Will Sun shine again?
Scott McNealy is giving away software to sell hardware. Will his razor-and-blades strategy put Sun Microsystems back at tech's cutting edge?
By John Battelle

(Business 2.0) - Can Sun make it? That question has been asked time and again in Silicon Valley ever since Sun Microsystems's fortunes rose and fell--and fell, and fell--with the Internet bubble. Sun once declared itself the "dot in dotcom," but while the Web is basking in the glow of a healthy resurgence, Sun remains under a cloud. Even a deal with Google failed to get Wall Street excited.

So how is Scott McNealy, one of the Valley's longest-serving CEOs, going to get his company moving again? Even during the darkest days of the bust, he never cut spending on R&D, laying out a cool half-billion dollars every quarter on new software, hardware, and microprocessors. In the past year, those investments bore fruit in the form of new systems designed for the ever-growing demands of computing on the Web.

Giving it away
Sun Microsystems CEO Scott McNealy has a new game plan.
Sun Microsystems CEO Scott McNealy has a new game plan.

Just one problem: How do you sell this stuff when open-source software is free and servers with Intel chips get cheaper every day? McNealy has crafted a "razor-and-blades" strategy to turn his iconic company around. In early December he showed his hand. First, Sun announced it will give away most of its software (that's the razor). Next, Sun rolled out an entirely new line of Web-friendly servers (the blades) that McNealy believes will best anything Dell, Hewlett-Packard, or IBM can offer. Business 2.0 caught up with McNealy to find out what's behind the giveaway.

You've spent billions of dollars to develop your software. Why start giving it away now?

If you google "McNealy software free," you'll see that we started talking about this as long as six years ago. But the big issue we've always had is getting all of the encumbrances out of the way.

The encumbrances?

As you build a product, you might need a driver here or a little encryption program there, and so you go out and buy a little piece of your software. It becomes an integral part. And when you want to go open-source, you need that little piece in there--but you can't open-source your supplier's product. So you have to reverse-engineer all of that code.

What about some of the crown jewels that remain under lock and key, like Java?

All of our products have open interfaces, open-source strategies, and community-development efforts. We are easily, wildly, the most partnered and community-development-oriented company in the technology industry. Microsoft ain't even close.

You built a detente with Microsoft a while back, but here you are poking them in the eye again. Microsoft's obviously square in the sights of this new effort, right?

I've seen many quotes lately along the lines of "Sun is the only serious alternative to Microsoft, and now it's free." And that's both a flattering and a challenging statement.

We aren't talking about IBM's WebSphere software because it doesn't play in the same league--it doesn't scale, it doesn't have the volume and the pricing and the open-source strategies and all the rest of it that we have. And there's nothing at Hewlett-Packard to dump on either. Longer term, the big challenger is going to be Microsoft. That's the most flattering thing I can say.

What about your acquisition of StorageTek? Some analysts have said it made about as much sense as HP-Compaq ...

Do we want to dredge that one back up? The difference is this: Compaq and HP are like putting your two hands together, absolute 100 percent overlap. It had to be one of the most complicated and gut-wrenching acquisitions ever because everybody in the company was wondering whether they lost their job to the other guy. With StorageTek, there was almost no overlap in terms of skill sets and folks.

Back to Microsoft. For a long period of time, Sun was really the point of the spear--the Valley company that stood up to Microsoft. It seems that that spear has been passed to Google. As a veteran of that war, what do you make of this new battle?

Microsoft has broadened its scope. It's MSN vs. Google, and it's Xbox vs. PlayStation, and it's .Net vs. Java, and it's Windows vs. Solaris. We're just not the Lone Ranger going out and competing with them anymore.

Google seems to embody--even more than Sun--your slogan "The network is the computer." Is that idea back?

Duh, it never went away! Anybody who bets against bandwidth--that's [Google CEO] Eric Schmidt's line--or anybody who doesn't think the network is the computer or who thinks it's smart to carry your personal mainframe around under your arm ... that just doesn't make any sense. We've just been plugging along, and during the ups and downs we have been very consistent.

About those ups and downs ...

I got all this "open-letter" advice from folks. The world's just sharing with me how they'd run Sun. People shared regularly.

How do you feel about that?

I said "Thank you for sharing" every time. I'm very California now, you know. I'm losing my Midwestern character. But one thing no one ever said to me was "Stop innovating."

Hence, the new lines of servers you announced last month?

They just blow away [the competition].... It's based on the UltraSparc T1 32-thread 9.6-gigahertz chip.

Could you speak English for the benefit of our audience?

Let's assume you're Google and you want to do billions of searches a day. Or you're eBay and you want to run hundreds of millions of simultaneous auctions. Or let's assume you're AT&T and you want to run a whole bunch of VOIP calls. You want a thread that uses very little space and very little energy and costs very little money.

What's a thread?

A thread is what the software application needs dedicated to it to go get something done. So a thread can run a VOIP phone call or a streaming video or an eBay auction or a Google search. It's a dedicated piece of hardware that can execute one of these transactions.

So has Google switched over to Sun hardware?

I could get in trouble there--Google likes to be secretive, and I honor that with my customers

Well, you did at least use the word "customer." Allow me to take a different tack. Part of Google's mythology includes the story of how it rolled its own computing grid using cheap, easily replaced Intel-compatible servers. Why doesn't that approach work anymore?

As more users join the network and energy prices soar, CIOs are concerned about data center environmentals like energy, cooling, and space. Google's single biggest discretionary expense is power. What's interesting about putting 32 hardware threads on a single processor is that the internal system bandwidth is enormous and at superhigh speeds. There's no way you can strap 32 Intel Xeons together and equal what we can do. You just can't get the bandwidth, the power, the space.

Four years ago we told our engineers to throw out everything that doesn't matter and add in everything we need to create something that is very, very different, which will run Solaris and Java but will be the cat's meow for an eBay or a Yahoo or a Google or a Merrill Lynch or a Citigroup.

In other words, anyone who needs to have a massively scaled Web presence or computational horsepower. So what are your expectations for sales?

We're building communities right now, so we're pricing it very, very aggressively. We're going to ship thousands this quarter.

So no upward guidance to Wall Street?

Sun doesn't give guidance.

Ah. You clearly are kindred spirits to Google. Wall Street has been disappointed in Sun these past few years, and many observers have written it off altogether. Does that ever get you down? Did you ever lose faith?

Nah. You know, I named my first kid Maverick. And my theory of operation has always been that if you don't have a very controversial strategy, you have no chance of making money.

If you can't differentiate, you can't differentially price; if you can't differentially price, you have no market power and you can't create any profits. So you've got to have a little different strategy, and ours has been very controversial the last four or five years. We did make some mistakes, but people were just looking at our mistakes and not trying to understand our strategy. I think the worm's turning. I've got to believe the next five years will treat us much nicer than the last five years did.

Click here for an exclusive interview with all four founders of Sun.


John Battelle is program chair of the Web 2.0 conference and author of "The Search" (Portfolio, 2005).

To send a letter to the editor, click hereTop of page

Follow the news that matters to you. Create your own alert to be notified on topics you're interested in.

Or, visit Popular Alerts for suggestions.
Manage alerts | What is this?