Printable View
The fact what you're oposing Xeon, 64 FX and P4 Extreme Edition suck for me. Those processor are for regular user machine. Those are just technology showoff.
Most of my reference website are in french but here some.
http://www.hardware.fr/articles/496/page12.html
http://www.materiel.be/cpu/2004cpu/page8.php
The graphic should talk for themself.
The xeon is used in 2+ cpu systems. You cannot compare the xeon to a p4 or 64 FX since neither of these processors can run in a system with more than a single processor. If you want to do a price comparison, compare similar processors. The price of an atlon mp that runs at 2.2 Ghz (2800+), and the cost of a xeon that runs at 2.2 Ghz is the same, $200.Quote:
Xeon cost 1000$ and Athlon 64 cost 200$. Is Xeon x5 faster that an Athlon 64?
I fail to see your point from the graphs. I understand french enough to get the gist of what the article was saying, but I do not understand how you get that the Intel processors are so much worse for gaming from those two graphs. In all of the graphs shown, the processors were all within a few percent performance region, with the P4 ahead some time and the FX 53 others.
The 'CPU test' portion of 3d Mark has always been a joke. It has never reflected anywhere else. The 3dmark processor test showed that the FX 53 was ahead by 5 fps (90 vs 95) on one 'cpu' test and then on the other, it fell behind by .6 fps. (13.2 vs 13.8)
A much better way to benchmark processors for gaming is to bench two systems both with the exact same video card. Then compare the results when both systems had the exact video card.
From the link I posted previously, all systems were fitted with the exact same Radeon 9800 XT, and then 3dmark 03 was ran on all of them. Across the board there was a 10% range in graphic performance, with both the FX 53 and the P4 3.4 at the top within less than a 1% variation.
source:
http://www6.tomshardware.com/cpu/200...n-fx53-17.html (second graph)
*Note in this link there is also a black bar that shows the results of an FX 53 overclocked to the max, the article was written specifcally for the FX 53 and results of overclocked P4's are not included*
Same Review... Game Wise!
http://www.hardware.fr/articles/496/page10.html
I have 8 AMD Processor who are better or egal score in Unreal Tournament 2003 that the best Intel Processor.
I have 6 AMD Processor who are better or egal score in FarCry that the best Intel Processor.
http://www.hardware.fr/articles/496/page11.html
I have 6 AMD Processor who are better or egal score in Warcraft III and in IL-2 Sturmovik that the best Intel Processor.
http://www.materiel.be/cpu/2004cpu/page12.php
I have at least 4 AMD Processor who are better or egal score in Unreal Tournament 2003 that the best Intel Processor.
I am not familiar with IL-2 Sturmovik. Unreal Tournament 2003 is a game that I know well. As I previously admitted, the Athlons, especially the FX-64's, are better with Directx8 and OpenGL instructions than the Intels. Unreal Tournament 2003 and WarCraft 3 are both Directx8. If you find a benchmark done for Quake (1, 2 or 3), Half-Life 1 or Counter-Strike, I would agree and be willing to bet that an AMD chip will beat the Intel for those games as well. Looking at those scores though will show us that we our comparing in the hundreds of frames per second.
It may be just my opinion, but I believe once a game is well over 30-50 frames per second, and then even excedes the refresh rate of the monitor, it does not matter what it's exact score is.
For all of the games that are out now, and the games that will be out soon, Unreal Tournament 2004, Half Life 2, Doom 3, Painkiller, Manhunt, X2 The Threat, and recent benchmarking programs 3dMark 03, PcMark 04, et cetera, both AMD and Intel score very similarly.
Far Cry is a very interesting game. It is Directx9 based. There are 2 major versions of it out now. The original was written for Directx9b; the patch that was relased included Directx9c instructions, which mainly just included Shader Model 3.0.
As for the site that you posted, it showed that 6 AMD processors beat out Intel for the FarCry 'benchmark'. I thoroughly read through that entire article, as well as several other on that site and was unable to find any more information on this 'benchmark' that they did. The interesting point that I am attempting to bring up is that there is no official way to benchmark FarCry. That is why it is rarely used as a benchmark program. There are no applications that use its engine to reproduce the same run over and over. The only thing that I can imagine is that possibly they used the FarCry demo intro that shows no gun fire or fighting what so ever, as well as an extremely small amount of any textures, sand, water, lava, jungle, and no vehicles. Even for this demo, I have seen people use cards such as Radeon x800 XT's and Geforce 6800 Ultra's, and 80fps+ is not common.
