Ted,
ree: "However I wonder with the focus so much on speed, that we are missing or underplaying the significance of other uses for the chip. It well could be that the chip will work best with applications that need more than speed, but rather a lot of power....like video streaming."
Exactly what I'm thinking. Does anyone need more MHz to run MS Office programs? I don't think so. Do todays PC's handle the growing applications such as digital photography, video editing, video streaming, more realistic games, etc as well as they could, no.
So if you are the chip designer, what are you going to design for, the applications that work just fine, or the growing ones that need improvement? Taking it to the next level, if you are an end user, are you going to buy a computer that runs Office incrementally faster, of one that runs graphics twice as good? In this context, my guess is that most of the current benchmarks become meaningless.
One of the implications of this is that if Intel does a 90 degree turn on it's functional benchmarks for PC's, does AMD follow? If they don't they run the risk that the folks that write application will do so with the capabilities of the P4 in mind, and they won't function well on AMD machines. Conversely, Intel runs the risk that the application guys won't design for P4 strengths, and AMD takes the lead. Also, Intel pays the penalty for being ahead of the optimized applications.
I think you get it that we're not talking about the fourth quarter or even most of next year, but what the PC will be doing for us in 2002 and beyond. I'm certain that Intel is thinking in a long term timeframe.
John |
