Introduction

What is this article about?

This is not a CPU review, it's not an overclocking guide, and it's not a how-to build your own computer guide. In a way, it's all those things wrapped into one, but mainly this article was meant to address all those emails we get about "what kind of computer should I buy?" This article, the Gigahertz Box Project, is basically an anecdote about the computer I built a few weeks ago, and my experiences with it.

Geek Trek 3 - "The Search for Framerate"

I've been plugging away at FiringSquad with my P3-450 system for over a year now. This box is not a bad machine, but it was never a monument to speed even in its more youthful days. Sure, I borrowed extra video cards from the hardware testbed to try and prop up the framerate, but the fact of the matter is that I play many games at low resolution. Even a top of the line 64MB GeForce 2 GTS board isn't going to help me much at the lower resolutions. The core of the problem here is clearly a slow CPU. Why not overclock it to 550, you say? Well, unfortunately this box here is company issued machine, so it's using an Intel motherboard. Those aren't exactly conducive to overclocking, and at this point in time, 550 is probably still a bit too slow for a spoiled tech brat like myself. The only solution, in my mind, was to buy my own dedicated gaming box, and relegate the old P3 to the mundane tasks of Word, net surfing, Photoshop, and HTML editing.

CalBear's workstation: 2 computers and 3 monitors

Making a decision

So once I decided to put the 450 out to pasture, the task at hand was to decide what parts should go into my new rig. Normally the job would entail reading review upon review at sites like our own, and Anandtech, and such. Three years ago, that's probably how I would have gone about it. But since the Editor in Chief of Hardware (James) sits three feet behind me, I did what any lazy bastard would have done. I asked him over ICQ.

"What's the sweet spot for price/performance right now?" Knowing that I'm more comfortable putting together Intel systems, James replied that I ought to pair up a Coppermine 700E with an Abit BE6-II. If I was lucky I might even be able to crank that puppy up to 931MHz. The truth of the matter is that I didn't go into this expecting to get up to 1GHz, or even 933MHz. I would have been happy to get to 840 or 875 MHz or so. Low expectations mean less chance for disappointment. That's overclocking rule number one - don't go into it expecting anything. If you're prepared to be happy with what you bought and paid for, then anything extra you get out of overclocking becomes icing on the cake.