Classic dead hardware. Only realistic chance is chuck it in the oven and see if a reflow helps. If not, too bad.

kolderman wrote on 2021-12-25, 00:08:

Classic dead hardware. Only realistic chance is chuck it in the oven and see if a reflow helps. If not, too bad.

But why is it (for now at least) fine with Windows XP?

Was xp a different pc? Maybe power regulation or something ...

kolderman wrote on 2021-12-25, 01:06:

Was xp a different pc? Maybe power regulation or something ...

The second test was a clean install of Win98 in an i855GME (same problem) then a clean install of Windows XP in the very same motherboard - all else the same except for the OS.

Repo Man11 wrote on 2021-12-25, 01:12:

kolderman wrote on 2021-12-25, 01:06:

Was xp a different pc? Maybe power regulation or something ...

The second test was a clean install of Win98 in an i855GME (same problem) then a clean install of Windows XP in the very same motherboard - all else the same except for the OS.

it works fine at the posting screen on both systems? and under xp? have u tried stress testing it under xp with something like fur mark the runs it at 100% usage and see if it artifacts or crashes? if it works perfectly at the bios and under xp then maybe the driver for 98 is bad try a few dif versions sometimes its not always best using the last driver made.

aaron158 wrote on 2021-12-25, 22:19:

Repo Man11 wrote on 2021-12-25, 01:12:

kolderman wrote on 2021-12-25, 01:06:

Was xp a different pc? Maybe power regulation or something ...

The second test was a clean install of Win98 in an i855GME (same problem) then a clean install of Windows XP in the very same motherboard - all else the same except for the OS.

it works fine at the posting screen on both systems? and under xp? have u tried stress testing it under xp and see if it artifacts or crashes? if it works perfectly at the bios and under xp then maybe the driver for 98 is bad try a few dif versions sometimes its not always best using the last driver made.

It works great with XP - last night I set it to run the 3D Mark 2001 demo in a loop, and went for a walk. An hour later, it was still running and it still looked just as it should.

With Windows 98, I cannot try a different driver, I can't even install a driver; when it gets to the above garbled screen, you cannot do anything except hit reset, then it comes right back to that screen, even in Safe Mode.

Repo Man11 wrote on 2021-12-25, 22:24:

aaron158 wrote on 2021-12-25, 22:19:

Repo Man11 wrote on 2021-12-25, 01:12:

The second test was a clean install of Win98 in an i855GME (same problem) then a clean install of Windows XP in the very same motherboard - all else the same except for the OS.

it works fine at the posting screen on both systems? and under xp? have u tried stress testing it under xp and see if it artifacts or crashes? if it works perfectly at the bios and under xp then maybe the driver for 98 is bad try a few dif versions sometimes its not always best using the last driver made.

It works great with XP - last night I set it to run the 3D Mark 2001 demo in a loop, and went for a walk. An hour later, it was still running and it still looked just as it should.

With Windows 98, I cannot try a different driver, I can't even install a driver; when it gets to the above garbled screen, you cannot do anything except hit reset, then it comes right back to that screen, even in Safe Mode.

what if u put a dif card in the 98 machine and pre install the proper driver for it before swaping the card.

what about well doing the windows 98 install everything is ok?

are u using vga or dvi. i know in 98 with dvi i had some weridness sometimes when lunching old games but then switch to vga and its it was fine with nothing else changed.

may just be some weird incompatibility with 98. plus being a Quadro was more probably meant for use under nt or 2000.

have u checked techpowerup and see if maybe there was a newer vbios for the card maybe there was and it fixed bugs that could cause issues such at that?

another option although risky i heard some people can take quadro and ati equivalent firegl cards and then get the vbios for the consumer version of the card in your case a Ti 4400 and then flash it essentially turning it into the TI4400 witch i've seen videos on utube were people did that and it fixed issues with games people were having under quadro or firegl cards but worked fine with them flashed to the consumer version. or sometimes they do it were the get the vbios for a card that has higher clocked core or ram and then then can get an overclock out of there card.

I booted up to Parted Magic and it ran perfectly, so Linux seems to be okay as well.

Repo Man11 wrote on 2021-12-26, 15:23:

I booted up to Parted Magic and it ran perfectly, so Linux seems to be okay as well.

my only guess is maybe that default 98 vga driver it don't like it. id say try booting up 98 with another card installed pre install the proper driver then shut down and swap to the q 700

aaron158 wrote on 2021-12-26, 16:21:

Repo Man11 wrote on 2021-12-26, 15:23:

I booted up to Parted Magic and it ran perfectly, so Linux seems to be okay as well.

my only guess is maybe that default 98 vga driver it don't like it. id say try booting up 98 with another card installed pre install the proper driver then shut down and swap to the q 700

That's what I did when I first got it, but as I said in the OP, I chalked that up to a Windows 98 issue with not having removed the drivers/video card, but just pulled one and stuck the other in. So I tried it again just to be sure, but the outcome was the same; I did a fresh install of Windows 98 on an Aopen I865 P4 motherboard with an FX 5600 installed, and the outcome was exactly the same as every other time I've tried to use this card with Windows 98.

