Doornkaat wrote on 2022-03-12, 07:59:

It's not -15°C outside here though.😅 […]
Show full quote

bloodem wrote on 2022-03-12, 06:40:

You're a better man than I am, Doornkaat!

It's not -15°C outside here though.😅

TrashPanda wrote on 2022-03-12, 07:23:

I wish I could hand some of the hot weather over to you guys .. then you wouldn't need heating !

Might need AC tho . .wait does Europe use AC like Australia does ? (I shouldn't Assume .. it could be US . .I know they dont use AC quite like Australia does)

Here pretty much every house/building has some form of reverse cycle AC.

Things like this are often very different between european countries. In Germany the vast majority of residential buildings does not have AC while some form of heating is basically guranteed. It's usually only uncomfortably warm for a few weeks a year here while winter can get pretty cold for multiple months.

Were pretty lucky here winters are mild and rarely below 10c - 15c but summers are a totally different game ...I would gladly hand off some of our summer heat, im sure Germany during winter would appreciate mild winters.

For me . .anything below 20c is too damn cold, I work in the heat all year round due to my work so I dont work well with the cold.

Honestly I like cold weather with a bit of frost and snow in winter. It's probably just a matter of what you're accustomed to.
I take it the warm Australian air is rather dry though? I would happily trade that for our usual high humidity during our summers' dog days.
I'd love to travel Australia some day. In a couple of years maybe...
Anyway, let's stop derailing bloodem's thread with our continued weather talk. 😉

Doornkaat wrote on 2022-03-12, 07:59:

Things like this are often very different between european countries. In Germany the vast majority of residential buildings does not have AC while some form of heating is basically guranteed. It's usually only uncomfortably warm for a few weeks a year here while winter can get pretty cold for multiple months.

Same in my city (Brasov, Romania), which is near the Carpathian Mountains. People don't usually have air conditioning, because the highest temperature is ~32 degrees C, and it only lasts for one or two weeks in August. Now, I myself do have air conditioning, because I can't stand hot weather at all (I also hate cold temps, but not as much as heat). If only I could have a constant ~ 20 degrees C all year long, that would be awesome. 🤣

Anyway, getting back to the actual topic, I plan on making a video tomorrow evening (UTC), so I'm still waiting for suggestions regarding any additional game benchmarks, as mentioned in this post. So, speak now, or forever hold your peace! 😁

bloodem wrote on 2022-03-12, 16:10:

Anyway, getting back to the actual topic, I plan on making a video tomorrow evening (UTC), so I'm still waiting for suggestions regarding any additional game benchmarks, as mentioned in this post. So, speak now, or forever hold your peace! 😁

Some dos benchmarks maybe? If the goal is to have the widest performance to cover as many years as possible they could come handy.
Looking forward to see the video 😀

Nemo1985 wrote on 2022-03-12, 16:48:

Some dos benchmarks maybe? If the goal is to have the widest performance to cover as many years as possible they could come handy.
Looking forward to see the video 😀

DOS will be covered in future videos, where I'll also explore how speed-flexible each platform is (and what quirks each platform might have).
In terms of speed, all late DOS games should work just fine no matter which platform you choose.

So, for now, my focus is entirely on Windows 98 gaming performance, but no worries, DOS will not be forgotten. 😀

bloodem wrote on 2022-03-12, 16:57:

DOS will be covered in future videos, where I'll also explore how speed-flexible each platform is (and what quirks each platform […]
Show full quote

Nemo1985 wrote on 2022-03-12, 16:48:

Some dos benchmarks maybe? If the goal is to have the widest performance to cover as many years as possible they could come handy.
Looking forward to see the video 😀

DOS will be covered in future videos, where I'll also explore how speed-flexible each platform is (and what quirks each platform might have).
In terms of speed, all late DOS games should work just fine no matter which platform you choose.

So, for now, my focus is entirely on Windows 98 gaming performance, but no worries, DOS will not be forgotten. 😀

Great news, thank you!

bloodem wrote on 2022-03-12, 16:10:

Anyway, getting back to the actual topic, I plan on making a video tomorrow evening (UTC), so I'm still waiting for suggestions regarding any additional game benchmarks, as mentioned in this post. So, speak now, or forever hold your peace! 😁

Drakan is a DX6 game with a built-in benchmark. To activate it, start the game like this:

1C:\Drakan\Drakan.exe +fpstest

The result will be recorded in fps.txt which gets created in the game's directory after the benchmark completes. Before that, run its config utility to select the wanted resolution and color depth. I leave the other settings at their default values for benchmarking. Make sure the game is patched to v445 before doing the benchmarks.

