> The thing that you're missing is that today's applications
> and usage models are quickly becoming obsolete. Just
> think about the number of tasks that Windows is set up to
> do in the background, yet most people disable them because
> they're too slow.
Yeah, I'm like that. But 99.99% of the populace isn't. I have to abuse people in this office to get them to merely walk and chew gum at the same time!
> Virus checking, disk defragmenting, security checks,
> archiving... these can all be running on the background
> to enhance the user's platform stability, security, and
> experience, but most people disable them because it slows
> down the computer, and then they say that their usage
> model only includes one application at one time.
Really, though, for the most part, the OEMs decide this stuff. But there's no sign that they're going to any time soon be doing defragging in the background, for instance. This is a Linux thing, not an XP thing. Maybe Intel is looking five years down the line, but there won't be any current processors being used (or, at least, very few) that far ahead. Benchmarks should be testing stuff that will be done within the next two or three years (imho).
> Jim and others think that Intel and Bapco cook up their
> benchmarks to reflect fake situations just to look good
> against the competition, but the fact is that Intel never
> shows these benchmarks against the competition, anyway;
The competition against the Pentium 4 is the Pentium III (AMD and VIA are too small to be considered competition in this particular respect), so Intel indeed does show their benchmarks against the competition. The problem is, in this case, that BapCo made these changes without properly, publicly justifying the changes. BapCo has a history of being uncomfortably close to Intel (such as when Intel hosted BapCo's website last year), so the burden of proof should lie on them, not on us, to show that their changes reflect anticipated future computer use. Unfortunately, instead of analyzing this benchmark for validity, the various review sites added it to their test lists with nary a word about it. I don't think that Intel "controls" BapCo, but it is still BapCo's responsibility to demonstrate their product validity in good faith.
> Intel simply wants to create new usage models to get
> people to upgrade their computers, and if the Pentium 4
> offers an experience in these new usage models, it makes
> sense to market that!
If that's so, then they should be adding more database oriented benchmarks to suites. Best I can tell, the closest thing to an up and coming fad we have is the whole file sharing thing, and that takes up lots of resources. But encoding an asf while making a presentation on Powerpoint? I'm not sure that this reflects the near future at all.
> I think the Pentium 4 is far better equipped than
> the Celeron in handling new kinds of tasks.
That's a very subjective statement. If we ignore the memory bandwidth advantage, then I am not sure that I agree with you. MMX is finally in compilers and is getting into binaries. And P4 suddenly halves the max throughput of MMX compared to an equally clocked Celeron. P4 has smaller data cache than a Celeron. It has L2 cache that on tests appear to have higher latency than the L2 on the Celeron. Celeron can decode more instructions per clock than Pentium 4.
Does this sound like a processor design that is uncontestably better made for the future? Or is the only real benefit to the memory heavy assumptions of times to come just the higher frequency and bigger memory bus?
The answer to this is far more complex than what I say here. There is no objective "new kinds of tasks", and you're taking it with too much presumption, in my opinion. I think the Pentium 4 is a better performer than the Celeron in certain types of tasks that may reflect future uses of computers.
> Benchmarks should reflect that, and they do, but some
> people like to pooh-pooh them, simply because competitor's
> products don't perform as well. It seems to me that these
> people are missing the point, and regressing continually
> back to the AMD vs Intel debate.
Well, in fairness, both sides do this. But you have to remember what I said above. Keep in mind that just because competitor's products might not perform well also does not automatically mean that the benchmarks are a valid description of future computing behavior.
> While AMD's gains in market share have removed growth
> from Intel's bottom line, it's not nearly as much as the
> PC market's loss in demand, which has been the primary
> reason why Intel has lowered prices so much
I think that the crappy economy has hurt AMD and Intel more than each other. I am in complete certainty that both companies would be profiting well right now has the economy not started acting so stupidly. |
