The differences are minor, within 5%!
Now it is the turn of the first "real"application: Farcry 1! One of the best game of 2004, highly CPU bound; it is also one of my favorites games. All settings to the MAX, geforce 7800gs is more than capable of handling AA and AF. I used HardwareOC Far cry benchmark tool; the results are displayed in Firefox so no screenshot, just the data!

AT7 with VIA KT333
The benchmark started at 03.05.2023 06:16:36

System Information
Operating system: Microsoft Windows XP
System memory: 2,0 GB
CPU: AMD Athlon(tm) XP 2800+
CPU speed: 2080 MHz
Sound system: : Realtek AC97 Audio
VGA Information
Graphics card: NVIDIA GeForce 7800 GS
Memory: 256.0 MB
Current GPU speed: 375 MHz
Current memory speed: 1200 MHz
Driver version: 6.14.0010.9371 (English)

Resolution: 1024×768
Ultra quality option, Direct3D renderer
Level: Volcano, demo: hocvolcano.tmd
Pixel shader: default model
Antialising: 8×
Anisotropic filtering: 16×
HDR: disabled
Geometry Instancing: disabled
Normal-maps compression: disabled

Score = 54,08 FPS (Run 1)
Score = 52,63 FPS (Run 2)
Score = 52,79 FPS (Run 3)
Average score = 53,16 FPS

Resolution: 1280×1024
Ultra quality option, Direct3D renderer
Level: Volcano, demo: hocvolcano.tmd
Pixel shader: default model
Antialising: 8×
Anisotropic filtering: 16×
HDR: disabled
Geometry Instancing: disabled
Normal-maps compression: disabled

Score = 53,32 FPS (Run 1)
Score = 51,75 FPS (Run 2)
Score = 52,09 FPS (Run 3)
Average score = 52,38 FPS

Resolution: 1600×1200
Ultra quality option, Direct3D renderer
Level: Volcano, demo: hocvolcano.tmd
Pixel shader: default model
Antialising: 8×
Anisotropic filtering: 16×
HDR: disabled
Geometry Instancing: disabled
Normal-maps compression: disabled

Score = 50,90 FPS (Run 1)
Score = 49,82 FPS (Run 2)
Score = 51,01 FPS (Run 3)
Average score = 50,57 FPS

Second is AT7-MAX 2 with VIA KT400:
The benchmark started at 12.05.2023 17:09:47

System Information
Operating system: Microsoft Windows XP
System memory: 2,0 GB
CPU: AMD Athlon(tm) XP 2800+
CPU speed: 2080 MHz
Sound system: : Realtek AC97 Audio
VGA Information
Graphics card: NVIDIA GeForce 7800 GS
Memory: 256.0 MB
Current GPU speed: 375 MHz
Current memory speed: 1200 MHz
Driver version: 6.14.0011.7519 (English)

Resolution: 1024×768
Ultra quality option, Direct3D renderer
Level: Volcano, demo: hocvolcano.tmd
Pixel shader: default model
Antialising: 8×
Anisotropic filtering: 16×
HDR: disabled
Geometry Instancing: disabled
Normal-maps compression: disabled

Score = 50,73 FPS (Run 1)
Score = 49,94 FPS (Run 2)
Score = 50,52 FPS (Run 3)
Average score = 50,39 FPS

Resolution: 1280×1024
Ultra quality option, Direct3D renderer
Level: Volcano, demo: hocvolcano.tmd
Pixel shader: default model
Antialising: 8×
Anisotropic filtering: 16×
HDR: disabled
Geometry Instancing: disabled
Normal-maps compression: disabled

Score = 51,96 FPS (Run 1)
Score = 51,28 FPS (Run 2)
Score = 51,91 FPS (Run 3)
Average score = 51,71 FPS

Resolution: 1600×1200
Ultra quality option, Direct3D renderer
Level: Volcano, demo: hocvolcano.tmd
Pixel shader: default model
Antialising: 8×
Anisotropic filtering: 16×
HDR: disabled
Geometry Instancing: disabled
Normal-maps compression: disabled

Score = 52,06 FPS (Run 1)
Score = 51,08 FPS (Run 2)
Score = 51,92 FPS (Run 3)
Average score = 51,68 FPS

