Pete, <<<the aggregate is called server revenue. Services are things like figuring what is needed to do a task, what software to buy, what hardware to buy, how to connect the various pieces together, integrating the software, the hardware, and the business systems into one seamless whole, and lastly figuring what went wrong and fixing it. >>>
You make a very articulate statement concerning the complexities of Information Technology, you show in-depth knowledge of the components that make up this industry, and you are quite aware of the relationships of these components. You should have stopped right there, when you were ahead.
Yes, the people in the Information Technology Industry have long known the costs involved. The conventional wisdom over the last twenty years is that software and services are 8 to 10 times the cost of hardware. And, the software industry is indeed quite large. During the Y2K hubbub, it was variously estimated that there were 600 billion lines of code (LOC) that needed Y2K remediation at a cost of $1 per LOC. The cost to produce the original source code is many times that figure.
Yes, we agree that the costs are high and the integration of all the components are complex. We also agree that hardware is a fraction of total costs and that servers are a fraction of hardware costs. And CPU is a fraction of server costs.
Where you seem to fall apart is in your conclusions:
1. All, most or even a significant portion of integration costs are included in server costs.
2. Intel is losing market share (WRT total cost) in the fastest growing server segment and that cpu becomes less significant.
1.) We agree that the server in itself contain many components and that the cpu is only one of the components. But, it is absurd to think that all systems integration costs or any significant portions of it is included in what is described as the cost of the server.
2.) Until very recently, and to many still today, Intel is known as the company that produces microprocessors for PCs even though it has been gaining significant market share in the low-end snterprise server market over the last 4 years or so.
The Itanium platform is designed to compete in the mid and large server market. It's rationale is that it will significantly lower enterprise IT costs. But, for Intel, the revenues will be higher and the margins much better than ever. The cpu, instead of becoming insignifcant will become more important over time as it takes over many more functions and that require less human intervention.
The "bet" on investing in Intel (IMO) is how well the Itanium platform penetrates the mid and high end server markets and the leverage that it can gain. Success in this market will ensure Intel's survival over the next 50 years or so. For Intel to survive a longer period of time, Intel's investments in the "other" categories will have to bear fruit.
Mary |
