For p4 owner!

[jhalada]

Wingznut PEZ:
sorry about the tone of my message.
I agree with you that you are not going to notice less 10% difference, but it doesn't mean the benchmarks are useless. The benchmarks links I posted to Anand's review of P4 are not synthetic. None of them. These are real apps. The Winstone, Sysmark and Office type benchmarks all test real life apps from Microsoft / Lotus / Corel Office apps to Photoshop, Corel Draw, rendering in browsers etc. This is represents typical stuff you will come across in the office, and performance in these apps predicts very well the performance of other business related custom developped apps.

The game benchmarks (again, I posted links to Anand's review) are also predictive of how 3D apps will perform. I wish some reviewers would review some non-3D games, like say Civilization Call to Power. My guess is that these perform similar to Office apps.

I agree about relative uselesness of the synthetic Sandra, Linpack and SPEC benchmarks.

But one thing about small differences in benchmarks is that when buying a new system, and say the difference in different CPUs is small, you can just go for the most bang for the buck at a given performance level

 

[jhalada]

Nox

You must be imagining things. I never said that Excel overwhelms P4, I said it may make 300 or so MHz CPU look slow, in particular my Celeron 366, which is no longer stable at 550 MHz, and I have to run it at rated speed to avoid crashes.

As far as P4 is concerned it is a good chip. Unfortunatelly, vs. the competition, 1 GHz P3 is about as good as P4 at half price and Athlon 1.2 GHz is faster on in most cases, again at half price of 1.5 GHz P4.

So if you are comparing performance in absece of price, it is fine, but once you introduce price / performance, it is a dog. Plus, you need to add the price of Rambus memory to the price of P4, making the price / perfomrance even worse.