Third is the KD7A with VIA KT400A:
The benchmark started at 18.04.2023 16:48:36

System Information
Operating system: Microsoft Windows XP
System memory: 2,0 GB
CPU: AMD Athlon(tm) XP 2800+
CPU speed: 2100 MHz
Sound system: : Vinyl AC'97 Audio (WAVE)
VGA Information
Graphics card: NVIDIA GeForce 7800 GS
Memory: 256.0 MB
Current GPU speed: 375 MHz
Current memory speed: 1200 MHz
Driver version: 6.14.0011.7519 (English)

Resolution: 1024×768
Ultra quality option, Direct3D renderer
Level: Volcano, demo: hocvolcano.tmd
Pixel shader: default model
Antialising: 8×
Anisotropic filtering: 16×
HDR: disabled
Geometry Instancing: disabled
Normal-maps compression: disabled

Score = 54,23 FPS (Run 1)
Score = 53,30 FPS (Run 2)
Score = 54,01 FPS (Run 3)
Average score = 53,84 FPS

Resolution: 1280×1024
Ultra quality option, Direct3D renderer
Level: Volcano, demo: hocvolcano.tmd
Pixel shader: default model
Antialising: 8×
Anisotropic filtering: 16×
HDR: disabled
Geometry Instancing: disabled
Normal-maps compression: disabled

Score = 54,67 FPS (Run 1)
Score = 53,90 FPS (Run 2)
Score = 54,52 FPS (Run 3)
Average score = 54,36 FPS

Resolution: 1600×1200
Ultra quality option, Direct3D renderer
Level: Volcano, demo: hocvolcano.tmd
Pixel shader: default model
Antialising: 8×
Anisotropic filtering: 16×
HDR: disabled
Geometry Instancing: disabled
Normal-maps compression: disabled

Score = 52,08 FPS (Run 1)
Score = 50,80 FPS (Run 2)
Score = 52,33 FPS (Run 3)
Average score = 51,73 FPS

Again the differences are small; there is no clear winner! By now the trend should be obvious: today with powerful CPU/GPU combinations the performance advantage by each successive chipset from VIA is very small.

The next game tested is Half Life 2! One of the best games, very optimized, it requires a good balance between CPU, RAM, GPU. All settings are Maxed out! Because Harwdare OC benchmark tool does not work with the original version of the game I tested with Fraps 3.4.7 a 60 second sequence in the beginning of the game, the park outside the train station, just as I did when I tested AN7 and KW7 Fraps outputs the results in Excel so no screenshot again!
AT7 with VIA KT333
Resolution 1024*768
Min--- Max--- Avg
28--- 47--- 38.333

Resolution 1280*1024
Min--- Max--- Avg
28--- 46--- 37.967

Resolution 1600*1200
Min--- Max--- Avg
29--- 48--- 38.3
It would seem that the CPU is strongly limiting performance!

Now it's the turn of AT7-MAX 2 with VIA KT400:
Resolution 1024*768
Min--- Max--- Avg
30--- 57--- 41.65

Resolution 1280*1024
Min--- Max--- Avg
32--- 54--- 41.25

Resolution 1600*1200
Min--- Max--- Avg
31--- 50--- 40.217

Again results are very close despite the increase in pixel density!

You put a lot of effort in your tests, nice to read with your vast enthousiasm.

And finally the system with KD7A motherboard based on VIA KT400A:
Resolution 1024*768
Min--- Max--- Avg
29--- 61--- 44.45

Resolution 1280*1024
Min--- Max--- Avg
24--- 58--- 43.717

Resolution 1600*1200
Min--- Max--- Avg
32--- 55--- 42.8

PD2JK wrote on 2023-05-28, 17:19:

You put a lot of effort in your tests, nice to read with your vast enthousiasm.

Thank you very much! I want all of these benches to mean something: people will choose a specific board with a specific chipset because they see these results and know beforehand what is best and what they should be looking for; to see the problems I encountered and the solutions I found so they do not do it all over again themselves.