I'd be willing to try a different BIOS, but I haven't yet found one.

Repo Man11 wrote on 2021-12-27, 04:57:

aaron158 wrote on 2021-12-26, 16:21:

Repo Man11 wrote on 2021-12-26, 15:23:

I booted up to Parted Magic and it ran perfectly, so Linux seems to be okay as well.

my only guess is maybe that default 98 vga driver it don't like it. id say try booting up 98 with another card installed pre install the proper driver then shut down and swap to the q 700

That's what I did when I first got it, but as I said in the OP, I chalked that up to a Windows 98 issue with not having removed the drivers/video card, but just pulled one and stuck the other in. So I tried it again just to be sure, but the outcome was the same; I did a fresh install of Windows 98 on an Aopen I865 P4 motherboard with an FX 5600 installed, and the outcome was exactly the same as every other time I've tried to use this card with Windows 98.

I'd be willing to try a different BIOS, but I haven't yet found one.

def a head scratcher as to why it works everywhere but within windows 98. pretty much the only suggestion would be to ask around until u find someone with a ti4400 and then ask them if the can run Nvflash and back up the bios on the card.

i checked techpowerup but there bios collection is mostly stuff from 2010 and up.

if u can get ahold of a back up of a 4400ti bios and then back up your own cards bios. then do a force flash of the 4400ti bios and see how it goes. if it don't work then just make sure u have a secondary card that goes in a pci slot or a system with onboard gpu so u can switch to that so u can see what your doing to reflash the old bios back.

Grab a basic bitch PCI video card and install 98 using that, install all the chipset drivers and such get the system update etc then put the quadro in (Dont remove the PCI gpu) and install the drivers for it while using the PCI videocard as the display, once drivers are installed for the quadro .. switch the display cable to it and reboot.

its a bit hacky but it might work, sounds like 98se has some horrible basic drivers for that quadro or the quadro is trying to use part of the chipset that needs drivers but the basic 98 install doesnt have them.

I can't fault the seller on this - if they tested it with a Windows XP system, they would have had no reason to think there was any problem with the card.

Not knowing what else to do, I set this card aside. From the beginning, I'd wanted to remove the heat sink and apply new thermal paste, but it was glued in place by the antique thermal paste and wouldn't budge. Last night I put it in the freezer, and this morning I was able to pop the heat sink off the chip. I cleaned the petrified thermal paste off, and applied new. I also cleaned the card's contacts.

The most convenient system for me to test it with was my Soyo 5EMA, so I put it in and powered it up. It booted into Win98 with no problem, and I was able to install the driver and run repeated tests of 3D Mark 2000 and 2001.

Then I tried it in my M930LMR which has a universal AGP slot. It had the same issue as before! I wanted to try reducing it to AGP 2X, but there's no option in the CMOS settings to change/manually set that. I then cleaned the motherboard's AGP slot with CRC 2-26 electrical contact cleaner (it used to be called TV tuner cleaner, but analog dial TV tuners have gone the way of the Dodo Bird) and put the card back in. It then booted into Windows with no problem, and I was again able to install the driver and run repeated tests of 3D Mark and Aquamark with no problems.

Lesson learned; if you see something like the second image in the first post, try cleaning both the card and the slot's contacts. I had little hope of getting this card to work properly, and I meant to use it in a Windows 98 system, so only working with WinXP meant that it was useless to me. So this outcome made my day.

the thermal crap companies used back then was freaking awful had a 9700pro and getting that stuff off there was near impossible to clean it was like cement. one thing that seems to help with super old crusty pasteis spray a tiny bit of wd40 on a qtip it seems to just eat right though that crap then clean it really good with alcohol to get rid of any left over wd40

My victory was short lived. After trying it in an AGP 8x motherboard, the problem came back. When I again put it back in the universal AGP board, it again had the issue there. But it will consistently work fine in my Soyo MVP3 AGP 2x board that has Windows 98. So I guess it is some sort of power issue?

Again, with the identical hardware, everything is fine with Windows XP! Heavy sigh.

That's very odd. I've not experienced anything like this, despite running NVIDIA cards in Windows 98 from the Riva 128 through the GeForce 6 series. Although I did experience varying degrees of corruption with certain chipset and CPU combinations, I haven't seen this exact one that you're describing.

Problems I have seen include trying to pair a card with pixel shaders with pre-coppermine, certain cards with pre-440BX chipsets, drivers after the 56.64 are always unstable, anything after the GeForce FX series is often incompatible with Windows 98, and general instability on VIA chipsets until the appropriate 4-in-1 driver is installed. Some newer chipsets will also operate the card in PCI mode, such as Intel 845, until the chipset drivers are installed to enable AGP support.

To see if the card is operating in PCI or AGP mode, check the NVIDIA display driver summary. You can also set the AGP mode in the display driver on later driver versions with the coolbits registry key applied.

Can you please provide exactly what motherboard, CPU, and NVIDIA driver version you're using?