After fading from the front page, the talk of an IBM/Sun Microsystems merger came roaring back to the forefront on late this week following a report in the New York Times that the talks were approaching the final stages, and Sun had dropped its asking price.
The original rumors put the purchase at between $10 and $11 per share, well above the $4 Sun (NASDAQ: JAVA) had been hovering around prior to the takeover news leaking out. Sun reportedly dropped the asking price to $9.50 with the assurance that IBM (NYSE: IBM) would fight anti-trust issues that may be raised.
According to the Times, around 100 IBM attorneys have spent weeks holed up in a Silicon Valley hotel poring over Sun’s patents and licensing agreements. They have a lot of ground to cover, from Java patents to Sun’s long-term relationship with Fujitsu, which makes and sells Sparc-based servers.
How much of Sun will survive remains another question. James Staten, senior analyst with Forrester Research, said it won’t be pretty, and Sun could lose up to one-third of its workforce, about 10,000 people.
“Storage has a huge bull’s eye on its back, Sparc has a big target on its back, software has a big target on its back,” he told InternetNews.com. “I think a lot of the stuff [Sun CEO] Jonathan Schwartz has been strategically investing in are not as important to IBM. IBM has similar products that are more successful than Sun.”
Things like Sun’s communications and application servers won’t survive when IBM has industry-leading software like WebSphere, but Sun’s identity server is a big exception that will likely become a part of IBM, he added.
It’s unlikely IBM will keep very much of Sun’s server business, given that IBM has been eating Sun’s lunch in that area already. They might keep Sun’s Constellation high performance computing (HPC) systems because it has a few advantages, but IBM would rather HPC customers went to Power than x86 processors, said Staten.
The rest of the hardware? It sets in the west. “The blade system and most of the x86 systems will likely be killed. The high-end Sparc stuff they have to keep around until they get the Power version of Solaris out. But I fully suspect Rock is dead,” said Staten.
Rock is the codename for a next-generation processor Sun has been developing for years but it has remained vaporware for some time. Sun has given no updates on the Rock roadmap for years.
So who is safe?
People working on Sun’s Solaris operating system can probably breathe a little easier. “They will obviously hold on to everyone in Solaris, and job one will be porting Solaris to Power,” said Staten, referring to IBM’s proprietary RISC
IBM will also likely keep Sun’s professional services, because most of that is in support of Solaris, and of course it would keep Java, Sun’s baby.
Surprisingly, Staten thinks IBM will keep the tape business, unless a government antitrust action forces them to split that up. “It all depends on whether you look at this as tape, or as storage. Tape is a miniscule part of storage and isn’t as big a deal,” he said.
Keeping Sun’s talent will be IBM’s real challenge. Staten described Sun as a “Do it first and get forgiveness later” culture, whereas IBM is much more deliberate. Still, people like James Gosling just might thrive in IBM. “IBM is number one in patent generation in the world. They don’t do that by having a stifling culture,” said Staten. “They allow creative people to flourish. But they try to channel them into areas that flourish and into a product line.”
Next page: The Fujitsu factor
Page 2 of 2
The Fujitsu factor
Sun’s lengthy relationship with Fujitsu is the hardest to evaluate. Sun makes Sparc and Solaris and Fujitsu resells the servers, so it is dependent on Sun’s continued development of both products. Would IBM leave the Japanese vendor high and dry? Or will Fujitsu take over its own R&D? Staten wasn’t sure.
“IBM could say ‘here, take Sparc,’ but for Fujitsu to value that, they need a long-term commitment from IBM that Solaris will stay a Sparc OS, but that’s not in IBM’s interest,” he said. It wouldn’t make sense for Fujitsu to support both OS and processor because it doesn’t have the economies of scale to support that, he added.
Another scenario is that Fujtisu could potentially nurse its market along until the transition to Power processors and they convert to Power, or convert their customers to Linux.
“Fujitsu might have the Sparc market to themselves for two years, so that might be worth something. To a degree, it buys them time to make a move. I don’t think IBM wants to force Fujitsu out of the relationship. They might see them as a partner to expand the presence of Power,” said Staten.
Meanwhile, those orbiting Sun wait
Sun employees aren’t the only ones chewing their nails right now. “For a lot of Sun customers, there will be a lot of trepidation about this merger, particularly if you are a Solaris loyalist,” said Staten.
“IBM will have to show a pretty clear roadmap and straight talk to customers to show a path is there. The longer IBM takes to show a path there, the faster customers will migrate,” he added.
For Mike Clesceri, vice president of marketing for Laurus Technologies, a Sun systems integration partner, it’s wait and see time. “I think there’s still an opportunity for Laurus,” he told InternetNews.com. “So long as IBM sees value in the core technologies Sun has, I’ll be able to service my customers and see an opportunity to expand the solutions I offer now because of what IBM will inevitably add to the product line.”
Right now he’s just a Sun reseller, but then again, he’s not selling that many Sparc-based servers these days. He’s selling solutions built on Solaris. “So if I have access to more systems that run Solaris, that can only be good for me,” he said. If IBM does migrate Solaris to Power, then that’s really good for Laurus.
“I’ll have access to customers that are IBM shops that would never have bought Sun gear, if IBM ports Solaris to Power. It also means more ISVs will move their apps to Power to get to larger more addressable markets,” said Clesceri.
