Intel Acquires AI Chip Startup Habana Labs For $2B

'This acquisition advances our AI strategy, which is to provide customers with solutions to fit every performance need – from the intelligent edge to the data center,' Intel's Navin Shenoy says of the chipmaker's latest acquisition.

Intel announced Monday it has acquired artificial intelligence chip startup Habana Labs for approximately $2 billion, expanding its arsenal of AI accelerator products for the data center market.

The Santa Clara, Calif.-based company's announcement came nearly two weeks after multiple reports stated that the semiconductor giant was in talks to acquire the Israeli startup, which has introduced programmable deep learning accelerators over the past year and raised a total of $120 million in capital from investors, including Intel's venture capital arm, Intel Capital.

[Related: Intel AI Exec Gadi Singer: Partners 'Crucial' To Nervana NNP-I's Success]

The move comes as AI computing becomes a central focus for Intel, who said that the AI silicon market is expected to reach more than $25 billion by 2024. The company has said that it expects its AI products to generate $3.5 billion this year, more than triple the $1 billion figure Intel reported last year.

"This acquisition advances our AI strategy, which is to provide customers with solutions to fit every performance need – from the intelligent edge to the data center," Navin Shenoy, executive vice president and general manager of Intel's Data Platforms Group, said in a statement. "More specifically, Habana turbo-charges our AI offerings for the data center with a high-performance training processor family and a standards-based programming environment to address evolving AI workloads."

Habana Labs, a fabless semiconductor company, has introduced two products in the past year: the Gaudi training processors, which the startup has said can outperform GPU training workloads nearly four times over, and the Goya inference processors, which it has claimed can provide one to three times better performance than data center solutions commonly used today.

Intel said Habana Labs will be an independent business unit led by its current management team within Intel's Data Platforms Group, where the chipmaker's other AI data center products are situated. The startup's chairman, Avigdor Willenz, has agreed to stay on as a senior advisor to Habana Labs.

"We have been fortunate to get to know and collaborate with Intel given its investment in Habana, and we’re thrilled to be officially joining the team," Habana Labs CEO David Dahan said in a statement. "Intel has created a world-class AI team and capability. We are excited to partner with Intel to accelerate and scale our business. Together, we will deliver our customers more AI innovation, faster."

The acquisition comes a little over a month after Intel launched its Nervana neural network processors for deep learning training and inference. At a launch event, the chipmaker said its NNP-I1000 inference processors provide nearly four times the compute density of Nvidia's T4 GPUs while its NNP-T1000 training processors offer 95 percent scaling capabilities on important training models.

Barry Speigel, business development manager at Ace Computers, an Elk Grove Village, Ill.-based partner and No. 261 on CRN's 2019 Solution Provider 500 list, said Intel's latest acquisition shows that the company has identified AI as a core competency and "they're doing something about it."

However, he cautioned, it's still early when it comes to the broader data center market.

"I think it's going to be a little while until they're ready for prime time," Speigel said.

With Intel now owning two separate product lines for deep learning training and inference, it shows that the semiconductor giant "wants to make sure they can get it right," according to Speigel.

"I think that's what they've learned. Maybe taking one track isn't enough," he said.