For those of you whom I have not had the pleasure of meeting yet, my name is John Meyer, and I'm one of the seven SPARC server technical specialists in Sun Microsystems' U.S. field organization. My travels have taken me to many a different customer since I joined Sun over 13 years ago, but it was not until my good buddy, fellow office practical joker, and mentor Dave Edstrom started blogging recently that I thought I would take up the keyboard and start getting my thoughts out to cyberspace as well. I'm very passionate about Sun in general and SPARC in particular.

"Why should my long-winded rants, sanctimonious barbs, and brutal enthusiasm be confined to Sun's internal e-mail aliases? This is good stuff!", I thought to myself. It was also a great chance for me to Photoshop a cool logo and name my blog after my favorite scene from one of the funniest movies ever made (not to mention a cult classic among the systems engineers in the local Sun office).

If you're wondering why I'm so excited about SPARC after all these years, have a look at this chart, which summarizes the state of the SPARC ecosystem in early 1994, the year I started work here. Our biggest competition in the systems area was DEC's Alpha and we stacked up like this:

MY WORLD IN 1994 SuperSPARC Alpha 21064A
Frequency (MHz): 33 - 75 200 - 300
Process (μm): 0.8 0.5
Operating System: Solaris 2.2 was slow and buggy Ultrix and VMS were world-class
Service and Support: Break-fix, immature Legendary

You should keep in mind that 1994 was a time before most people knew what the Internet was (I had to explain it to my mother when I landed the job at Sun; now she has a webpage, downloads music from iTunes, and tells me how to auction stuff on eBay). It was also a time when clock rate really was the leading indicator of performance in a system. Caches were still fairly new and uncomplicated, and interconnects were relatively simple. We were getting beaten hands-down on benchmarks for all the right technical reasons by Alpha, and I mean by miles.

But you know what the punchline is: DEC and Alpha are both gone forever, and SPARC is not just still around but on the leading edge of microprocessor and open systems technology. If we could win with what we had back then, we can certainly win with what we've got now.

I'm convinced that the reason for our victory is that our customers sensed the same thing that I've always known about Sun. That is that we are one of the very few systems companies left who are doing anything really interesting. When not a single customer or analyst thought there was anything wrong with proprietary hardware and software, Sun saw open systems as a business strategy and forced all competitors to follow suit. When most people thought a network was no more than a tailpipe to a mainframe, our motto was "The Network is the Computer" (pretty obvious now, eh?) and we developed Java, the lingua franca of the Internet. And now we're often told by the purveyors of conventional wisdom that Sun's investment in SPARC and Solaris are a waste of money and resources. Why develop a processor and operating system when Intel and Linus (or Bill) can do those things for you? The reason is simple: if we only did what everyone else is doing, why in the world would anyone buy from us? That's usually when those purveyors of conventional wisdom fall silent.

I firmly believe that chip multithreading is the single most important technological advance the microprocessor world has seen in at least two decades, and we're the only ones who have it. A few months ago, I met with the CIO and CTO of a major financial firm in New York for a SPARC futures briefing. When I got to the part about ROCK, I told them, "This is the point where we finally put IBM out of the server business forever." They thought I was kidding and smiled, but when I told them that I was dead serious, they knew I meant it as only a fevered lunatic can mean something.

I also know that Solaris is unquestionably the most reliable, scalable, performant, and secure (not to mention technologically interesting) operating system in the world today. Don't take my word for it: when was the last time time you saw anything in AIX, HP-UX, Linux, or Windows beat out thin-film solar energy panels and powdered inhalable insulin as the Wall Street Journal's number one most innovative technology of the year?

In short, it's just fun to work at Sun right now. Our customers know they get a competitive differentiator, not just a world-class piece of electronics, from us.

So I hope you'll tune in occasionally to read my thoughts, tolerate my rants, forgive my mistakes, and participate with me in this stunning new world of SPARC.


John - you, rant? never! passionate discussion perhaps.

Posted by Chris Spina on May 14, 2007 at 04:25 PM EDT #

Welcome aboard John! Now I have to go watch Spinal Tap. :)

Posted by melanie gao on May 14, 2007 at 07:01 PM EDT #

I think Sun has a winner on it's hands with Rock and Solaris 10 but call me when the Rock CPU starts shipping in low-end and mid-range Sun boxes.

What is Sun's current answer in the low-end? UltraSparc III+ !!!!! That 3+ not 4+.

Power, which should ship within the next month, is just going to put the heat on. As for putting IBM out of the server business. Intel was caught completely off guard when AMD released 64bit Opteron and as you can see Intel survived. Big profitable companies usually do. And now Intel is for the moment back on top with quad cores.

Posted by Bryan Althaus on May 17, 2007 at 06:51 AM EDT #

Post a Comment:
Comments are closed for this entry.



« October 2016