $350 million.
The acquisition strengthens Intel’s position in an artificial intelligence (AI) space that is expected to expand rapidly in the coming years as the internet of things (IoT) grows, with company officials predicting more than 50 billion smart, connected devices worldwide by 2020 and continuing to increase after that.
Intel’s Xeon processors run more than 90 percent of the servers in data centers globally, but the company is now looking to become a larger player in the areas of AI and machine learning. It’s a market that a broad array of tech vendors are making hard pushes into. That includes GPU maker Nvidia, which has made AI a central part of its future strategy, and other chip companies, such as IBM and its Power and OpenPower efforts and ARM and its range of chip-making partners.
AI technology already is in use today, from fraud detection and photo tagging to self-driving cars and precision medicine, according to Diane Bryant, executive vice president and general manager of Intel’s Data Center Group.
“Encompassing compute methods like advanced data analytics, computer vision, natural language processing and machine learning, artificial intelligence is transforming the way businesses operate and how people engage with the world,” Bryant wrote in a post on the company blog. “Machine learning, and its subset deep learning, are key methods for the expanding field of AI.”
Nervana already has a software and hardware stack that is fully optimized for deep learning, and the company’s “IP and expertise in accelerating deep learning algorithms will expand Intel’s capabilities in the field of AI,” she wrote.
“We will apply Nervana’s software expertise to further optimize the Intel Math Kernel Library and its integration into industry standard frameworks,” Bryant wrote. “Nervana’s Engine and silicon expertise will advance Intel’s AI portfolio and enhance the deep learning performance and TCO of our Intel Xeon and Intel Xeon Phi processors.”
Patrick Moorhead, principal analyst with Moor Insights and Strategy, said the acquisition of Nervana is an important move in a market that he said represents a crucial inflection point in the tech industry.
“It moves them in the right direction,” Moorhead told eWEEK. “I’m a lot more comfortable in their future in AI and machine learning, but there’s a lot of execution that needs to be done.”
There essentially are two key parts of machine learning: training (where neural networks are taught such things as object identification) and inference (where they use this training to recognize and process unknown inputs). Neural networks used for training are large, and most training is done on Nvidia GPUs. The inference networks are smaller, and most of that work is done on CPUs from Intel.
The chip maker wants to be a larger player in both areas, and the Nervana acquisition—combined with its Xeon, Xeon Phi and field-programmable gate array (FPGA) chips—gives them the chance to make that push, Moorhead said.
“That’s a benefit of being a big company,” he said. “You can make these massive investments.”
Intel Accelerates AI, Machine Learning Efforts with Nervana Acquisition
The opportunity is significant, according to Intel’s Bryant. The vendor’s Xeon E5 chip is the most widely used processor in systems for deep learning inference, and the latest Xeon Phi 7200 “Knights Landing” chips—which offer up to 72 cores—bring the scalability and performance needed for training, she wrote.
“While less than 10% of servers worldwide were deployed in support of machine learning last year, the capabilities and insights it enables makes machine learning the fastest growing form of AI,” Bryant wrote.
For Nervana, the deal means that the company can merge its growing AI and deep learning technologies with the resources of the world’s top chip maker, according to co-founder and CEO Naveen Rao.
“With this deal, we can now shatter the old paradigm and move into a new regime of computing,” Rao wrote in a post on the Nervana blog. “We’ll look back in 10 years and see this time as the inflection point of when compute architectures became neural. The semiconductor integrated circuit is one of humanity’s crowning achievements and Intel has the best semiconductor technology in the world. Nervana’s AI expertise combined with Intel’s capabilities and huge market reach will allow us to realize our vision and create something truly special.”
According to Rao, the company will keep its brand and workforce, and will continue to operate out of its San Diego headquarters.
Intel has been the dominant chip maker in both the PC market and the data center. However, while the data center business continues to grow, the global PC space has been contracting for several years, and Intel officials are working to reduce the company’s reliance on PCs and expanding into new growth areas, including the IoT and AI. Intel also has turned away from efforts in the mobile device space after several years of fruitless efforts to make inroads into a space dominated by ARM and its partners.