Joseph_Joestar wrote on 2022-03-12, 17:43:

Drakan is a DX6 game with a built-in benchmark. To activate it, start the game like this: […]
Show full quote

Drakan is a DX6 game with a built-in benchmark. To activate it, start the game like this:

Copy code to clipboard

1C:\Drakan\Drakan.exe +fpstest

The result will be recorded in fps.txt which gets created in the game's directory after the benchmark completes. Before that, run its config utility to select the wanted resolution and color depth. I leave the other settings at their default values for benchmarking. Make sure the game is patched to v445 before doing the benchmarks.

Thank you for the suggestion. Added it to the list. 😀
Never played this game myself (I think I tried it at some point and didn't like it?), but I do remember it being popular among my friends back in the day.
Will also add Unreal Tournament, I think someone suggested it in a previous post (and UT was known at the time for being a very good CPU benchmark).

UT is indeed a good CPU benchmark, be sure to run UTBench which is more indicative of actual gameplay, instead of the flyby intro.

Garrett W wrote on 2022-03-13, 10:52:

UT is indeed a good CPU benchmark, be sure to run UTBench which is more indicative of actual gameplay, instead of the flyby intro.

Yes, of course, everything is already prepared. 😀

As promised, NEW VIDEO (Asus P5A & AMD K6-2+ 550 OC @ 633 MHz Benchmark): https://www.youtube.com/watch?v=NVUXrRnkrBE

TEST SYSTEM REMINDER:
MB: Asus P5A rev 1.04
CPU: AMD K6-2+ 550 MHz OC @ 633 MHz / FSB105 & 6 multi (2.1 volts)
RAM: 128 MB SAMSUNG SDRAM PC133 (there are actually 2 x 128 MB modules, but only half the size is detected by the motherboard; using more than 128 MB of RAM decreases performance by 5 - 10%, even when using a chip with on-die cache).
VIDEO: Asus V7700Ti GeForce 2 Ti (OC @ GeForce 2 Ultra clocks)
SOUND: Creative Sound Blaster Live 5.1 SB0220
SOUND2: ESS AudioDrive ES1688F (non-PNP / irrelevant for this test).
HDD: Seagate 40 GB IDE/PATA

Sorry, this video is quite a bit longer than the others, since I've decided to first test without sound and then rerun all game benchmarks a second time, with sound. I wanted to show what the performance impact is when using a sound card with slower platforms (even if the sound card actually has DirectSound hardware acceleration). Plus, it doesn't hurt to have the no-sound FPS stats as well, since they can be compared with other benchmarks out there (as far as I can tell, most people usually run these benchmarks without a sound card).

As mentioned previously in this post, just this one time I am using a chip that is more overclockable than usual. Also, as seen in the video, I did quite a bit of fine tuning, so this is not your typical SS7 build, it's probably in the top 1%, or at least in the top 5% (especially since, based on my experience, the Asus P5A is the fastest SS7 motherboard out there), so don't feel bad if your scores are much lower than this - it's normal!
I do plan on testing a more "average" SS7 build (in terms of speed) in the very near future, so stay tuned!
All in all, I hope that @BitWrangler doesn't accuse me of trying to make the SS7 platform look bad, because I literally did everything in my power to make it look as good as possible. 😀

I also plan on redoing the Ezra-T benchmark video, using the updated methodology and game list.

And, a few pics with the test system

Nice one, once again!

Garrett W wrote on 2022-03-14, 07:47:

Nice one, once again!

Glad you liked it 😀

bloodem wrote on 2022-02-16, 07:48:

Warlord wrote on 2022-02-16, 07:03:

I have a ver 1.7 of that motherboard. Wouldn't mind if you shared the VRM mod if it will work.

The mod is nothing spectacular, it's just a matter of replacing the existing VRM chip (HIP6004ACB), which only supports 1.8V or above, with a HIP6004BCB chip that supports voltages greater than 1.3V (they are available on eBay for $5 or less). I don't have a GA-6BXC rev 1.7 (I only have rev 2.0 and 1.9 boards), but I'm guessing the 1.7 should be pretty similar to the 1.9.

