Skip to Content

Why IBM is still hot and bothered over mainframes

New York City -  Ross Mauri, manager IBM mainframe systems, poses with new System Z machine on Wednesday, January 14, 2015.  (PHOTO: Augusto F. Menezes/Feature Photo Service)New York City -  Ross Mauri, manager IBM mainframe systems, poses with new System Z machine on Wednesday, January 14, 2015.  (PHOTO: Augusto F. Menezes/Feature Photo Service)
IBM's new System Z machine.Photograph by Augusto F. Menezes — Feature Photo/IBM

The cloud may be IBM’s future—a reorganization announced internally right after the New Year underscores CEO Ginny Rometty’s desire to accelerate that mission. But it’s still betting big on its mainframe past.

Watch more about Ginni Rometty from Fortune’s video team:

To that end, the company just spent five years and more than $1 billion to develop what it bills as the most sophisticated computer on the planet, the z13. The technology was designed in collaboration with 60 of its most strategic mainframe customers. It can scale up to 8,000 virtual services and is capable processing more than 2.5 billion transactions per day. That’s the equivalent of the online shopping volume generated on 100 Cyber Mondays.

IBM (IBM) is pitching the system as a platform well-suited for the mobile economy, able to handle billions of interactions faster and more securely—everything from fraud detection for fund transfers via a smartphone banking app to credit-card validation for retail purchases. It will start taking orders this quarter, but a senior executive wasn’t inclined to discuss pricing specifics.

“You touch a mainframe every day, but you might not even know it,” says Tom Rosamilia, senior vice president of the IBM Systems division. Industries that are still particularly dependent on the technology: financial services, telecommunications, and retail.

“They are so embedded in the IT organizations of most enterprises that replacing them is extraordinarily risky, expensive and difficult,” says Richard Fichera, vice president and principal analyst with Forrester Research. “Most mobile transactions result in some access to mainframe-resident data being triggered.”

Matt Eastwood, group vice president for research firm IDC, figures mainframes still represent 10% of annual spending on server technology. “They remain very strong platforms for a variety of mission-critical workloads including business applications, transaction processing, and large data warehouses,” he says.

IBM is also talking up another potential use case: cloud data centers. “IBM often expresses the notion that the mainframe is the ‘original’ cloud, in that the virtualization technologies and automated management tools that are key to cloud computing originated decades ago in mainframe systems,” says another analyst, Charles King, with Pund-IT. It isn’t that common, but there are a few IBM customers who run private clouds using the System z platform, King says.

Over the past month, IBM has disclosed several high-profile cloud “wins.” Many of these deals, such as one with Dutch bank ABN AMRO, build on existing outsourcing contracts. Still, what the computing giant counts as cloud revenue grew 50% for the nine months ended Oct. 31; for 2015, it is projecting $7 billion for this line of business.

Although it hasn’t talked much about the move externally, IBM is restructuring to create customer-centric business units that prioritize its cloud business and several other strategic imperatives, such as Watson business analytics services. There are eight divisions being created, according to a report by Technology Business Research (TBR). The main units include Watson, Analytics, Commerce, Security, Cloud, and Systems. The existing services teams remain relatively intact.

TBR notes:

Competitors will face an IBM that can fully enter client discussions with a choice of consumption options in specific areas. As buyers broker decisions between components and cloud, for example, IBM has an advantage over competing firms that can only sell one part of the mix. The challenge for IBM is to continue to train and align its traditional salesforce to sell across this new model, as revenue recognition, value propositions and customer conversations will differ.

This item first appeared in the Jan. 14 edition of Data Sheet, Fortune’s daily newsletter on the business of technology. Sign up here.