Not only is it uncommon to achieve those high rates that they showed, I find it hard to believe that a few hundred Mhz difference in processors affected game performance that substantially in images per second. When I overclocked my processor's core by 35% (from 2.8Ghz to 3.8Ghz), I only noticed an increase of about 5% maximum on any games and benchmarks that I used, (no where near the 40% change that they show). Any game benchmarks hardly changed for me until I changed the core clock and memory clock of the graphics card.
If you could find out any more information about how they did this 'benchmark' on FarCry, I would be very grateful, as I have never seen such results anywhere else. I will continue to search their site and references to their site, but I am doubtful in my chance of finding anything else.
Ding! Dong!..............end of round three :D
I have a few comments to make, that I hope will slightly introduce other aspects than the purely technical?
Firstly, my position is very crude and very simple................if it does the job I go on cost..£$€, that is the deciding factor, surely it has to be. I use both AMD and Intel, and must say that I have NEVER had a bad experience with either. Surely, if they were that bad, they would not sell? If one were so superior to the other, the "other" would go out of business.
I still say that the secret lies in balanced components, and the purpose for which the system is required.
In my view, it is a bit like intelligence tests............they mostly show how good you are at intelligence tests, not how intelligent you are. In a way it is a bit like those crossword puzzles (motes croisses?)....................you may be very good at them, but can you articulate yourself like a confidence trickster with only a grade level education?
I am typing this on a machine with an Intel MoBo and PII/266 processor.............why?.............because it does the job. On the other side of the room I have an AMD 1900+ and an IBM with a PIII/533...............they have 1024 and 512 of much faster memory than this one's 128 Mb of 72 pin EDO RAM :D
I use this machine for general internet browsing, and typing letters etc.
I have a Digital Venturis PI/133 that will work as well............except that my ADSL provider's software won't accept it.............yet it benchmarks to within 12% of this one (OK, I admit that I did that as a "project").
Let's face it, gaming and graphics are pretty specialist, and in my opinion (based on serious commercial graphics Apps) are less dependent on miniscule processor performance differentials, as memory and the actual graphics card.
Let my epitath be:
"He loved the technology, but he hated the price"
:D
Usually, AMD are cheaper that Intel.Quote:
if it does the job I go on cost..£$€, that is the deciding factor
See the USUALLY?
AMD are usually cheaper than Intel for similar performance. I know that if you compare the processors side by side, Intel will be more expensive most of the time.
However, there are great exceptions on either side. Which is why I never compared prices, because I believe them to be very similar as well.
My 2.8C M0 SL6Z5 for instance, which is a gallatin 3.4EE just labled as the 2.8, only cost me $200. With the stock cooler I could hit 3.2 before it got too hot, with the new swifty MCX478-V I put on there, I can now hit 3.8 with it. A 35% increase in performance from stock is great, especially considering how cheap the processor was. Now it is higher than even the newest LGA Intel processors that only hit 3.6 with an 800Mhz bus. A 3.4EE still costs rougly $1000. Mine is perfectly stable, I have been operating it for months and have had 0 crashes related to hardware, even though I push it and try to get it to. It is stable running Prime95 for 24+ hours any test.
So sure you can say that AMD are usually cheaper than Intel. But with great exceptions on either side, I have read several articles on athlons that clocked up very high very easily, it is hard to purchase either one based solely on price or make your decision on one brand because of price.
There is not a lot of user that overclock their machine. I don't consider that fact when buying a CPU.
Hey guys,
This is a really great thread, and I am enjoying it. Now I have a problem.............I would tend to prefer an Intel for over AMD for an overclocking competition......it has been a while but I took a P4 /2.26 and stoked it up to 3.066...................I wanted to break the 3 Gig barrier.
At the same time I built the Wife (notice the capitalisation) a Duron 1.3............I spent the money on software, because that is what she will and does use.
I have great affection for the 1.3 Duron........solid as a rock............I am not well impressed by Celerons though, as I have seen too many problems with them.
Some of us are lying in the gutter and looking up at you guys playing with the stars (appologies to Oscaar Wilde)
Cheers, and a good week-end (fin de la semain)
:D
I was not happy with the Prescott core. I bought a 3 GHz with 1 meg of cache and had a horrible heat problem. I bought an AMD 64 3000+ and never looked back. nuff said.