Steve Ballmer says Microsoft is seriously considering making more of its own hardware. It seems that even with the old guard of IBM, Dell, Toshiba, Sony and HP, and the new guard of Samsung, Lenovo, LG and Asus, and all those white-box manufacturers, Microsoft just can't find a reliable design partner.

Have we come full circle to where we started 50 years ago, when hardware and software were inextricably intertwined and necessarily provided by the same company?

Looking at the actions of some of the world's largest software houses, you could easily come to that conclusion. Microsoft now makes Surface, and it has made the Xbox for years. Google bought Motorola's phone division, probably for its patent portfolio but also to make the Droids it thinks people want. Google also makes the Chromebook. Even Oracle had to have its own hardware division, though the performance of Sun doesn't bear out the wisdom of that move. Its real play is purpose-built machines such as Exadata and Exalogic.

For the moment, put aside the question of whether the Oracles of the world are doing the right thing and focus on the trend as it relates to end users. Such moves start with Apple envy. After all, if Microsoft works with partners such as Intel or an ARM licensee to create reference designs, it would be hard to believe that manufacturers wouldn't help make prototypes.

Microsoft, Intel and system makers have collaborated to good effect in the past. Ultrabooks are a prime example: While they're not revolutionary, Ultrabooks address most every problem users have with laptops, from weight to battery life to processing power to boot time. They're a major improvement over last-generation notebooks and netbooks. Intel produced the reference designs, and a few months later systems showed up from a variety of manufacturers.

So what's Microsoft's problem with that model? The biggest problem is that it's not what Apple does, and Apple's share price has outperformed Microsoft's consistently.

It's not that Microsoft's products aren't good enough and that better hardware integration is the answer. The problem is that they're not good enough to displace entrenched products tied to an ecosystem investment, namely, Apple's iTunes and App Store. Without some fundamental added value, Microsoft's tablets are likely to go the way of the Zune. That's not because Zune wasn't a solid product; it might have been terrific, I never tried it. And the reason I never tried it is because I have an investment in iTunes and I know the iPod/iPad/ iPhone interface for playing music.

What's Microsoft's big value add? For businesses it might be compatibility with Office. For consumers I have no idea what it would be -- Xbox compatibility? This Apple-Microsoft dynamic reminds me of Ford's breakthrough with the Mustang. Chevy was a close follower with the Camaro, and both were big sellers. Dodge made the Dart and Charger, both technically great cars that never really caught on. Microsoft's third to market here. It had better have something more important than its Metro technology if it wants to compete, no matter who makes the hardware.