Hewlett-Packard

Larry’s a hardware man now

After weeks of speculation about whether Oracle would keep or sell off Sun’s hardware business, we have the answer. Well, the answer for now.

In an e-mail interview with Reuters, Oracle CEO Larry Ellison made it clear he intends to keep not just Sun’s chip and server products but its disk storage and tape backup businesses too. So with one short interview Ellison has confirmed he will attempt to significantly change the competitive landscape among major vendors competing for the billions of enterprise dollars at stake.

And he is not lacking for confidence about his chances. In the Reuters interview Ellison said he has the in-house talent — both from among Sun and Oracle engineers - to compete successfully against the likes of hardware giants including IBM, Hewlett Packard and Dell.

“We have lots of hardware experience inside of Oracle. Hundreds of Oracle’s engineers came from systems companies like IBM and HP. Even I started my Silicon Valley career working for a hardware company that worked with Fujitsu to design and build the first IBM compatible mainframe,” Ellison said in the Reuters interview.

I am not sure how much of Larry’s own hardware experience will successfully translate to competing against The Big Three in a cutthroat low margin business. I suspect it will have more to do with retaining key Sun engineers and their managers working on key hardware technologies. But you have to like his optimism here.

It could very well be that Oracle has no intention of engaging in hand-to-hand combat with his major competitors in the low end, Intel-based server market. According to his comments in the Reuters interview, he intends to invest heavily in Sun’s Sparc- and Solaris-based servers where margins would be significantly higher.

“Once we own Sun we’re going to increase the investment in SPARC. We think designing our own chips is very, very important. Right now, SPARC chips do some things better than Intel chips and vice-versa. While most hardware businesses are low-margin, companies like Apple and Cisco enjoy very high-margins because they do a good job of designing their hardware and software to work together. If a company designs both hardware and software, it can build much better systems than if they only design the software,” Ellison said.

Yup, that’s right. Apple is a model, if not the inspiration, for Ellison believing he can deliver high margins servers if he can form fit Oracle’s software with Sun’s chips and servers ala Apple’s iPhone and iPod.

There may be at least a little concrete evidence to back up his ambitions. Oracle’s Exadata database machine, which tightly couples Oracle’s flagship database with HP’s server hardware, has received good reviews, particularly for its speed and performance.  It must be noted however, that the Exadata server uses Intel chips, and not RISC-based chips such as Sun’s SPARC processor.

Both Ellison, in the Reuters interview, and Oracle President Charles Phillips at last week’s Collaborate conference, said Exadata was the most successful product launch in the company’s 30-plus-year history. Oracle, of course, declines to release sales figures for the system, so there can be no iron-clad confirmation of this.

But if Oracle successfully applies its Exadata model to other server hardware-software combinations, perhaps targeting each offering at a specific vertical market, it may not only succeed in the market but also lay down the law for how server bundles will be sold.

There are a couple of unanswered questions remaining, of course. One, is if Oracle proceeds with its plans to sell SPARC-based servers bundled with its software, where does this leave HP?  HP still competes with Sun in some segments of the server market, and may not take too kindly to Ellison’s aggressive commitment to SPARC.

Second, how will Ellison deliver bundled combinations of servers to Oracle and Sun customers?  If he intends to focus on complete solutions using only Oracle-Sun chips, servers, operating systems, databases, middleware, and tools, the emphasis would seem to be on largely selling  these systems direct.  If he does an end run around the resellers, will this drive the channel into the arms of IBM, HP, and Dell that can reach customers across a greater number of markets?

We may not get these questions answered for another few months. But I’ll say this, with the Sun acquisition Larry has brought back some of the fun and excitement that has been missing from this market for some time now.

Living in a three company world

Several years ago or so Oracle chairman and CEO Larry Ellison made a typically brash prediction that eventually there would only be two or three major IT vendors left standing. Of course he believes Oracle is going to be one of them.

It is the kind of prophecy most people discount as self serving, and that couldn’t possibly come to pass.

But with persistent rumors swirling around the last couple of weeks involving IBM, Hewlett-Packard and Oracle all interested in buying Sun Microsystems, as well as rumors circulating that Oracle again wants to gobble up Red Hat, the possibility of a three vendor IT world seems more possible.

