what the hell is that card? dayum! and i thought my Videologic 928 Movie was an odd card (it does MPEG-1 decoding on ISA)

I finally purchased a PS/2 to Atari-ST mouse adaptor, leading me to dig out my Atari 520STFM! Running the sysinfo program at the moment. Yes, it's on the floor... .

I don't have any actual games for it, but I have about 20 various 'ST FORMAT' magazine demo discs with demos of various games including 'Invaders', Civilization, and Elite! Need a joystick though.

I am testing some old cards and once again I am scratching my head after using an ATi Rage. I have a rather basic low-profile card with a BIOS sticker that says R128P, a silk-screened logo that says "Powered By ATi Rage 128 ULTRA" and the system detects it as a Rage 128 Pro 4x AGP 16MB. The clock speeds according to powerstrip are 130/130. Everything I see online says that this should be a Rage 128 Ultra... but that the Ultra is always 128bit. Which seems... not true. This card has two memory chips with two additional spots left empty. It actually matches the card on this page, which says it has 64bit DDR, but that seems to be incorrect.

And yet, when I play an older 3D game on it for testing (Jedi Knight 2 Mysteries of the Sith demo, in particular) it seems ridiculously fast compared to other cards I've been testing. Other non-Nvidia\non-3dfx cards from the late 90s seem to barely be usable in this game at 640x480, and yet this card blows through it at whatever resolution I set it to. Evan at 1600x1200 it is still running better than several of the weaker cards did at 640x480. This just seems surprisingly potent for a basic card with 64bit memory.

So, just to check myself... It has two of these memory chips. I see that it is 512kbit x 32bit x 4 banks... so does that mean that it absolutely has to have a 64bit memory bus? There is no way that it's 128bit, because the two chips are only 32bit each... correct?

This is just a sanity check. Apparently once you've played on several different S3 Virge, Matrox Millennium 2, SIS 6326 and Trident 3DImage9850 cards, even a budget card from a year or two later feels like running a Geforce 2.

Living wrote on 2025-11-07, 04:11:

what the hell is that card? dayum! and i thought my Videologic 928 Movie was an odd card (it does MPEG-1 decoding on ISA) […]
Show full quote

what the hell is that card? dayum! and i thought my Videologic 928 Movie was an odd card (it does MPEG-1 decoding on ISA)

It is a professional graphics card intended for 3D and CAD work in Windows NT. Despite its looks, it is laughably slow for real time 3D graphics, though to be fair I haven't had a chance to test it myself in a proper Windows NT environment running an OpenGL game. It's powered by the much improved descendants of the GLiNT chip on the Creative Graphics Blaster VLB, which was really unique and is super collectible now but was... very slow.

Around this same time the single chip Permedia 2 would basically have all of the normal "non professional" 3D capabilities of this card, as well as proper DOS support (so it didn't need a separate S3 chip). The late 90s were interesting times for gaming and professional 3D markets, for sure. 😀

Ozzuneoj wrote on 2025-11-07, 21:18:

I am testing some old cards and once again I am scratching my head after using an ATi Rage. I have a rather basic low-profile ca […]
Show full quote

I am testing some old cards and once again I am scratching my head after using an ATi Rage. I have a rather basic low-profile card with a BIOS sticker that says R128P, a silk-screened logo that says "Powered By ATi Rage 128 ULTRA" and the system detects it as a Rage 128 Pro 4x AGP 16MB. The clock speeds according to powerstrip are 130/130. Everything I see online says that this should be a Rage 128 Ultra... but that the Ultra is always 128bit. Which seems... not true. This card has two memory chips with two additional spots left empty. It actually matches the card on this page, which says it has 64bit DDR, but that seems to be incorrect.

And yet, when I play an older 3D game on it for testing (Jedi Knight 2 Mysteries of the Sith demo, in particular) it seems ridiculously fast compared to other cards I've been testing. Other non-Nvidia\non-3dfx cards from the late 90s seem to barely be usable in this game at 640x480, and yet this card blows through it at whatever resolution I set it to. Evan at 1600x1200 it is still running better than several of the weaker cards did at 640x480. This just seems surprisingly potent for a basic card with 64bit memory.

