IBM Corp. is acquiring Turbonomic Inc., a Boston-based company with a platform that helps enterprises reduce cloud expenses and improve their applications’ performance. 

IBM didn’t disclose the value of the deal, which was announced this morning. But sources told Reuters that the transaction values Turbonomic at between $1.5 billion and $2 billion. That’s a significant premium over the $963 million private valuation that the company is said to have received after its last funding round.

Turbonomic offers a so-called application resource management platform that uses artificial intelligence to help enterprises optimize their public cloud environments. The platform focuses on two main use cases. The first is optimizing cloud applications’ performance, while the other is reducing the infrastructure costs they incur. 

Turbonomic’s AI models analyze a company’s cloud environment to see if any workloads are allocated fewer hardware resources than they require to operate optimally. Turbonomic can, for example, detect if an e-commerce website’s database needs to be given more network bandwidth during weekend traffic spikes. The AI models then provide administrators with recommendations on how to fix the identified issues.

The platform provides similar recommendations for reducing wasteful cloud spending, Turbonomic’s other major focus. For example, Turbonomic can show information technology teams when they’ve provisioned more storage capacity for a workload than it needs. The platform can likewise detect if an application is running on an unnecessarily large and expensive cloud instance.

IBM’s plan is to integrate Turbonomic with its existing monitoring product, Instana, which it also obtained through a recent acquisition. Instana provides observability one level higher up the stack. Whereas Turbonomic focuses on detecting infrastructure-level issues, such as when an application is running slow because it hasn’t been assigned enough virtual processors, Instana is geared mainly toward spotting software-level issues, for example latency spikes caused by a buggy application update.

Another existing product that IBM sees Turbonomic complementing is IBM Cloud Pak for Watson AIOps. Introduced last year, the product provides AI-generated recommendations similar to those Turbonomic offers that can help IT teams troubleshoot infrastructure issues.

The acquisition advances IBM’s ongoing effort to refocus from its legacy business areas, such as professional services, to the high-growth cloud and AI markets. As part of the strategy, the company late last year announced plans to spin off its $19 billion managed infrastructure services unit. IBM Chief Executive Officer Arvind Krishna (pictured) stated at the time that the transaction will free up the company to focus on the “$1 trillion hybrid cloud opportunity.”

“IBM continues to reshape its future as a hybrid cloud and AI company,” Rob Thomas, the senior vice president of IBM’s Cloud and Data Platform group, said in a statement today. “The Turbonomic acquisition is yet another example of our commitment to making the most impactful investments to advance this strategy.”

The company also pointed to the broader industry trend towards AIOps, the practice of automating repetitive IT tasks with machine learning, as a factor behind the deal. “We believe that AI-powered automation has become inevitable, helping to make all information-centric jobs more productive,” said Dinesh Nirmal, the general manager of IBM’s IBM Automation group. Other major industry players such as ServiceNow Inc. have also made investments recently to expand their AIOps capabilities.

IBM expects the acquisition of Turbonomic to be completed in the second quarter.

“IBM’s acquisition of Turbonomic represents a logical next step to deliver aggregated value to enterprise IT leaders seeking to accelerate cloud migration and value gains,” Gartner analyst Chirag Dekate told SiliconANGLE in an email.

Turbonomic CEO Ben Nye spoke in 2019 with theCUBE, SiliconANGLE’s livestreaming video studio, about the growing importance of AIOps:

Photo: SiliconANGLE