While Schwartz is much calmer than his predecessor, Ed Zander, and his chairman and CEO, Scott McNealy, he is no less acerbic in cutting up the competition or being resolute in his view of where Sun has been, where it is going, and why.
What seemed obvious from Schwartz’s one-man show, which was billed as the Take Back Wall Street event and which was somewhat ironically held in New York’s Union Square, not on Wall Street proper, is that Sun has been humbled in the past four years, and that Sun now understands how far it has fallen behind its competition in the server markets and how it has let other vendors set the pace that it once set with great engineering and savvy marketing. And while Schwartz said some words of contrition about how Sun had threw away its fishing poles and maps in the late 1990s and early 2000s, when the fish were just jumping into the boat as Sun was growing revenues at 50% a year, like Schwartz, you would have to admit that this was perfectly human behavior. Even for a multi-billion company.
Schwartz rolled out Sun’s A100 workstation, its first product and one that he said Wall Street firms bought in droves to do their number crunching and then started using as a new thing called a server when they wanted some cheap networking power back in the 1980s. Wall Street is the swamp from which we were spawned, quipped Schwartz. This is exactly the place we think we need to return to.
He said that when hard times fell on the financial services market in the wake of the dot-com blowout, CIOs and IT managers had asked for Sun to get Solaris on cheaper x86 iron. They asked Sun to stop being an isolated island and to start interconnecting with other systems and deal with the fact that legacy systems will persist. They also, he said, asked Sun to get back to its roots and innovate as it once had, building a Ferrari out of spare parts. When Sun was not willing or able to do this, Wall Street and other financial centers around the world, which loved Solaris but which were angered by Sun’s high-priced Sparc platforms, voted with their budgets and cut Sun back to the bare minimum. The double whammy of the telecom downturn (perhaps Sun’s biggest business in the glory days) hit Sun hard, and it still took the company several years to realize its roadmaps were all heading in the wrong directions.
Today, Schwartz is second in command of a different company. One that has refocused on Solaris with its upcoming Solaris 10 platform and has Solaris 9 certified on 249 different x86 systems, including four of its own. Sun has partnered with Fujitsu Ltd to get faster SMP systems, is innovating with its massively multithreaded Niagara and Rock chips for low-powered, dense computers, and has adopted the Opteron processor for its core Solaris x86 platforms. More importantly, Sun is really focused on systems performance–something that has hurt the UltraSparc-III generation as IBM Corp’s Power4 and Power5 systems and Hewlett Packard Co’s Itanium 2 systems have pushed the performance envelope and brought the cost of Unix computing way down. We know we can win on performance, he stated emphatically, demonstrating how Solaris 10 on a Xeon server bested Red Hat Linux running on a Xeon box using the Volano benchmark, and how Solaris 10 running on an Opteron server smoked it. And he made no bones about the fact that just like Red Hat and Novell have been targeting Unix shops for their money, Sun was going to turn the tables on the Linux world and show better bang for the buck with Solaris 10 on Opteron and still be able to run Linux applications inside Solaris 10 containers through Project Janus. We are absolutely targeting Red Hat, said Schwartz, as has been evident from the past several months of rhetoric from the marketeers at Sun. Sun is giving Solaris to Red Hat customers at half price if they trade up to Solaris 9, and is giving big discounts to Xeon server customers who buy Sun’s Opteron boxes, too.
Schwartz once again talked about Sun’s plans to adopt subscription based pricing for its Solaris platforms, explaining that because Sun owned an OS and a hardware platform it could use that combination against vendors who have either a server platform (Dell, HP, and IBM in the X86 market) or an operating system (Red Hat or Microsoft). I don’t think anyone will survive in the one-way or two-way x86 server space who doesn’t own an operating system, he declared. But Schwartz seems to want to have it both ways, and the reality is that Microsoft has the deepest corporate pockets in history (tens of billions of dollars even after giving a lot of its cash hoard back) and could give Windows Server away, too, and shift costs to services or even desktops. And that Dell and HP, with higher server volumes, can make servers for less money than Sun. If Sun can become the volume leader in server operating systems or servers, this changes everything. And clearly Sun would love this. But this would be a very difficult task.
That said, Schwartz said that Sun has an installed base of 3 million Solaris licensees, and that is a good place to start.
In the interim, Sun has to do a lot of deals, and think outside of the box. It’s Compute@Sun offering, which he said would sell computing capacity for $1 per hour per CPU (that’s different from what the press releases said yesterday), is an interesting offering. While this is not for every workload, said Schwartz, in the long run, all computing will be done this way. He said that Sun would not necessarily be the utility behind this type of computing, but would partner with other service providers to deliver the hardware and software infrastructure that can enable number-crunching over long distance links today and that should be able to offer transactional capabilities some day in the future.
We don’t have all of the answers, he said. But we do know that a buck an hour [for computing power] resonates with CIOs. I think there is a little bit of inevitability to a pricing model that is so simple. At the very least, it will cause CIOs to question ISVs about how they price their software.
And for the doubters in the Wall Street crowd who don’t quite buy into this subscription-based pricing and that the commoditization of IT does not mean it is not of value or is not something where a company can make money, Schwartz chastised them. I want to remind everyone in the financial services business that you are in the commodity business. He said that all commodity businesses differentiate with technology and services, and that is exactly what he intended for Sun to do.