So, just to check myself... It has two of these memory chips. I see that it is 512kbit x 32bit x 4 banks... so does that mean that it absolutely has to have a 64bit memory bus? There is no way that it's 128bit, because the two chips are only 32bit each... correct?

This is just a sanity check. Apparently once you've played on several different S3 Virge, Matrox Millennium 2, SIS 6326 and Trident 3DImage9850 cards, even a budget card from a year or two later feels like running a Geforce 2.

Living wrote on 2025-11-07, 04:11:

what the hell is that card? dayum! and i thought my Videologic 928 Movie was an odd card (it does MPEG-1 decoding on ISA) […]
Show full quote

what the hell is that card? dayum! and i thought my Videologic 928 Movie was an odd card (it does MPEG-1 decoding on ISA)

It is a professional graphics card intended for 3D and CAD work in Windows NT. Despite its looks, it is laughably slow for real time 3D graphics, though to be fair I haven't had a chance to test it myself in a proper Windows NT environment running an OpenGL game. It's powered by the much improved descendants of the GLiNT chip on the Creative Graphics Blaster VLB, which was really unique and is super collectible now but was... very slow.

Around this same time the single chip Permedia 2 would basically have all of the normal "non professional" 3D capabilities of this card, as well as proper DOS support (so it didn't need a separate S3 chip). The late 90s were interesting times for gaming and professional 3D markets, for sure. 😀

Hmm I've got a couple of those 2-memory-chip ATI Rage cards in a box - I've tried out the definitely 128-bit 4-chip-SGRAM variant and I seem to recall that one being in the performance range of a voodoo banshee?
It should definitely have 4 of those ram chips to be 128-bit and I don't think you can use bank switching to use a 64-bit memory bus as 128-bit.

Could you check what it shows in HWinfo or everest? That should give a clear indication.

I have been trying to repair the fan on my ATI Radeon 8500 All-In-Wonder DV card which came to me without a fan but with the original heatsink still fitted. Instead of trying to buy a fan to fit the heatsink I've instead got this 5 volt Sunon fan from a 1u switch which was chopped out of its 40mm fan housing and I melted some thin metal wires onto the remaining legs:

Then bent them into little fan screw mounts:

Now tied into the card's fan power which was 5 volts - it's way too loud at 5 volts and a resistor was no good to drop the speed so it's got this cool adjustable regulator for fan control 😀

Also there's something weird with this card that I'm pretty sure is not okay, it works when doing memory tests and I haven't got a windows install to use with it yet but the big resistor was getting up to 160c and smelled of burning, that seems to connect to the 12v line somehow and I think hooks into the tv tuner and disconnecting the nearby inductor unhooked the TV tuner and now it doesn't get hot. So at last I have a radeon 8500! Not the good Radeon 8500 but good enough.

That's a nice creative solution. If it works, it works.

Opened a CREATIVE CD-ROM drive and found a BIOS chip inside.
That's a first for me. Maybe it's common but a full 27C512QC-12 wasn't expected.

Not sure if this counts as modern or retro, but this video popped up on my feed. It shows the recording of the German voice overs in the upcoming Gothic remake. Really glad that they managed to get some of the original actors to reprise their roles, especially the protagonist and Xardas.

Also, I appreciate how they said that some of the German dialogue would be lost in translation, so instead of creating one unified script, they just made the English localization slightly different. Personally, I love playing the Gothic/Risen games in German for that authentic feel, and it's great to see this handled with such love and care.

"There is an English voice-over, but with Gothic it’s a bit different, since it was originally a German project, we're actually creating our own voice-over instead of doing a direct translation. So, in some cases, the characters say completely different things than in English, even though the goal in the end is the same."
- Tim Henneman, Voice Director.

It's time to see who's really worth something… 🙃

Locutus wrote on 2025-11-08, 23:02:

It's time to see who's really worth something… 🙃

The attachment IMG_2714.jpeg is no longer available

This pic + morning coffee = greatness.
I really hope at least half work. The other half must become a repair thread here 😀

Maybe not stricte retro activity but I think it counts 🙃
Started to buiId XT-CF-Lite with most vintage parts I’ve found in my shop.