Looking for this on ebay, just trying to make sure I dont get the wrong part but would any of these work?

https://www.ebay.com/itm/361916368536?hash=it … B8AAOxyQQRR67sH

https://www.ebay.com/itm/402855282169?hash=it … 3AAAOSwNfRgnzgx

https://www.ebay.com/itm/402829193668?hash=it … JIAAOSwwlVglM6e

Yes, they seem to be OK. Personally, I've last bought 1 x 5 pcs from this seller, and already replaced 3 without any issues.

I was planning on redoing the Ezra-T video, but since I've recently modded a couple of K6-2+ CPUs and turned them into K6-3+, I figured I might as well also test the K6-3+ (on the same platform, same 633 MHz frequency, same Windows installation, same settings)...
So... NEW VIDEO (Asus P5A & AMD K6-3+ 570 OC @ 633 MHz Benchmark): https://www.youtube.com/watch?v=jrV6vPGR3Yc

TEST SYSTEM REMINDER:
MB: Asus P5A rev 1.04
CPU: AMD K6-3+ 570 MHz OC @ 633 MHz / FSB105 & 6 multi (2.1 volts)
RAM: 128 MB SAMSUNG SDRAM PC133 (there are actually 2 x 128 MB modules, but only half the size is detected by the motherboard; using more than 128 MB of RAM decreases performance by 5 - 10%, even when using a chip with on-die cache).
VIDEO: Asus V7700Ti GeForce 2 Ti (OC @ GeForce 2 Ultra clocks)
SOUND: Creative Sound Blaster Live 5.1 SB0220 (disabled)
SOUND2: ESS AudioDrive ES1688F (non-PNP / irrelevant for this test).
HDD: Seagate 40 GB IDE/PATA

This time I've decided to only benchmark the K6-3+ without sound, to keep the video "shorter", especially because, based on the data in this video and the data in the previous K6-2+ video (which does include tests with sound), you can extrapolate and make a pretty good estimate as to what performance you should expect from the K6-3+ with sound enabled.

Due to "popular demand", as a bonus video, I've now benchmarked the same K6-3+ platform using a Voodoo 3 3000 (OC @ 3500 speed / 183 MHz clocks): https://www.youtube.com/watch?v=jFFNyexouGU

Make sure to also watch the previous two videos related to SS7, where I benchmarked the K6-2+ @ 633 MHz and the K6-3+ @ 633 MHz (using a GeForce 2 Ti OC @ Ultra speeds):
https://www.youtube.com/watch?v=NVUXrRnkrBE
https://www.youtube.com/watch?v=jrV6vPGR3Yc

TEST SYSTEM:
MB: Asus P5A rev 1.04
CPU: AMD K6-2+ 570 MHz modded to AMD K6-3+ 570 MHz OC @ 633 MHz / FSB105 & 6.0 multi (2.1 Volts)
RAM: 128 MB SAMSUNG SDRAM PC133
VIDEO: 3dfx Voodoo 3 3000 AGP (OC @ V3 3500 clocks)
SOUND: Creative Sound Blaster Live 5.1 SB0220 (disabled for these tests)
HDD: Seagate 40 GB IDE/PATA

DRIVERS:
Chipset drivers: ALi Integrated Driver 2.13
3dfx Voodoo 3 drivers: 1.07.00
MiniGL driver for GLQuake / Quake 2: v1.48 (tested 1.47, 1.48 and 1.49, and 1.48 was the fastest by a very small margin)

Spoiler alert, the results might be a bit surprising: It seems that, with such a fast SS7 system, the benefits that a Voodoo 3 generally offers for slow platforms are beginning to fade, and GeForce graphics cards are starting to shine (even at low resolutions!).

A few things to keep in mind, though, when comparing the Voodoo 3 to the GeForce 2: OpenGL games like GLQuake, Quake 2, Quake 3 do take advantage of the Hardware T&L capability of the GeForce 2, more specifically of the hardware "Transform" step of the 3D pipeline (they do not take advantage of hardware accelerated lighting, because Quake games have their own lighting engine, which is based on lightmaps - essentially another set of textures overlayed on top of the base textures).
So, because of this, the Voodoo 3 has a small disadvantage here. The caveat is that, with slower CPUs, the Voodoo 3 (when using Glide, which is a low level API) is still usually able to defeat the GeForce cards (at least at lower resolutions). However, as you increase the CPU power and/or graphics resolution, GeForce cards will start to show their true potential.