The development with the most potential to create this three vendor world is not the one where IBM buys Sun, but the one involving Oracle and HP dividing up Sun. According to those rumors Oracle is willing to put up $2 billion to buy Sun’s software business, most notably its crown jewels, Java and the open source data base, MySQL. At $2 billion it would be the steal of this young century.

Rumorologists have yet to attach a figure to what HP is willing to plunk down to take over Sun’s server-based hardware business. It is safe to say it would cost HP $3 to $4 billion, and that too could be worth it to secure HP’s top position in the overall server market.

But its Oracle’s possible move on Sun and Red Hat, in tandem with its increasingly chummy relationship with HP the last few years, that is at the center of all this.

First, there is the prospect of Oracle taking over control of Java. It is unlikely that even Oracle would consider monkeying around and changing the technical working of Java to serve its own development needs and so put a major competitor such as IBM at a disadvantage.

But it could make life difficult for competitors, most notably IBM, by raising the fees on Java next time Big Blue’s Java licenses came up for renewal. It could put IBM server products at a price-performance disadvantage against those of Oracle.

If it grabs hold of MySQL, Oracle could significantly enhance its credibility in the open source world, as well as gaining a low-end data base that could effectively compete against Microsoft. As more IT shops strongly consider open source products in these recessionary times, the prospects for MySQL are looking better and better.

Some might suggest that acquiring MySQL would threaten the margins Oracle makes on its much higher end bread and butter Unix-based data bases. I don’t believe it will. With Linux-based operating systems and their applications taking on increasingly mission critical applications, along with the high-end Unix market slowly shrinking, Oracle can avoid MySQL canabilizing the lower end of its proprietary databases and make this work.

Couple MySQL with Red Hat’s Linux, particularly the Enterprise versions of that product, and Oracle gains direct control of half the LAMP stack (Linux, Apache, MySQL, and PHP) and suddenly Oracle becomes the strongest vendor in the open source world — certainly the richest.

Then there is the increasingly tighter relationship between Oracle and HP. Oracle and Sun once had a very close relationship. Back in the hay day of the dot com boom, the “miracle stack” was Oracle’s databases, Sun’s SPARC servers and operating systems, Cisco’s communications hardware and the Apache Web server.

But a few years ago Oracle and Sun drifted apart over issues involving Oracle dissatisfaction about the cost of Java licensing fees, and the competition imposed by Oracle’s Unbreakable Linux.

Stepping in to take Sun’s place has been HP, as evidenced by deals such as the one last year between the two that resulted in the Exadata appliance server. That product, which is the marriage of HP hardware and Oracle software, that allows 11g to run insanely fast. Oracle hasn’t shown that kind of tight cooperation with a major vendor since its dealings with, well, Sun. And given that HP can provide all the servers Oracle could need (especially if HP acquires Sun’s SPARC servers), along with storage products, and a large worldwide maintenance organization would make for a very formidable team. And oh yes, Red Hat already has a good working relationship with HP, which makes Red Hat Linux available on its servers.

Oracle’s continued control of the proprietary data base market, its strengthened position in the open source world, and a tight relationship with HP, would put every major competitor, possibly excluding IBM, at a major disadvantage.

Even mighty Microsoft would have difficulty keeping up. As the world gravitates more towards open source for higher end applications involving cloud computing and SOA initiatives and turns  towards a rich well positioned supplier like Oracle, Microsoft would have to go on the defensive. And with Oracle working more closely with HP to deliver higher-end margin rich solutions, Dell too could be commoditized down to a second tier player in the enterprise market.

Two other things lend further credence to this scenario materializing. One, Oracle has a proven track record of making large acquisitions work, and two Sun, despite IBM engaging it in talks first, prefers to sell to a west-coast based company, according to rumors.

Lord knows what Sun’s poor board of directors is thinking given the possibilities potential buyers have presented to them. But if Mr. Ellison can entice Sun, Red Hat and HP to go along, his outrageous prediction of just three IT companies left standing, namely Oracle, HP, and IBM, is not so outrageous.