I had the same sense for Oracle's report last quarter. Hardware sales hit $1.2 billion then, but that was no more than in Sun's last independent quarter. Factor out that number, and Oracle would have done 11.8 percent year-over-year growth, not the 39 percent that it claimed.
This quarter the company claims $1.08 billion in hardware sales. You can only get the $1.7 billion Oracle touted by adding hardware systems support, otherwise known as the services part of Sun's business. Now look at the 10-Q for Sun's quarter ending Sept. 27, 2009. Service and storage products revenue (which does include software, so it's impossible to break the numbers apart) was $1.2 billion. Support services were $852 million, not counting another $204 million in professional and educational services. The total is over $2 billion.
Remember, too, that Oracle now sells its Exadata database machine, co-developed with HP (HPQ). [Update: Although Oracle originally worked with HP, it switched to Sun hardware last year.] CEO Larry Ellison said, "The worldwide Exadata pipeline now exceeds $1.5 billion for the full fiscal year." Pipeline? For those who aren't familiar with the sales term, it refers to all potential business, only a portion of which will likely turn into sales, depending on how the company calculates the number.
That leaves its current Sun hardware and support business at clearly under $1.7 billion, compared to $2 billion the year before. If Oracle has had any substantial success with the Exadata, then the Sun legacy business could be off even more than the minimum 15 percent you can calculate.
Now to R&D. Ironically, new co-president Mark Hurd, the former axe-man of HP research and development spending, said that Oracle would spend $4 billion in R&D this coming year. Wait, my BS detector has just gone off. Last quarter, Oracle spend $1.1 billion on R&D. At that rate, the company would have spent significant more than $4 billion. And the year before in the same quarter, Oracle spent $660 million in R&D, while Sun spent $354 million, for a total of ... $1.1 billion.
Granted, there's likely some economies the company can get by cutting duplicated overhead, but it sounds like there might only be enough money for Solaris and Sun software-related R&D to stay at roughly the same levels as when Sun was a dying concern.
Acquiring Sun might hold strategic promise for Oracle, but so far the integration of the business has much to prove. I have a feeling that HP, IBM (IBM), Dell (DELL), and Cisco (CSCO) have little to worry about Oracle as a hardware competitor.
- HP Hunts Former CEO Hurd To Hurt Oracle Over Hire
- Oracle Hires Mark Hurd. Will Ellison Regret This?
- Oracle Google Suit Could Kill Android and Threaten Open Source
- Oracle Owes Big Quarter to Sun Hardware, Software, and Fast Talking