The "fast, faster, fastest" headline doesnt pack the same punch anymore when it comes to processors. Sure, gamers want to blast their cannons a little faster, but the rest of the people cant really care about this stuff, can they?
Intel is counting on that they can and will. The company is hoping, of all things, that video editing, and ripping and burning CDs, will drive consumers to gobble up the new Pentium 4 processors running in the gig-and-a-half-hertz range and beyond. What Intel isnt really promoting is widespread corporate adoption. This puts the home users in the interesting position of actually driving the market, which means sooner or later, corporate users will have souped-up Nvdia graphics cards so that they can play Quake really fast on their breaks.
I admit the P4 piqued my interest, but only in the most narrow sense of the phrase. I upgrade my processor every year or so, and its still not enough for the stuff I wanted to do years ago. My Pentium chugged on it, my Pentium II chugged and my Pentium III currently is chugging away on simple video projects. Meanwhile, my word processor, database and presentation graphics package dont seem to run any faster.
The P4 has streaming SIMD extensions—144 of them, to be specific—and many of them are made for video rendering. So I guess Ill go out on a limb and believe Intel, just like I have every year. I believe that application vendors will develop their wares with the new extensions in mind. It might take a year or so, but it will be worth it.
In the past, however, I used to think my demands for faster performance really mattered. Now, the only reason I—or anyone else, for that matter—want a faster processor is for hobbying. Therefore, the P4 cant have nearly the corporate impact that any of Intels previous processors have had.
One reason corporate users should avoid the P4 is that it wont make office applications faster. As much as Microsoft is pushing voice recognition in Office 10, very few people are going to use either Office 10 or voice recognition.
Another reason to avoid the P4 is that it uses RDRAM. My quick sampling of hardware vendors at Comdex led me to believe that RDRAM will be dropped from most of its platforms and that Rambus will lose its lawsuits.
Finally, corporate users shouldnt choose the P4 because its far too expensive, and it may stay that way. There are timing problems with RDRAM right now, and motherboard manufacturers cant maintain adequate yields yet.
Right now, the P4 and RDRAM are bound together. As long as this is the case, this beast is best for home users.