Oracle's (Nasdaq: ORCL ) $7.4 billion acquisition of Sun Microsystems (Nasdaq: JAVA ) won't include an advanced chipset code-named "Rock." At one time, Sun executives thought the new chip would be far superior to server processors from IBM (NYSE: IBM ) and Intel (Nasdaq: INTC ) , The New York Times reports.
Rock was supposed to be a next-generation chip featuring many "cores" -- i.e., separate processing units strung together on a single chip, or "die" -- and thereby capable of crunching reams of data very efficiently. Think of it as similar to Cell, the multi-core chip from IBM, Sony, and Toshiba that captured headlines in 2005.
Adding horsepower via Rock's high-powered cores would have given Sun's servers a technical edge over servers built by IBM, Dell (Nasdaq: DELL ) , and Hewlett-Packard (NYSE: HPQ ) , among others. But now the project has been canceled, the Times reports, and Sun will have to tune its existing SPARC chips or find faster alternatives from Intel, Advanced Micro Devices (NYSE: AMD ) , or current big server partner Fujitsu.
That's understandable. Sun reported higher losses and a 20% drop in revenue in its latest quarter -- not exactly the sort of numbers that justify high-cost, low-yield projects.
And let's be honest: Oracle doesn't need to take on new hardware projects. Under Oracle, Sun's servers will become delivery vehicles for new database software, business applications, and middleware. An option for the discerning IT buyer who'd rather deal with one vendor, you might say.
Rock would have mattered more had IBM been Sun's white knight, as previously expected. Big Blue has a history of investing in innovation; Oracle has a history of squeezing efficiency from every business that it acquires. I expect everything that isn't necessary -- isn't proven -- will either be sold or canceled. Count Rock among the casualties.
It just didn't belong in Oracle's garden.
Get your clicks with related Foolishness: