i find it disturbing...

[drifter106]

after reading so many posts and benchmark results (and the insueing follow up discussions) that there is no clear cut indicator that one chip is truely out performing another. That is, one chip standing on its own and being tested against the merits of the competition.
My contention is this, any true investigation needs a control and a variable. In the data that I have seen, chips are tested with one being overclocked or having a different motherboard, ram is changed and anything else that you might want to do to change the outcome.

In the end its all about $ and power... isn't that what the real world is all about?

Maybe I am looking at it from a different perspective, but in my way of thinking any experiment consists of two things being compared and one variable changed and the results are compared to a standard. It's just not happening that way.

Granted yes, consumers want to know...manufactures want to know... hell, even i want to know how the chips perform but still its not a pure test of performance...

p.s.

just had to write that after reading some of the posts here where it was obvious that certain people wanted certain results to support there canidate...

hey ...
don't forget... opinions are like assholes...
everybodies got one...
so do i...

 

[Johanthegnarler]

Any variable you have to compare will be biast to either or AMD or Intel.

<A HREF="http://arc.aquamark3.com/arc/arc_view.php?run=277124623" target="_new">http://arc.aquamark3.com/arc/arc_view.php?run=277124623</A>
46,510 , movin on up. 48k new goal. Maybe not.. :/

 

[jihiggs]

its impossible to have a constant motherboard between the two platforms.

wpdclan.com cs game server - 64.246.52.144:27015
now featuring (optional) cheating death!

 

[pauldh]

Seems in CPU's you just have to take the best platform for each processor and compare the one as a whole to the other. But definately there is no clear winner now. It all depends on what benchmarks/uses are important and which are not. It was easier when the P4 3.0C and i875p was introduced, as it clearly was the top performing cpu in (almost) every test. Now you can't claim one is better without including better in what (value-XP2500+, gaming-A64, media encoding - P4, etc.)

ABIT IS7, P4 2.6C, 512MB Corsair TwinX PC3200LL, Radeon 9800 Pro, Santa Cruz, TruePower 430watt

 

[drifter106]

very true, taking it a step further, it could be said that it is a test of systems rather than a single integer dependent on factors within the system.

 

[Mr_Nuke]

To have 2 perfectly identical tests, you need the same mobo, wich is abviously impossible when testing 2 different platforms. Some peolple say use the latestst mobos that appeared when the CPU appeared, others say use 2 equivalent versions of the mobo (same max mem, chipset, max CPU freq or equivalent, same vendor, same BIOS settings) and both say use identical cards and pheriperals,(fact that's not so impossible). My advice is: be a mobo engineer and either get the closest performance mobos or make them yourself(if you have acces to a prduction line).

 

[hogfather]

Its true there is no clear winner, but you can still see those which stand above the others. As yoou say, the real issue is price / performance ratio. No one's buying P4 EEs because they're stupidly expensive for their relative power benefits. On the other hand, a p4 c 2.8 is the dogs bollocks, as is the 2500 xp for budget (as said by paulh).

Another point I think lots of hardware sites misss (and its not a popular view with manufacturers I suppose) is that if you're compressing a huge file to .rar, for example, you don't sit there watching it thinking "dammit, if i'd bought a p4, this would have taken 25 seconds less time", you leave it and come back. So while AMD or Intel may pat themselves on the back for their chip being better at certain tasks, in the real world whose noticing?

The only real exception to this is perhaps gaming and rendering, but on a normal day to day average joe user basis, I don't see that either chip wins, or needs to - whatever you buy these days is llightening fast, and I would think it a bit anal to be rueing losing 5% to the latest chip when it comes out.



XP2000, 512 ddr 2700ram, GF4 MX440, XP Pro

 

[Mr_Nuke]

Even in gaming 2 or 3 extra framerates is really not noticed, and hell!!! Hyper Threading and Hyper Transport bring almost the same performance increase, just that the first acts only within thw CPU and the second one acts on the edge(not phisical) of the CPU. So, really other things make the difference, like ammount and speed of RAM(dependent on the FSB), video card. But comparing 2 equivalent CPU's cannot be done accurately, but who cares, the difference is only 100 3D marks out of thousands. And as I stated, they are EQUIVALENT CPU's, so it's almost like they're meant to have the same performance.

 

[Kanavit]

I prefer Intel. Their chips are higher quality. I can just feel it when I handle them. that's why AMD don't use heatspreaders, because their cheap.

-------
(2x512mb) 1GB DDR333 Dual Channel
INTEL Pentium 4 2.8B
ATI RADEON 9800 PRO
Asus P4P800 Bios 1016
38,476 Aquamarks

 

[ChipDeath]

AMD do use heatspreaders now. And that's a flamebait post if ever I saw one...

---
Epox 8RDA+ rev1.1 w/ Custom NB HS
XP1700+ @205x11 (~2.26Ghz), 1.575Vcore
2x256Mb Corsair PC3200LL 2-2-2-4
Sapphire 9800Pro 420/744

 

[jihiggs]

i give that flamebait a 3. try again.

wpdclan.com cs game server - 64.246.52.144:27015
now featuring (optional) cheating death!

 

[Spitfire_x86]

I usually don't respond to flamebaits, but you need to know one thing.

The shiny metal thing on the CPU core is not heatspreader, it's core protector. It's completely useless for cooling, just adds protection to the CPU core

------------
<A HREF="http://geocities.com/spitfire_x86" target="_new">My Website</A>

<A HREF="http://geocities.com/spitfire_x86/myrig.html" target="_new">My Rig</A> & <A HREF="http://geocities.com/spitfire_x86/benchmark.html" target="_new">3DMark score</A>

 

[Zoron]

It increases the surface area in contact with the HSF. Not exactly useless. More heat can be removed with the heat spreader than without it.. why do you think Intel integrated it onto their P4 chips?

<font color=red> If you design software that is fool-proof, only a fool will want to use it. </font color=red>

 

[Cybercraig]

Ohh, absolutely! Just look at the brand name!
INTEL = INTELligence!
AMD = Another Melted Device!
LOL! Flame on!


"I am become death, the destroyer of worlds. Now, let's eat!

 

[endyen]

The core protector has poorer thermal transfer than a copper hs. It doesnot make the die larger, only seperates it from the hsf. If Intel were to remove it from the P4e, the chip would run cooler.

 

[P4Man]

>why do you think Intel integrated it onto their P4 chips?

I think it serves two purposes:
1) protect the core from crushing/cracking
2) with mediocre or bad heatsinks, badly installed, no thermal compound (or too much or whatever), it would most likely help decreasing the temps.

However, with a good HSF, a heatspreader will only increase thermal resistance, as it adds another layer, and one that is not likely as good as a well installed, quality HSF.

Also, I'm not sure about mobile P4's, but I've seen mobile A64's come without the heatspreader. Why do you think that is ? IMHO, its to decrease thermal resistance (much needed in a cramped notebook) and protecting the core is non issue, since no one builds there own notebooks.

Maybe some hardcore overclockers can comment on this ? I'm sure some of them have ripped off the heatspreader, I'd be curious to see if that increased or decreased core temps. I'm fairly certain if anything, it would have helped to decrease them (especially with the top notch cooling they ought to have).

= The views stated herein are my personal views, and not necessarily the views of my wife. =

 

[P4Man]

How do you feel about buying a car ? You find it disturbing that "not one car is truly out performing another" ? Each car has its own strengths and weakness, and some cars will match your needs much better than the other, because you have different requirements. And for $100k you're more likely to find a car that fits your wishlist than for $20k.

Pretty much applies to cpu's, or videocards, motherboards, even harddisks. Performance is not one single criterium, its a spectrum

= The views stated herein are my personal views, and not necessarily the views of my wife. =

 

[Johanthegnarler]

lol

<A HREF="http://arc.aquamark3.com/arc/arc_view.php?run=277124623" target="_new">http://arc.aquamark3.com/arc/arc_view.php?run=277124623</A>
46,510 , movin on up. 48k new goal. Maybe not.. :/

 

[phial]

I prefer Intel. Their chips are higher quality.

oh youve visited both manufacturing plants eh? got inside contacts?



my Athlon Palomino has been overclocked from 1.4 to 1.75ghz, wiht the vcore raised to 1.9v up from 1.75v for 1.5 years. that cpu is COOKING


yes, thats really low quality! and that is a cpu made in 2000. their process has only gotten so much better that the same core with MORE CACHE is hitting almost 3ghz now .. wow what crappy quaity that is

-------
<A HREF="http://www.albinoblacksheep.com/flash/you.html" target="_new">please dont click here! </A>

 

[scottchen]

I agree with ya, both manufactures has high quality. Just that for me, a gamer, P4 at 3.4Ghz seems to have to struggle against Athlon64 3200+(2000mhz), now with this thing clocked at 2300mhz, i bet in most games i could compare against 3.8Ghz or more.

-AMD Athlon64 3200+@2300mhz -GAK8N -3x512 Hynix PC3200 ram@192mhz 2.5-3-3-7 -Albatron 5900PV @ 509/1000 -SB Audigy -Adaptec 1210SA Raid -2x120G Seagate SATA150 7200.7 Raid0 -Enermax 460watt psu

 

[phial]

yep ^_^


but its **INTEL**

look! ^ its the word "INTEL". you must all drool and swoon like the sheep you are over a name

-------
<A HREF="http://www.albinoblacksheep.com/flash/you.html" target="_new">please dont click here! </A>

 

[blackphoenix77]

Well why isn't god petrieved as black, and a woman? I have studied Dogma and thats all I've gotten out of it. I also learned how to make a "Holy Bartender" :lol: I hate my Fx5200

<A HREF="http://kevan.org/brain.cgi?michael" target="_new"> You Want To Click Me, Go Ahead </A>

 

[justaguy]

I agree with the original thread on a purely theoretical basis. That being said, you will always need RAM, mobo, etc. to use your CPU. So, in order to make the CPU function, reviewers try to choose comparable components (high-end AMD and high-end Intel) and let you know how the systems perform in comparison with one another.

Not to mention, when one reads said theoretical review, are you curious about the literal performance of the CPU? More likely, the reader will need to discern which platform they would choose to purchase. In that context, a platform (mobo, memory, etc.) is a very valid component to include. Being that you cannot test a CPU without other components and the results of any system are always interrelated and synergistic, a CPU only test (even if it were possible) would be useless.

To summarize, a review may not measure strictly the performance of the CPU, but that's not the information anybody wants to begin with. It's not realistic to consider whether A64 is better/worse than Prescott and ignore the differences in the available chipsets, etc.

Slightly cliche, already mentioned car analogy applies. If you want a fast car, would you choose the car with the most horsepower? The car with the most torque? Or the car that gets around the track you'll be racing on the fastest? Logically, you pick the one that gets around the track the fastest, regardless of the raw power of the engine.

Athlon XP 1900 (11x200) 42C (Load w/AX-7 & 8cm Tornado) - MSI K7N2 Delta - Corsair Value PC3200 - Gainward GF3 @ 250/550 - 80Gb WD 8Mb Cache -