Servers, Solaris and SPARC key to Oracle's growth

Oracle's tilt at Sun Microsystems was opportunistic, Larry Ellison admitted to financial analysts yesterday, but Sun is now a critical component new five year growth targets that aim to more than double the company's sales to well over $50 billion.

Speaking at a briefing in San Francisco yesterday, Oracle's CEO joked that he and the company's two presidents, Safra Catz and Charles Phillips, were driving by Sun headquarters in Silicon Valley and there was a sign on the lawn saying Sun was for sale.

"Safra slammed on the brakes and almost threw Charles and me though the window. Charles jumped out and asked 'How much?' "

Ellison said the new products that would really "move the dial" were servers, Solaris and Sun's SPARC processors. He said he was confident of positive results starting to emerge two quarters after the buy-out, which is being examined by European regulators, is completed.

Ellison continued to reassure Sun customers that development of the company's major products would continue under Sun.

"We think we can build a very competitive chip technology and make money doing it," he said of SPARC. Enhancements could come in security and improved Java execution he suggested.

That said, Ellison pointed out that the second generation of Oracle's Exadata online transaction processing machine, produced with Sun, runs Intel processors and not SPARC. Oracle will continue to walk down both the Intel and SPARC technology roads, he said.

Ellison also had reassuring words about MySQL and Open Office, saying development of both would continue. The Open Office team in Germany was being kept together, he said, and he expected the product to be profitable.

"It's not going to be a huge business, but strategically it's very important that there's an alternative to Microsoft Office" he said.

Ellison said he expected a lot of interesting things to come out of the eventual merging of the two companies, citing the Sun Ray stateless thin client as one technology that Oracle could help commercialise.

"The Sun acquisition is going to pay off handsomely for shareholders and for customers," he said.

Earlier in the briefing, Catz also said she expected quick improvements in Sun's operating income, despite the fact the company's hardware business had cost of goods sold where Oracle's software portfolio did not. She said Oracle's support can be scaled to embrace Sun without any real increase in costs.

"I think it will surprise a lot of you the margins a hardware company can achieve," she told the analysts.

Catz also said there was a lot of technology at Sun that had no market, was undifferentiated and where there was no chance of winning in the market, though she did not give examples.

Asked why Oracle had not followed the likes of HP and Dell and bought an IT services company, Ellison said he and his team were "either brilliant or crazy".

He said services businesses aren't scalable and reiterated that Oracle is an engineering company first and foremost. He said he was comfortable with the company's current split of revenues at 80% from technology and 20% from services.

He also hinted that other hardware acquisitions could follow Sun, though they would probably not be of the same scale as the Sun buy-out was a unique opportunity.

With Oracle's long-promised Fusion Applications now scheduled for launch next year, Ellison downplayed any suggestion they would be rapidly adopted by users. He said they could be adopted any time between 2010 and 2020.

"It will take customers a long time to get to Fusion, but they're happy it's there," he said.

Ellison has been reassuring Sun customers and baiting IBM over recent days at Oracle's keynote Openworld Conference in San Francisco, revealing an advertisement offering a $10 million prize to anyone who can get a database application to run half as fast as they run on Exadata.

"IBM, you're welcome to enter," the ad says.

Ellison says new Exadata servers will be produced for other applications to make it easy for customers to deploy Oracle's technologies. He says it was a mistake to expect users to integrate and optimise a huge range of hardware and software technologies to make things work. Suppliers should be the integrators, he said, and Oracle would continue to pursue a strategy of "integration by engineering.