Golden but modest 😆

Had a small retro chat with a technician working on fiber optic for FTTH.
He worked in a company (the ones that offered everything for businesses) and they installed all sorts of backup solutions + training.
Was nice hearing a guy still complaining on how bad companies stored important backup tapes and other media and then asking for impossible recovery of sensitive and critical data.
And they paid premium for those solutions. 🤣
Also people following instructions like crazy and never having issues.

Guess some things you can never forget.

Nexxen wrote on 2025-11-09, 09:16:

Locutus wrote on 2025-11-08, 23:02:

It's time to see who's really worth something… 🙃

This pic + morning coffee = greatness.
I really hope at least half work. The other half must become a repair thread here 😀

Good news !
Only 4 of 14 are dead 😀

Uhhgg... man... why is it always so bad? This was one of the days that I take a deep breath and decide that I'm going to test a bunch of ATi cards.

Holy cow. 😭

I've spent the last hour trying to test one stupid HD 2600 Pro AGP. Some drivers are plenty new enough but ATi had specifically removed support for most AGP cards after the X1000 series. It doesn't say that anywhere unless you dig out the right inf file and realize that it is dated several years before the driver package was released. NICE!

So, I figured, hey, why not download a slighly newer version with the AGP hotfix (hotfix is ATi-speak for "we just kinda removed support for random cards from the other drivers...") version. I downloaded Catalyst 11.2 because it still supported the 2000 series. Everything seemed okay, then during a run of 3dmark 2001SE I look over and see the system is sitting at a BSOD. NICE!

I'm not sure if the card is dead, because 90% of my ATi cards from this era have a piece of tape on them with a note saying they don't work half the time... so... it could be drivers, it could be a compatibility issue with the system (nforce 2), or it could be a genuine defect.

Sooo... why not try a different driver. I try the 9.12 driver with AGP support, I get through most of the driver setup and it tells me at the end that installation completed but that there was a problem and to look at the log file. I look at the log file and the "problem" is that it installed all the software but the driver install "failed"... and that's it. It failed. No indication of how or why it failed to put some files in the right places and tell windows to use them. At this point in the process it has already determined that a compatible card has been found, so I don't understand how it could not proceed with the driver installation.

I have used Driver Cleaner Pro to remove remnants of old drivers between attempts and somehow there are still ATi applications that can't be uninstalled sitting in Add\Remove Programs. I tried using Revo Uninstaller Pro to remove them but... what a surprise. This was just a trial version and it can't be used now. I had to install it the last time I was dealing with ATi drivers...

So, here I am, having used a chunk of my afternoon to test some cards. To show for it, I have:
4 ATi AGP cards from the 9000 series to the HD 3000 with issues that are going into the repair\diagnose bin... maybe dead, maybe drivers, maybe compatibility issues... who knows.
... and that's it.

Two days ago I tested over a dozen ISA cards from 1987 to 1992 and almost all of them worked. Over a handful of hours spread over a couple days before that I tested probably 20-25 PCI and AGP cards from the 90s and only had 3 that didn't work... and this was fully installing their drivers in Windows 98SE, running them at their max resolution and playing games on them if they offered 3D acceleration. Very few issues with that, despite dealing with several early 3D accelerators.

I will try to get a different AGP board with a Via or Intel chipset set up to test these ATi cards again before assuming they are dead, but I'm not looking forward to it.

i recently found next to my house an old pc in a pile of junk probably someone is doing some serious house cleaning, complete packed with motherboard, cards, hdd, floppy drive and CD-ROM. the board is slot 1 MSI MS 6117 ver 1.1 with PII 300mhz, 64mb of ram, pci graphics card ATI 3D Rage II + dvd with 4mb, a generic ISA sound card (ad1816) and kingston Ethernet ISA card and 230w power supply that works fine and after cleaning all the dust, all the capacitors in the power supply looks fine. what shocked me is the motherboard itself is in great shape no buldging/leaking capacitors at all, no dust build-up, aside from the cpu fan and bottom of the case. the hard drive was a western digital caviar 35100 5gb, aka "spinning rust", sadly that hard drive was pretty dead with unhealthy clicks non-stop despite my nostalgia hearing those clicks, the CD-ROM was also dead, it was picked up in the bios and opening but not reading discs, im aware very old cd-rom don't read burned cds, so i tried official disc, nothing. the motherboard itself is working fine, but i had to flash the bios due to storage limitations, the board came with vanilla bios ver 1.1 (AMI bios) flashed to ver 2.0 the last version of the AMI bios. after resetting the bios and clear cmos, it finally auto detected the hard drive but when i tried testing large hard drive like the seagate 80gb HDD it froze, later found out that this board is only limited to 32gb of storage, but i can manage. got 30gb hard drive replacement, LG 52x CD-ROM drive (all been cleaned and tested) WIN98 SE installed and planning to upgrade the ram 256mb (probably get pc-100 type to be on the safe side, after all it's a board is from 1997), get it with voodoo 3000 agp card, ESS ISA sound card, and pci usb 2.0.

