Having proved in 2001 that those who fly highest fall hardest, Sun Microsystems spent much of the new decade searching for a new identity, or so it seemed to outsiders. Judging from recent results -- a showcase of bright hardware and software achievements -- it turns out that the old identity was plenty good enough. Credit steady, longtime CTO Greg Papadopoulos for sticking with the old formula of great engineering and innovation. It worked for Sun before the bubble, and it's working for Sun now.
Deep down inside, Papadopoulos is a hardware guy. He designed supercomputers before joining Sun (way back in 1994), and he's still focused on total system design, as demonstrated by his lead role in the creation of Project Blackbox, the modular "datacenter in a can" Sun launched in late 2006.
[ Discover what insights you can take advantage of from the other 2008 InfoWorld CTO 25 winners. ]
For Papadopoulos, Blackbox was an important step toward solving a design problem that the industry has essentially dropped into customers' laps: figuring out how to integrate all of the elements of a datacenter in the most ecologically and economically optimal way.
"The Blackbox represents taking responsibility and helping lead a transition," Papadopoulos says, "where there is a co-design that takes place between the problems of power, cooling, and packaging, and the design of servers and storage and networking equipment. As an industry, we've essentially said to the customer, 'Good luck with that.' There's no substitute for real engineering there."
It's a perspective that explains why the CTO and his team, which manages Sun's technology portfolio and advanced research, did not give up on the Sparc microprocessor when, during those dark days of 2001, Wall Street called on Sun to abandon it. Instead, with the Niagara project, Papadopoulos and company steered Sparc in a new direction, eventually getting a jump on IBM and Intel with an eight-core, multithreaded marvel that's tailor-made for Web applications and heavy database workloads. Niagara-based servers brought Sun $1 billion in revenue last year, and they're on pace to bring in another $1 billion in 2008.
In short, Papadopoulos and Sun plan to remain true to their silicon roots. Considering the developments in multicore and multithreading, and where memory systems and networking are heading, Papadopoulos sees Sparc silicon ("being able to craft it and shape it") as vitally important to the company's long-term success. The other vital piece is the Solaris OS.
In fact, as an early catalyst for Sun's open sourcing of Solaris, Papadopoulos may be the man most responsible for the operating system's renaissance. Solaris has always had its admirers, but the OpenSolaris project has certainly helped to boost its popularity. Of course, it also helps that Solaris has a brilliant set of features and contains the otherworldly ZFS file system. As Papadopoulos himself would say, there's no substitute for real engineering there. That new Sun we were all looking for a few long years ago -- it's a lot like the old Sun.