C. 3d mark 2001:

D. 3d mark 2003; by now the CPU limitation is gone and the GPU is the one which is limiting the performance of the system:

E. 3d mark 2005 - the GPU is the bottleneck now and the results are provided more for your information than for any useful purpose - 7800gs is clearly not suited for 3d mark 2005; the settings screenshot provides information about what I meant by " maximum settings":

First of all benchmarks are total waste of time, they are not representative for the real world performance in programs or games.......
If you want to have real world numbers test something like win rar (not the benchmark) or 7 zip, also tests in games, what game will show better performance, video encoding tests or editing. That will show you, not the dumb retarded benchmarks, i have seen much of them, and they NOT show the real performance, most of the time benchmarks shows intel or nvidia based hardware to be FAR superior to the AMD and Radeon hardware, but in reality the difference is like not 70% that the benchmark showed but only 15-20%.......

Other thing the KT880 is newer, BUT it does NOT have AGP/PCI lock so that made it useless. Also that new VIA chipset is not better at handling 939 greater modules, is still sucking up when you put some nice 2GB DDR500 or 600 modules, and the machine will start spit errors (if you try any higher then DDR400 speed) after some time, at least that did happen on my NF7, and i did heard from other enthusiasts that they did have KT880 and it was not any better...
Also is known that K7 processors benefits from the higher possible bus speed, but seems impossible to use more then 200FSB if the modules are 1GB per stick...
Also other thing to point, there are tons of BIOS mods and upgrades over the NFII chipset, while there are from small amount to none for the KT880, and that really matters..

nVIDIA NFII +
AGP/PCI Lock, way more BIOS mods, bigger stability, way better overclocking, sound chip and quality

VIA KT880 +
integrated SATA, newer, 4GB RAM support,

END line is even KT880 users have spoken that they are not very happy with that chipset, as they will go for NFII if possible...........
Of course im sure there are some specific places where that new chipset will perform faster then the nForce II, but still overall will be underdog no matter what....
Also all extreme OC on AthlonXP and Mobile XP-M chips are made on NFII boards (mostly ABIT NF7) with moded BIOSes...

Benchmarks are not useless they are actually fairly representative and not everyone buys motherboards to overclock certainly not for retro gear considering faster stuff is almost always readily available so stop stating your ideas as if they were facts.

If you don`t know that the benches are just useless junk is your problem, not mine. In some benchmark intel processor can be shown as to be way way faster then AMD processor, and in real world programs or games the AMD will perform not 80% slower as the bench stated, but only 20% slower.............. So if you don`t know for what you are talking about don`t tell me im wrong as im not, and you are.........

Radical Vision wrote:

If you don`t know that the benches are just useless junk is your problem, not mine. In some benchmark intel processor can be shown as to be way way faster then AMD processor, and in real world programs or games the AMD will perform not 80% slower as the bench stated, but only 20% slower.............. So if you don`t know for what you are talking about don`t tell me im wrong as im not, and you are.........

Jeez. No one is even comparing AMD vs Intel here, or ATI vs NVIDIA—did you read the first post of the thread?

The same CPU and GPU are being compared in motherboards with different chipsets: surely even you would agree that benchmarks are not skewed in such a case... 🙄

Radical Vision wrote:

Other thing the KT880 is newer, BUT it does NOT have AGP/PCI lock so that made it useless. Also that new VIA chipset is not better at handling 939 greater modules, is still sucking up when you put some nice 2GB DDR500 or 600 modules, and the machine will start spit errors (if you try any higher then DDR400 speed) after some time, at least that did happen on my NF7, and i did heard from other enthusiasts that they did have KT880 and it was not any better...
Also is known that K7 processors benefits from the higher possible bus speed, but seems impossible to use more then 200FSB if the modules are 1GB per stick...
Also other thing to point, there are tons of BIOS mods and upgrades over the NFII chipset, while there are from small amount to none for the KT880, and that really matters..

God smite ANYONE wanting to put these assertions to test for the fun of it, right? I mean, we're only in a forum about old hardware, that has no place here at all.

Take it down a notch, maybe suggest games you would like to see benched or something.

---

OP: nice work, can't wait for more results!

I would also be interested in seeing AMD's chipset in the comparison.

slivercr wrote:

Jeez. No one is even comparing AMD vs Intel here, or ATI vs NVIDIA—did you read the first post of the thread? […]
Show full quote

Radical Vision wrote:

If you don`t know that the benches are just useless junk is your problem, not mine. In some benchmark intel processor can be shown as to be way way faster then AMD processor, and in real world programs or games the AMD will perform not 80% slower as the bench stated, but only 20% slower.............. So if you don`t know for what you are talking about don`t tell me im wrong as im not, and you are.........

Jeez. No one is even comparing AMD vs Intel here, or ATI vs NVIDIA—did you read the first post of the thread?

The same CPU and GPU are being compared in motherboards with different chipsets: surely even you would agree that benchmarks are not skewed in such a case... 🙄

Radical Vision wrote:

Other thing the KT880 is newer, BUT it does NOT have AGP/PCI lock so that made it useless. Also that new VIA chipset is not better at handling 939 greater modules, is still sucking up when you put some nice 2GB DDR500 or 600 modules, and the machine will start spit errors (if you try any higher then DDR400 speed) after some time, at least that did happen on my NF7, and i did heard from other enthusiasts that they did have KT880 and it was not any better...
Also is known that K7 processors benefits from the higher possible bus speed, but seems impossible to use more then 200FSB if the modules are 1GB per stick...
Also other thing to point, there are tons of BIOS mods and upgrades over the NFII chipset, while there are from small amount to none for the KT880, and that really matters..