Now for the Half life 2 results: there is a clear winner: KD7A/KT400A. But we can also see the big difference between AT7-MAX2 - KT400 and At7 - KT333. Please note that box MAX and MAX2 are using the same Realtek ALC 650 codec so the only difference is the chipset itself! KD7A benefits from the VT1616 codec besides the KT400A chipset to gain a huge advantage: 11.74% against AT7 at the high resolution of 1600*1200 is beyond anything in this comparative; it is the greatest difference in all tests that I have done! The VIA KT333 platform is seriously limiting performance in Half Life 2; even KT400 is left behind the KT400A by over 6% at 1600*1200!

We arrived to the final benchmark: Doom 3! This game is very demanding on the video card so I disabled AA because Geforce 7800gs is too weak to run this game with antialising; the rest of the settings are all maxed out! Test has been run 3 times and the first 2 results have been discarded!
First is the AT7 based system:
1024*768, than 1280*1024 and last 1600*1200
As the resolution increases we see no scaling whatsoever, the CPU must be limiting performance!

Second is AT7-MAX 2 based system at the same 3 resolutions:

And finally KT400A based system with the KD7A motherboard at 1024,1280,1600:

There is no mistake here! AT7/KT333 wins by a large margin and that margin increases as the resolution increases! Geforce 7800gs is limited by the CPu in all 3 systems, but KT333 system manages to feed the video card faster with data than the newer KT400 and KT400A systems! All tests repeated 3 times and the first 2 results discarded; XP reinstalled 4 times along with every game and program installed one of a time after fully formatting the HDD. The results are the same!

A bit more info about the drivers and other software used:
VIA platform driver version 4.56v
Nvidia driver version 93.71
Realtek driver version 4.06
Everest 5.50 & CPU – Z legacy for system monitoring and all info
MSI afterburner 4.30 & Fraps 3.4.7 for certain games that did not have built in test tools
Irfan view for all screenshots

Would be nice to have the results in charts when you're done with all benchmarks. It's quite difficult to follow this way and compare between platforms.

Point taken! I am now making a chart that I will put in the forum, similar to the one I made when comparing nforce2 against KT880!

Before analyzing the results some impressions about the 3 systems: all of them feel really snappy and responsive; there was not a single moment when one felt sluggish or difficult to use, the upgrade to a newer platform does not translate in an increase in usability! XP flies on all 3; benchmarks load quickly on all 3, the Raptor used in the system with the AT7 – MAX 2 board does loads Windows noticeably faster but otherwise, in general usage, does not provide with that WOW factor that I expected – this is also true in the nforce2 system.

A few pros and cons about each motherboard:
Abit AT7 – VIA KT333
Pros:
1. I do not have any other KT333 motherboard to compare with but after reading the 2002 reviews, forum posts and seeing the results I believe that this is the faster VIA KT333 motherboard worldwide. So if you want the fastest possible system for your Voodoo 5 5500 this is the board to get!
2. hardware monitoring is very good
3. rubycon capacitors: there was not a single time when the system locked up for no reason; after finding compatible RAM it was smooth sailing.
4. BIOS options that are making every other board jealous!
Cons:
1. Very picky with the RAM; Corsair works but other brands – Kingston, Kingmax, Geil, etc – are a no go!
2. it needs a power supply with a strong 5V rail because it has no auxiliary 4 pin connector; I got a Corsair RM1000i PSU that works fine but with a Corsair RM650 the system refused to even start! In the end the period correct Enermax works excellent and it is also very quiet for a 20 years old PSU.

Abit AT7 – MAX 2 with VIA KT400
Pros:
1. Auxiliary power connector present – it can work with any modern PSU; I tested with Corsair RM650 and it passed several hours of 3d mark 2001
2. hardware monitoring is excellent
3. rubycon capacitors: there was not a single time when the system froze!
4. BIOS options that are putting other motherboards to shame!
Cons:
1. It is also very picky with the RAM; Corsair works but other brands – Kingston, Kingmax, Geil, etc – are a no go!
Subjectively, the look of the board and the assembled system is, in my opinion, gorgeous: the black board together with the black Corsair ram coupled with the black cooler master case makes for one good looking system!

Are those Rubycon caps original? I used to own a MAX2 with leaking caps, they even had a 'solid' look / package.