IBM could buy SUN: and then?

Some rumors say that IBM could buy SUN for 6.5 billions of dollars. This news can appear rather common in the current economic context, but it generates many questions in the small world of computer science.

The purchase of SUN by another company is not really surprising: SUN has negative results for several months. Despite of good ideas and good technologies, the company had some hesitations in the past, and lost market share, especially in servers market. Many doubt about its ability to reach a good position.

By reading the article, I wondered about the consequences of this operation:

  • on the servers’ market,
  • on the future of processors,
  • on a software point of view.

At the beginning, I considered that this repurchase could generate very heavy consequences, but at end end …

Servers

We can divide the market of servers into two categories: servers based on Intel x86 technologies (Pentium), and the others. About the first category, the repurchase wouldn’t generate any impact. The market share of SUN is not high enough today to create big changes. This kind of hardware is very standardized, and can be rather easily, replaced by another.

For the servers based on Itanium, Sparc, PowerPC…, the world will be cut into two parts: IBM side, and HP side. Is this really good for competition? I am not sure. For the prices, there will be perhaps an advantage, but for a technical point of view with two competitors only …

The evolution will depends in fact on the feedback of the users: either they adopt this new bi-part world, or they leave their chances to smaller speakers for precisely leaving the constraints which will be imposed by the two large companies. It would be interesting to follow the behavior of these customers regarding their size.

Processors

For processors, things become more critical: until the end of the years 1990, we had the choice between various technologies: PA-RISC at HP, Sparc at Sun, PowerPC at IBM, Alpha at DIGITAL (repurchased by Compaq). Since the things evolved/moved: HP chose to replace PA-RISC with Itanium (Intel), then repurchased Compaq. The processors PA-RISC and ALPHA thus disappeared. If IBM repurchases SUN, we can suppose that technologies SPARC and CoolThread will also disappear.

So, we will have only technologies: Intel with Itanium and x86, and IBM with PowerPC. Same situation than the servers’ one? No: Intel largely dominates the market of the processors, and the current trend goes towards a generalization of the servers with x86.

Software

This domain is so wide, it would be difficult to be exhaustive.

From operating system point of view, we will certainly loose the Solaris system, which is, for me, one of best Unix. But the fact that this system exists in Opensource version could perhaps save it (but competition with Linux will be difficult).

IBM and SUN are both major actors on the market of the free software, in many projects. So, there will be no problem for applications such as OpenOffice. Both of these companies support the development of this application, offer office suite based on it.

I don’t think also that the repurchase changes anything for Java. But the Java development platforms could suffer from competition with the products supported by IBM.

The situation of MySQL is less clear: IBM has already a set of database engines, with especially DB2. I do not say IBM could kill MySQL, but it could slow down the developments of new functionalities on MySQL to avoid conflict with its own products.

For a general point of view, I am not especially anxious for the most popular software. But what about less known applications, like Glassfish. Free or not, the applications massively supported by SUN could suffer from this repurchase.

You can notice the big advantage of the free applications on the others: their development could be restarted. The other software can simply disappear.

Conclusion

If these rumors are true, we will have to wait a long time before the repurchase will really occur. We will have to wait the official announcement, the agreement of all administrations …

At the end, the main consequence of this repurchase could be a larger « INTEL-isation » computing world. Even if we are already in this way, this repurchase would accelerate the move the trend. I think it’s not a good thing, neither a competitive point of view, nor from the technical point of view.