Sun and IBM held a joint press conference today to announce that IBM will extend support for Solaris x86 to most of Big Blue’s x86 server offerings. IBM will distribute Solaris x86 and basically support it like any other x86 operating system. Sun will provide service and maintenance for the operating system, and both companies will spend money to qualify and test systems to ensure full compatibility and performance. You can find the joint press release here, and some pretty good analysis from The Register here (yeah, I’m linking to them because they quoted me.) So now that the “official” details are stipulated, let’s into discussion

First up, an examination of the deal from both sides:

Sun Upside: this is a big win for Sun, at least in the short to medium term. It is the biggest validation of the Solaris x86 to date and opens up a huge new market for Sun at least from a software perspective. While HP has been supporting Solaris x86 on their systems for quite a while now, their relationship with Sun is more “arms length” according to Jonathan Swartz. The partnership with IBM is much closer, with both partners pledging to work together to optimize how the o/s works on IBM boxes.

Having IBM jump on Solaris x86 should spike uptake from customers, along with motivating more ISVs to port more applications which is absolutely crucial to the long-term success of the operating system. If Sun and IBM can successfully work out the details, Solaris x86 will be able to reach scalability heights that were heretofore impossible. IBM is the only major x86 vendor offering greater than 8-way x86 systems (their x3950 scales to a single-image 16 socket server). This hardware, coupled with Solaris virtualization, could give VMware and Xen a real run for their money in the burgeoning x86 virtualization market.

Sun Downside: By getting IBM on board the Solaris x86 train, Sun has just invited a potent new competitor to each and every Sun x86 deal. With their direct sales force, IBM has a solid presence in almost every large IT account in the world, and is the dominant vendor in many. Sun is banking on their belief that their x86 and SPARC-based boxes are more than a match for anything IBM brings to the table. Of course, they know they won’t win every deal, but they figure they can hold their own in most deals. This may or may not be true, but the point we want to make is that in many cases deals are won or lost on factors other than hardware specs or pricing.

This is where IBM truly excels in the industry. They have a massive array of services, ranging from installation and break/fix to architectural to outsourcing to full-on business consulting. Sun’s service prowess fades quickly after maintenance and break/fix. IBM can also deliver financing for almost any size deal, and tailor terms to fit any situation. While Sun may be able to match prices and payments, it gets pretty expensive to match custom terms on any but the largest deals. The bottom line is that on a deal of any size, IBM has a lot of tools they can use to capture the business tools that Sun either hasn’t developed yet, or will probably never possess.

IBM Upside: IBM is kind of like the local big box hardware store. They have pretty much every tool you can think of, along with all the nuts, bolts, wires, paint brushes, and anything else you might need for a project. They can also hang a new door, install a new garage door opener, or just give you some advice about what kind of saw blade you need to cut concrete blocks. Like that hardware store, IBM has a competitive product in almost every IT category. It’s not necessarily because they want to it’s because this just what you have to do to be a whole-enterprise vendor.

Bill Zeitler (IBM Senior VP of IBM’s systems group) talked about the importance of offering customers a wide range of choices a key concept, but relatively new to IBM. Since IBM’s near-death experience in the early 1990s taught them that offering customers alternatives works, their success in the 2000’s has only reinforced the need for them to provide a wide range of products even if these products rely on third parties to complete the offering. I think that IBM is looking at the long-term picture, and believes this move will ultimately help them gain share. Sure, they’re validating Sun’s operating system strategy, but they’re also getting into deals they didn’t have access to before and getting the chance to capture services, financing, and maybe over time paint the account a deeper shade of blue.

IBM Downside: There are some risks to this deal. Just as Sun is giving IBM a shot at their installed base, IBM’s endorsement of Solaris x86 certainly gives Sun a leg up in some accounts. Solaris x86 can now run on existing IBM x86 gear (or at least the newer models). This gives true-blue IBM x86 customers an easy way to give Solaris x86 a shot if they like it, the next step is to kick the tires on some of Sun’s x86 gear. Sun still gets the large majority of their revenue from hardware sales and their sales force is one of the most aggressive and hungry in the industry. They’ll attack a hardware deal like a pack of blood crazed badgers although we’re not sure that badgers run in packs

More interesting points.

Although this wasn’t in the press release, we learned in the Q&A session that IBM and Sun have been working on porting Solaris x86 to IBM’s System z mainframe platform. This is truly a surprising development. Combining the scalability, performance, and application portfolio of Solaris with mainframe virtualization, stability, and security could result in quite a solid value proposition. However, I believe they are porting Solaris x86, which has a much smaller (although still 3,000 apps) set of ISV applications. Still, it is an interesting concept although I’m not sure why Sun is interested in pursuing this opportunity. In my mind, successfully enabling native Solaris x86 on the mainframe would almost guarantee that many of the mid-sized and smaller applications currently running on Solaris x86 would eventually be consolidated onto mainframes. We’re already seeing this happen with Linux workloads. We can only conclude that Sun believes they can make significant money from o/s licensing and maintenance fees, which is certainly possible as mainframe customers are generally willing to spend significant money to gain the advantages mainframes provide.

Another interesting tidbit came from a question concerning IBM’s AIX Unix operating system. Someone asked if it was in the plan to port Solaris to IBM’s POWER platform. Jonathan certainly seemed willing to do it, and even Bill Zeitler was somewhat positive about the notion. However, I think the technical aspects of this would be pretty complex, requiring not just o/s porting, but probably a lot of application porting as well. For Sun, this is a slam dunk decision they have nothing to lose and everything to gain by porting Solaris to competitive platforms. For IBM, they really don’t need Solaris in order to be competitive in the Unix market. They, along with HP, broke Sun’s Unix dominance several years ago and it isn’t likely that either IBM or HP would move to Solaris in order to save money on o/s development. In Unix, the operating system is perhaps the greatest source of differentiation, why give that up if you don’t need to?