IBM CEO Urges Partners To Choose Their Cloud Architecture Wisely, Noting That Decisions Made Today Will Impact Their Businesses For Decades

IBM CEO Ginni Rometty told IBM partners Tuesday that the cloud architectures they implement for customers at this "interesting moment in time right now" will likely impact their respective businesses for decades to come.

"Architecture decisions determine an awful lot of things," Rometty told attendees of the IBM PartnerWorld conference. "They have a big influence on everything downstream."

Rometty swung through the partner conference in Las Vegas between travels in Africa, and on her 60th birthday, she said, because "I need to be able to keep telling you how committed we are to your success."

[Related: IBM CFO Tells Investors Goals Of Cognitive Transformation Are In Sight]

In Watson, IBM developed the first enterprise-grade cognitive cloud platform—encompassing artificial intelligence and data processing capabilities—that can deliver competitive advantages to businesses. Cognitive computing is poised to become a $33 billion industry, Rometty said.

"Cloud is the platform for business," the CEO said. And "AI has burst onto the scene full throttle."

Those developments pose "an interesting set of architecture decisions" for enterprises, and the business partners they hire, that will reverberate across their organizations for many years, if not decades, she said.

Customers, she said, have to make key technology decisions in three areas: data platforms, artificial intelligence platforms, and cloud platforms.

Watson has the crucial attribute of being an open data platform, breaking down silos and enabling an entire ecosystem to build solutions integrating data sources. That yields unique benefits for systems integrators, solution providers and resellers, she said.

As an AI platform, Watson offers a range of cognitive services, from natural language processing to machine learning to vision recognition. And equally important, it's transparent, Rometty said.

"To trust these systems you need to know who taught it and with what data," she told partners.

A unique quality of a cognitive solution like Watson is that because the system learns, there's a built-in advantage to those who start using it earlier, she added.

Ultimately, Watson's broad range of data, analytics, and machine learning functionality can only be delivered on the IBM Cloud, a platform optimized for data and cognition, Rometty told partners.

"We've got to win this architecture battle," Rometty said in her keynote. "You're going to be setting the table for the future."

Rometty's message to partners essentially restated a time-honored industry principle as it applied to a new technology, said Lief Morin, CEO of Key Information Systems, an IBM partner based in Southern California.

"A long time ago the [hardware] platform was the differentiator, whether it was the mainframe, AS/400, or the rise of Intel. They were sticky, with applications written for certain platforms," he said. The same was true for the software stack.

And the same rules, and business imperatives, apply when putting data into an AI platform, Morin said.

Since any AI solution needs to be trained over time with extensive amounts of data, it also carries the potential to be highly sticky for a vendor.

IBM has spent years building its AI platform, and with Watson has a product that clearly can differentiate itself in a market in which everybody is trying to implement a digital strategy, Morin told CRN.