As expected, the Voodoo 3 does score a big win in Unreal when using Glide (no major surprises here, since Unreal is known to have been optimized very well for this API).
This is not the case in Unreal Tournament, where the GeForce 2 and Voodoo 3 (using Direct3D or Glide, doesn't seem to matter) are within 1 FPS of each other. So, either Direct3D was very well optimized for slow systems in UT (compared to Unreal), or Glide was poorly optimized. 😁

All in all, the Voodoo 3 remains a very good (and arguably better) choice for SS7 (even if we're talking about a very fast SS7 platform like this one), especially if you want to play a lot of Glide games. It does cost a bit too much nowadays, though.

Another good choice for SS7 (when it comes to price, performance and power requirements) is a GeForce 2 MX / MX400 (the ones with the full 128 bit memory bus, avoid the 64 bit OEM crap that's out there, which were marketed as MX400 instead of MX200!).

bloodem wrote on 2022-03-27, 05:36:

All in all, the Voodoo 3 remains a very good (and arguably better) choice for SS7 (even if we're talking about a very fast SS7 platform like this one), especially if you want to play a lot of Glide games. It does cost a bit too much nowadays, though.

Another good choice for SS7 (when it comes to price, performance and power requirements) is a GeForce 2 MX / MX400 (the ones with the full 128 bit memory bus, avoid the 64 bit OEM crap that's out there, which were marketed as MX400 instead of MX200!).

Interesting results. Might also be worth trying the 3DNow! optimized version of Quake2 on that system, just for comparison purposes. Here's an Anandtech article which covered this back in the day.

I'm also curious how an MX440 would fare against an MX400 on such a system, if both cards are of the 128-bit variety of course. At a guess, the MX440 might be faster in 1024x768 and up, while performance would be fairly similar in lower resolutions. Basically, I'm wondering how much of an impact the newer Nvidia drivers that the MX440 requires would have on such a platform, compared to the improved memory architecture of that card.

Joseph_Joestar wrote on 2022-03-27, 08:24:

Interesting results. Might also be worth trying the 3DNow! optimized version of Quake2 on that system, just for comparison purposes.

You must've missed it. I did try the 3DNow! 3dfx Glide version (link to specific timestamp), just to show that it doesn't work on my Asus P5A boards.
Not sure if this Q2 3dnow driver bug affects ALi Aladdin V in general, if it has issues with a specific revision of this chipset, or if it's specific to the Asus P5A. I will test this on a Gigabyte GA-5AA rev 3.2 in the future, and see if it works (I do know that it works fine on VIA based motherboards that I've tested).
I could've tried the standard 3DNow! OpenGL driver, but it's pretty slow on Voodoo 3 cards.

Joseph_Joestar wrote on 2022-03-27, 08:24:

I'm kinda curious how an MX440 would fare against an MX400 on such a system, if both cards are of the 128-bit variety of course. At a guess, the MX440 might be faster in 1024x768 and up, while performance would be fairly similar in lower resolutions. Basically, I'm wondering how much of an impact the newer Nvidia drivers that the MX440 requires would have on such a platform, compared to the improved memory architecture of that card.

Good question, definitely worth testing! Though, based on past experience, I'm pretty sure that drivers which support the GeForce 4 MX (the first ones being something like detonator 28.xx, I believe) will be much slower on SS7 than 7.76.

bloodem wrote on 2022-03-27, 09:32:

You must've missed it. I did try the 3DNow! 3dfx Glide version (link to specific timestamp), just to show that it doesn't work on my Asus P5A boards.

That's why I'm asking actually. I don't think the 3DNow! 3DFXGL option can work on Voodoo3 and Banshee cards. For those, you need to do the following:

1. Extract the Quake 2 MiniGL file provided by 3dfx to your root Quake 2 directory (X:\Quake2\). The MiniGL is available for download at www.3dfx.com if you dont already have it.
2. Rename the MiniGL file (3dfxgl.dll) to the following: opengl32.dll. Be sure to delete any previous opengl32.dll files that were present in your root Quake 2 directory before doing so.
3. Extract the AMD Quake 2 3DNow! patch to your Quake 2 directory as documented in AMDs installation FAQ.
4. Start Quake 2. Under the video options menu, choose 3DNow! OpenGL as your rendering device, not 3Dnow! 3dfxGL

That always worked for me, meaning I got a non-trivial performance boost on my Voodoo3 when using it with my AthlonXP. Those instructions are from the Anandtech article that I linked to earlier.