Since the beginning of the personal computer market, one argument was defined by an initial disagreement between Apple and Microsoft about whether PC hardware and software needed to be closely coupled. Microsoft initially appeared to win this disagreement. (A copy of the letter to Apple's CEO from Bill Gates suggesting Apple license can be found here.)

How does it look now? Well, look at Palm which, operating under the assumption that Microsoft was right, separated its hardware and software units to improve financial performance. The software side failed; the hardware side is currently treading water. In addition, Apple's current health and Microsoft's own efforts with Zune and the Xbox tend to indicate that maybe Apple was right.

In the end, I think that while it clearly did work for Microsoft, had Apple followed Microsoft's lead, we wouldn't have the iPod, the iPhone or Steve Jobs running Apple. So we might not have Apple either. Let's now apply this to ClearCube, which just split into two companies, one focused on hardware and one on software.

ClearCube: The Most Amazing Technology You've Never Heard Of

ClearCube is the pioneer of blade PCs, first to market with a solution largely based on the blade server model but applied to desktop computers. The solution had a number of advantages: It was more secure, more reliable, quieter (in the workspace), and potentially more cost-efficient over a long period if PC costs remained reasonably stable.

In its simplest form, ClearCube supplies a solution that provides the flexibility of a PC with the reliability of a mainframe. Had this technology entered the market a decade earlier, there is a reasonable chance the PC market as we now know it wouldn't have existed. But it didn't and, as a result, the market had already locked down on a less efficient distributed technology model.

The reason that ClearCube, and blade PCs in general, didn't become the new standard included the lack of cross-vendor standards for the blades themselves, the risk of something new and different, the size of ClearCube as a company, and the initial cost and risk of the solution. Once competitors entered the segment -- both HP and Hitachi compete in this space -- the lack of standards became more pronounced and all three vendors were more successful showcasing their differences than in coming together to drive this new platform. This lack of standards is probably limiting the existing market for blade PCs to less than one-fifth of its potential.

Finally, with a huge focus on mobile computing, blade PCs seemed out of touch, even though they could potentially be the best cloud computing solution in the market.

Why Separation May Work for ClearCube

There are two components to a good blade computing platform. The first is the hardware, which must perform at PC levels for both processing information and for graphics. The second is the software that manages the blades and creates the capability for failover if a blade fails, or to dynamically assign blade resources to the user as needed.

These same tools can be used to manage traditional PCs and are arguably more mature than most of the tools they might have been able to compete with. But, because ClearCube was seen as a vertical integrated hardware company, its desktop management solution could not compete, let alone win, against these less mature alternatives.

On the other hand, companies that have already deployed advanced desktop management solutions are unwilling to mix software, even though blade hardware might actually be preferred to traditional PC hardware. So, because the hardware was hard-linked to the software, it couldn't be sold into organizations that might otherwise find it useful.

The theory then is that, separately, both sides are better able to take advantage of existing opportunities and move on them. Together, one side crippled the other.

Another path is one of acquisitions, and we appear to be in an acquisition and merger period. As separate entities, the parts will be attractive to firms that otherwise would have seen conflict in either the software or hardware components. In the end, this could end up being the final path for either or both of the now separate units.

The Apple/Palm Test Revisited

If this is successful, and it should be, it will showcase a framework around which future decisions to hold units together or split them apart can be better analyzed. But it would appear that it depends on the market maturity of the competitors of the components to be separated. If they are immature but deployed, then separating the company may result in more revenue more quickly. If they are immature but not deployed, then keeping the company together as a solution may allow the firm to have a sustaining competitive advantage if the result is less complex and less costly than the alternative. If the competitors are both mature and distributed, the firm is in trouble together or apart.