God smite ANYONE wanting to put these assertions to test for the fun of it, right? I mean, we're only in a forum about old hardware, that has no place here at all.

Take it down a notch, maybe suggest games you would like to see benched or something.

Seems more bull***t detected here...
When i say benchmarks shows faked scores they do. When there is test on AMD systems in benchmarks and after that tests in games or some programs the AMD processor shows WAY better resultats, meaning the benches are useless junk for uneducated people, bcuz in real world they will show different performance, and in the end the processors and videocards here will show different results on real world software, so the benches are USELESS, if you and others think other way you are all delusional..

Also the + on NFII i mentioned are gamechanger for that platform. Also when i OC Mobile Barton CPU i get 2.5-2.6GHz and guess what i will get ton of more performance on that platform, and no need to even bother testing the VIA crap chipset, that will have stability problems and other.....

Let's cool down a bit a Radical Vision.
Banned for 2 weeks.

Radical Vision wrote:

Seems more bull***t detected here...
When i say benchmarks shows faked scores they do. When there is test on AMD systems in benchmarks and after that tests in games or some programs the AMD processor shows WAY better resultats, meaning the benches are useless junk for uneducated people, bcuz in real world they will show different performance, and in the end the processors and videocards here will show different results on real world software, so the benches are USELESS, if you and others think other way you are all delusional..

Also the + on NFII i mentioned are gamechanger for that platform. Also when i OC Mobile Barton CPU i get 2.5-2.6GHz and guess what i will get ton of more performance on that platform, and no need to even bother testing the VIA crap chipset, that will have stability problems and other.....

Well, you know what you should do? Take your OCed system to hwbot and upload some benchmark scores with it to see how it compares with others using the same CPU.

First of all I must state that I will not ever overclock my retro systems! Long life and stability is far more important!
Second of all the benchmarks represent the same CPU/RAM/GPU combination and only the chip-set is different so no Intel bias here 😊 !
Now back to the test with the NFORCE2 platform represented by the Abit AN7 motherboard:

A. Code creatures pro refuses to run on the nforce2 platform regardless of the settings used! I will investigate but for now let's skip this one!
B. 3d mark 2000 - as you can see increasing the resolution will result in a lead for the KT880 platform despite nforce2 being faster at 1024*768:

C. 3d mark 2001 - nforce2 dominates this DirectX8.0 benchmark by great amounts at lower resolutions:

D. 3d mark 2003 - it's a tossup here because the video card is already bottle-necking the system:

E. 3d mark 2005 - again too close to call because the video card is overwhelmed in this test, a geforce 8800 or better is required:
I am sorry for the lack of result at the 1680*1050 resolution; it seems I forgot to make the screenshot 😊 but the results are really close, within the margin of error!

F. 3d mark 2006 tested out of curiosity and the result confirm that 7800gs is no match at this level:

PC mark 2002, 2004 and 2005 have been added to the test suite but I don't have any results at this time!
the testing shows that when the system is stressed as a whole the nforce2 has the upper hand - slight but undeniable advantage - but when the onus is on the video card both systems are really close to each other! 3d mark 2001 is a balanced test that spreads the stress evenly between system components unlike 3d mark 2000 which is CPU limited on both systems and 3d mark 2003, 2005 that is GPU limited on both systems; I think I will have to redo the test again for the 4 time on the VIA KT880 platform to see if I can get better results because the system was online during this particular test!
Personal impressions:
Pluses for VIA KT880: for me the KT880 is far more easy to setup and less fussy about which components I use than nforce2; every single module of RAM works out of the box on VIA but on NVidia only the Corsair works without errors and hard locks; VIA also recognizes any SATA hard disk connected without problems but on NVidia there is a hard limit at 500 GB on the capacity that gave me problems when first setting up the system; the system runs cooler by about 5-10 degrees on VIA and the north-bridge/south-bridge is barely warm where on nforce2 I have to use active cooling on the north-bridge case fan to cool the south-bridge.
Pluses for Nvidia nforce2: I have surround sound using Altec Lansing 5.1 system and nforce sounds clearly better and the sound positioning in games is better - of course this is subjective as I don't know any test for sound positioning and accuracy - in Far cry 1 and Doom 3 the sound is awesome on nforce2; performance is better when playing old games using DIrectX 8/7 such as NFS series, Undying, Quake 3, Red faction.
To be continued... 😀

I actually tend to agree with Radical Vision on this one. I've been thinking about it too lately. Intel has been known for playing dirty in the past, so it won't surprise me if all major benchmarks are optimized to run best on Intel's processors.

Dreamer_of_the_past wrote:

I actually tend to agree with Radical Vision on this one. I've been thinking about it too lately. Intel has been known for playing dirty in the past, so it won't surprise me if all major benchmarks are optimized to run best on Intel's processors.

Well, maybe then you could speak on his behalf and explain to us what that has to do with testing the same AMD CPU on different chipsets using synthetic benchmarks?