I have to admit: I get a kick out of the name for the annual Oracle Openworld event because "open" isn't exactly core to how that company does business, at least not if you are a third-party supplier of services. It has largely implemented an Apple-like strategy, which increasingly ties all revenue related to Oracle technology back to Oracle. It appears to be doing this while still avoiding the mass exodus of customers who are clearly frightened by these moves. Most are either not frightened enough or so locked in they don't move (though some apparently are suing). Oracle actually seems to be getting sued by customers quite a bit at the moment, suggesting financial pressure that may not be reflected in its financials. It is an interesting dance, but one the company is having a great deal of difficulty with given its falling hardware market share.
Covering up a Problem
Sun went into x86 hardware - first with AMD and then with Intel - in an attempt to address that firm's rapidly falling market share. However, Sun was not structured to compete in the low-margin server business and seemed to be trying to become a software-only vendor at the same time with catastrophic results. Software companies doing hardware have a bigger problem than hardware companies trying to do software because software margins are traditionally far larger than hardware margins. This means, had Sun been successful, the market would have seen increasing margins for Sun as more and more of its business shifted from hardware to software. For Oracle, the problem is that the more it sells, particularly low-end servers, the worse its margins look and success in volume would likely lead to a declining valuation for the company. In other words, it really can't win and this isn't a skills problem anymore than bigger buckets would have saved the Titanic.
Now, add to this that the market it deals in is very conservative, none of the other major vendors are failing (granted HP has been having some messaging and leadership stability issues) and relationships are well established. This means that the two or three vendors typically bidding on these high-volume servers are institutionalized, requiring Oracle buy-in by either providing incentives or discounts that would likely take the deals into the red for Oracle.
As a result, Oracle is bleeding share badly and now argues that it isn't really interested in this x86 business anyway. Unfortunately for them, it isn't that easy.
In or Out
Unfortunately, IT really doesn't like blended vendors and tends to prefer as homogenous a shop as it can to lower problems with service. If vendors could just come in and sell the high-margin stuff, most would have abandoned the lower-margin commodity servers long ago. This also goes to account control because if you can't contain the entire shop, or most of it, you won't be the primary vendor and secondary vendors are typically easier to displace. In short, if you have IBM and Oracle in a shop and IBM can provide all of the IT needs but Oracle is picking and choosing, over time, IBM should be able to displace Oracle because it can fight on more fronts and increasingly structure deals to require hardware Oracle doesn't want to provide or increasingly no longer sells.
Sun Strategy Failing - Appliance Defense
By giving up x86 hardware, Oracle effectively gives up on its account control strategy, which was core to the reason for buying Sun in the first place. In effect, it appears to be coming around to the idea that the Sun acquisition has failed, and, as a result, has brought out Oracle appliances. These are platform products that wed hardware and software into a simple solution. These force a blended buy and lock in the Oracle software to the Oracle hardware. It isn't a bad idea for a defense, but it has to be recognized as a defense. Much like it would be if an Army, upon winning a war, recognized it couldn't control the country it had conquered and had to build forts (French Foreign Legion analogy). This would work as a holding exercise for a while and be particularly interesting in smaller shops, but in larger enterprises would remain at risk from the more prevalent vendor, be that IBM or a stronger HP (which is rapidly building a software capability).
While the appliance approach is tactically sound as a defense, and, admittedly, may be stronger for mid-sized companies, it leaves Oracle exposed strategically in the enterprise and the firm will still need to return to an offensive account-control strategy if it still plans on taking IBM on broadly.
Wrapping Up: Next Steps
Even if Oracle could afford to buy HP, that would only exacerbate Oracle's margin problems and as IBM and HP build software revenue, their margins should increase, putting Oracle on the defensive. Appliances may slow the erosion and might even reverse it in the mid-market, but in the enterprise they are only a tactical solution. In the end, the company likely will come to the conclusion that it will have to spin hardware out and hold it as a separate entity and treat it like a long-term investment.
After separation, Oracle would then fund and structure it so it can compete in the market broadly with both low-end and high-end solutions and better match HP and IBM after taking a onetime write-off to stabilize the financials and ensure that funding. While Oracle is doing well financially - amazingly well, which some customers are also objecting to - its ability to continue down a path of account control is far from certain.
Until it realizes this long-term problem with hardware, it is likely to continue to lose hardware share and move farther and farther away from its goal of account control and will increasingly cede that position to IBM.