athlon-power wrote:

I find it funny that the only time-accurate GPUs I have are known to be the worst at their time.

People don't know much.

Also, I want to mention that I am aware that more VRAM does NOT equal more performance, which is why I made this post in the first place. If a video card has a 100mhz processor that uses a 66mhz clock for it's 32MB of VRAM, it's going to still be slower than a video card with a 200mhz clock with a 100mhz clock speed that has 16MB vram. That was a very poor example at best, but that's the best lazily thrown together attempt I can make as an analogy for that.

Intel 740 and Savage 2000 were similar class but 2.5 years apart - there is no way that older Intel can keep up with the Savage.

Putas wrote:

Intel 740 and Savage 2000 were similar class but 2.5 years apart - there is no way that older Intel can keep up with the Savage.

I'm confused. I was of the thought that the S3 Savage was released in 1999, while the Intel i740 was released in 1998. It's very possible that the S3 Savage was released in 2000. If that's the case, that would put the S3 Savage off of the list of time-accurate GPUs I have. The reason as to why I'm being so stingy about this is because I already have a known time-accurate one installed (the i740), and the S3 Savage has had issues with running Half-Life, which is a big deal for me. I'm not sure if that's Windows 98 specific, but if it is, it still puts it off the list because I want native DOS compatibility with games such as Quake and DOOM, among other things.

You are right, 1.5 years apart. Still...
Something is up with your setup, no reason why Savage wouldn't play Half Life.

Putas wrote:

Something is up with your setup, no reason why Savage wouldn't play Half Life.

I think it's the motherboard. Just to give you a list of the problems I've had with it in the past:

PCI Slot 03 always comes up with a resource conflict, no matter the card AND with all integrated devices I can disable in the BIOS disabled, as well as only the video card installed. I've tried several different video cards, both PCI and AGP. Doesn't matter. Slot 3 always results in a resource conflict.

The nVidia GeForce 2 MX400 refuses to play anything 3D while on that motherboard. I've tried it on another system, it works great and even plays HL2 okay.

It's had weird conflicts with an external IDE controller I had in it. I installed that because the front panel connectors were proprietary to Gateway cases, and so I had to figure out the pinout. For the longest time, I could not find the HDD indicator pinout, so I installed an external IDE controller so I could see HDD activity. Later, the system refused to boot off any CD drive attached to the integrated IDE controller when the external one was installed.

It's a Gateway Tabor III, if that helps. I even updated the BIOS with a good, compatible one a while back. It had these issues prior to the BIOS update, at least, with the GPU. I bought it off of eBay, not realizing it was an OEM Gateway board. For all intents and purposes, it looked to be a generic Intel board, and now I know that assuming instead of looking into it got me here in the first place. Hindsight is 20/20, I guess.

Not so sure it's the motherboard. As far as software is concerned, that's one bog-standard i440BX board, and the BX was pretty much the gold standard when it came to trouble-free AGP. Moreover, if it was the board, I'd expect the i740 to have the same problems. I'd sooner expect software/driver issues.

Have you done a clean OS install since these problems started? If not, that would be what I'd try.

Also, you say Windows 98. There's quite a difference between FE and SE when it comes to driver models. Are you using SE or FE?

S3 savage 3d is more inline in my opinion. I740 aside if the rig is based on glide for 3d i dont see the big deal. The i740 basically has the performance of a riva 128 yet is very dos compatible. But between the 2 i would never run the i740 by itself so the savage2000 gets my vote.

dionb wrote:

Moreover, if it was the board, I'd expect the i740 to have the same problems. I'd sooner expect software/driver issues.

That's what's so odd. I've tried multiple different driver versions, at least, with the GeForce 2, and it never functions properly. I'm not sure, but I think I may have tried different versions with the Savage 2000 as well.

dionb wrote:

Have you done a clean OS install since these problems started? If not, that would be what I'd try.

I've done multiple clean installs, complete with full slow formatting the HDD. I have tried Windows 98 SE, Windows NT 4.0, Windows 2000, and Windows XP SP2. All of them exhibit the same or nearly same exact issue that Windows 98 SE does.

dionb wrote:

Also, you say Windows 98. There's quite a difference between FE and SE when it comes to driver models. Are you using SE or FE?

I've always used SE instead of FE, mainly on terms that I believe that SE was more stable. I might be wrong, but I've been fairly sure for a while that SE is the way to go.

I do recall there were some people having problems with Geforce 2 cards on the BX chipset back in the day. Possible fixes: Use 1x agp speed, turn off sideband addressing, fast writes, video bios shadow, UCWC etc.

I´m guessing the bios options of that motherboard are fairly limited though.

You can try using Powerstrip or Rivatuner to change AGP settings.

From what I've experienced with the Diamond Viper II Z200 (Savage 2K), I would only consider the card as an experiment, and only for UT99 with Metal API or Quake 3 perhaps. Those seem to be what the drivers were most coaxed to work fairly well with. And only with an Intel or possibly NVidia chipset.

GeForce 2 problems on an Intel motherboard seem likely to be indicative of some kind of motherboard AGP power delivery problem. In my experience with 2000-2003 cards, when the GF FX series added the aux power connection stability improved for some motherboards that had problems with older cards. And don't use NV drivers after 56.64 with 440BX because they freeze up for some reason. Stick with 45.23 or older.

swaaye wrote:

GeForce 2 problems on an Intel motherboard seem likely to be indicative of some kind of motherboard AGP power delivery problem. In my experience with 2000-2003 cards, when the GF FX series added the aux power connection stability improved for some motherboards that had problems with older cards.

That's the thing, though. I have an AGP FX 5200 256MB and it's perfectly stable on that thing without any aux power connector. The GeForce MX2 400 came from an old Dell, some sort of Optiplex, but I can't remember the model number off of the top of my head. It wouldn't make sense for it to not work on the PIII mobo because they're both Intel chipsets, albeit that they're probably vastly different versions. The only thing I can honestly try is maybe take a look at the older drivers, but I'm pretty sure nVidia blocked access to download those. I tried a while back and on every link but the newer ones it just told me that access was denied. Guess they don't want people using older drivers.

athlon-power wrote:

The only thing I can honestly try is maybe take a look at the older drivers, but I'm pretty sure nVidia blocked access to download those. I tried a while back and on every link but the newer ones it just told me that access was denied. Guess they don't want people using older drivers.

http://www.vogonsdrivers.com/index.php?catid= … &menustate=14,1
😀