Five years ago, Oracle and Sun Microsystems consummated a controversial marriage few people believed would last. Today, the fruit of that marriage includes a radical new approach to the data center, a stunning combination of hardware and software, and a hiring spree of cloud engineers no one would have predicted, but which clearly reflect Oracle’s focus, leadership, and strength in innovation.
Shortly after Oracle disclosed its intention to acquire Sun Microsystems in 2009, then-CEO Larry Ellison said the following about the superb technology created by Sun, which he called “a national treasure”:
"We're keeping everything: we're keeping tape, we're keeping storage, we're keeping x86 technology and Sparc technology, and we're gonna increase the investment in it.... Sun has fantastic technology—we think it's got great microprocessor technology—it needs a little more investment but we think Sparc can be extremely competitive; [Sun's] got the leading tape archival storage systems—we think Open Storage, their new system, is absolutely fantastic; Java speaks for itself; [and] Solaris is overwhelmingly the best open-systems operating system on the planet. Sun has been a national treasure for the last couple of decades and we think with that combination of Sun technology and Oracle technology, we can succeed and compete and beat IBM. And that's our goal."
The person most deeply involved in marrying that “national treasure” with Oracle’s software technology and prowess is Oracle executive vice president of systems John Fowler, who had previously held many top engineering positions at Sun. And in an interview on January 21 during Oracle’s launch of a new set of engineered systems designed to become the “Data Center of the Future,” Fowler shared his insights on how the Oracle-Sun marriage, which was completed January 27, 2010, confounded conventional wisdom and set the stage for an entirely new model of enterprise computing.
“When Larry bought Sun, the press at the time was confused, because here was a big packaged-software company buying a hardware company—what’s the idea here?” Fowler said in a video interview with John Furrier and Dave Vellante of Silicon Angle TV.
“Larry always had this idea that he could actually put the engineers together and build better products—that was the beginning of the vision. From Day One, it was, ‘Go hire some engineers, John; go build me a more-aggressive product roadmap; go tell me what companies we need to buy.’ So it was the opposite of the then-current perception.”
Indeed, many observers at the time said the Sun acquisition would be a disaster because, those observers reckoned, what the heck does a software company know about building hardware? “We knew what we were building was something new, something different,” Fowler said, rejecting the suggestion that Sun and Oracle had antithetical visions.
“As I like to tell customers, I start off by saying, ‘I’m not conventional wisdom—conventional wisdom is to buy stuff from a dozen different vendors and put it together yourself. I’m gonna build you bigger and more-powerful systems that are cost-effective because they’re integrated.’ ”
Rather than having clashing visions, Fowler said, Oracle and Sun were ideal partners (a match surely foreseen by Ellison) because business customers were beginning to demand new computing models that the traditional approach—“buy stuff from a dozen different vendors and put it together yourself”—simply could not handle.
Fowler: “Now, before Oracle bought Sun, they had already been working on Exadata—so Exadata’s a vision that was actually started years before the acquisition of Sun. But what Larry discovered in building that [with HP hardware] is it’s just a lot harder when the engineering activity consists of contracts, and nondisclosure agreements, and all of that as opposed to ‘Everybody get in the same room—we’re the same company!’
“So you get all the engineers in the same room and we can change the software as we need to, we can change the hardware too. So that genesis was interesting because in one of my very first meetings with Larry, Larry wanted to talk about the SPARC processor. And he says, ‘John, can we put something into the processor to help accelerate the database and make the database more effective?’ And I said, ‘Sure—of course, Larry.’ So we went back to the teams and began working on how we could improve our silicon design to make a better product.”
Oracle plans to announce the first of its Software in Silicon products this year, Fowler said. “There’s a lot behind all of this that was driven actually by engineering—that is, how do you engineer a better product? That was the drive behind all of this.”
He added, “Today, if you go into an Exadata meeting, you’d have a hard time figuring out which men and women are working on hardware, and which men and women are working on the database. Because there’s this single conversation about, ‘What can we do to make that work better?’ And it’s very exciting.”
And as Oracle accelerates its move into all layers of the cloud-computing business—software as a service, platform as a service, and infrastructure as a service—that blended approach of engineering hardware and software together is the animating force behind Oracle’s unique approach.
“I know that from working closely with Larry, part of his vision was that Oracle could build better cloud infrastructure, both for customers and for our own internal abilities. And by having all of these engineers who build the applications, and build the middleware, and build the database, we could build a better mousetrap.”
Of all the tech companies clamoring today to position themselves as cloud heroes, none can match Oracle on that breadth and depth of scale in addressing every layer of cloud products and technologies, Fowler said.
Engineered for Innovation
Oracle’s mission is to help customers get out of the expensive and lengthy and low-value processes of cobbling together dozens or hundreds of mismatched parts, and in so doing to help those customers liberate precious investment dollars to apply toward customer-facing innovation. (Here’s a fascinating look at Fowler’s vision for highly-tuned systems.) The key to unlocking that potential, said Fowler, is to allow world-class engineers to shatter the old silos separating hardware and software and to work together to create entirely new capabilities.
“The idea of the platform—embraced by both Oracle and Sun—was very straightforward: you want to preserve people’s applications,” Fowler said. “You want them to inherit new properties when they use the products without them having to change their software—at all—so they have durability over decades while you keep making [the underlying systems and platforms] better and better. That fresh and unprecedented approach to building next-generation systems has significant appeal to young engineers, Fowler said, because Oracle offers them the chance to explore and discover new approaches to computing that are helping to revolutionize how businesses develop products, market products, engage with customers, manage operations, and communicate with the world.
“A few years ago when we started this, I realized that Larry’s passionate about college hiring, and I’m passionate about college hiring. We had a complete meeting of the minds,” Fowler said.
“For example, we have a very aggressive program for college hiring—over the past few years, I’ve hired in the U.S. alone 900-plus Masters and PhD’s from only the top schools—and it’s been tremendous. And now we have just a flood of people coming in because the word is out: ‘You wanna tackle hard problems? Well I’m tackling hard problems!’ ”
Those “hard problems” stem from Ellison’s long-held mandate that Oracle will aim to build nothing but the best products in the world—and Fowler said that when he puts that challenge out to great young engineers, they’re eager to embrace it. At Oracle, that new talent mingles with experienced Sun engineers, most of whom are still with the company.
“The plan is to always build the best hardware, right? Cause you know Larry! And while you could argue whether we’re doing it or not, there is no argument about what the plan is. When I go talk to a top engineer, I say ‘Here’s the plan: we’re not gonna build the second-best hardware—I need you to come here and help me build the very best.
“You’re saying, ‘Let’s build the best—no head-faking—the best!’ And you go to market like we’re doing today with the next generation, and we’re gonna get some smiles and some frowns, and then we’ll go back and change the frowns into smiles.
“And that’s a very simple architecture, right?”