HI everyone, I did some work on my old Olivetti Philos 45 (486 SL) which doesn't POST: it had leaky caps which I replaced and cleaned up the PCB which didn't have extensive damage, no solder mask nor copper seemed corroded. The Nicad "backup" 6 cell pack was dead and acid got out but luckily it's located in a seperated little area away from the MB. The large lithium button cell was also dead but didn't leak. When powered it apparently tries to start as the current spikes up but then probably gets cut as it peaks more than 3 Amps ( a bit too much) at 18Volts. Initially it would show some LCD activity other than the battery charge status (I kind of revived it by charging it seperately though the 12V pack is definitely dead) but now just that. I also had to recaps it's PSU as these were also completely rotten, but again here no pcb damage luckily.
download/file.php?mode=view&id=230692
download/file.php?mode=view&id=230691
download/file.php?mode=view&id=230690
download/file.php?mode=view&id=230689
download/file.php?mode=view&id=230688

I bought the proper smd caps even if I temporarily used regular through hole caps for testing. Unfortunately I can't seem to find a service manual anywhere for this model. It appears there is a cap shorted somewhere along the power busses. I will need to do more in depth research when I have more time. The curious part is that this model has one classic lithium cell (well a very large one as you can see, not the regular cr2032) and a small 6 cell Varta NiCd pack which is wired to the DC-DC power board (and the corrosion reached the plug along the 15cm cable!). From what I can gather and remember from when it was working this was intended to aid the swapping of the battery without kiling power to the system: once the keyboard was lifted (it's on plastic hinges) a microswitch would put the system into a sort of stand-by or low power mode also turning off the screen and you could swap out the battery just py lifting it up and putting in another one. Once the keyboard would be pressed down again the PC would resume. The hard disk could also be swapped, though not hot swapped. The floppy could be removed and the external PSU would fit in it's place with a small riser on the bottom. The trackball can popout from the right side. It also had a numerical keypad on the top right of the keyboard. All in all a neat design. As there aren't many of these left I am hoping to get mine up and running again! If anyone has any links for service manuals or schematics of course please let me know - cheers!

vintageonthemoon wrote on 2025-11-10, 21:52:

i recently found next to my house an old pc in a pile of junk probably someone is doing some serious house cleaning, complete packed with motherboard, cards, hdd, floppy drive and CD-ROM. the board is slot 1 MSI MS 6117 ver 1.1 with PII 300mhz, 64mb of ram, pci graphics card ATI 3D Rage II + dvd with 4mb, a generic ISA sound card (ad1816) and kingston Ethernet ISA card and 230w power supply that works fine and after cleaning all the dust, all the capacitors in the power supply looks fine.

What an Awesome Find! thats great!

I was lucky enough to be given a mystery box from a friend his workplace was getting rid of, nothing as good as a slot 1 Im afraid, but was pretty good early 2000's Athlon XP system, and like you couldn't believe how clean it was inside!
Video Below:

https://www.youtube.com/watch?v=FYxRqXf0VDk

Well, this is a weird one.

This PNY FX 5200 card clearly has all of the RAM locations populated, and is clearly marked as "128MB" on the sticker, yet the BIOS screen at startup says it has 64MB, and various programs in Windows agree. With 8 8Mx16bit chips it should also have a 128bit bus, but programs say that it is 64bit.

What's the deal here? I can't find any physical damage on it. It's actually in very good condition.

I've tested a lot of cards but this is an unusual issue to find, especially on a card with no visible damage. In testing the card works perfectly, other than performing like a 64bit card.

Ozzuneoj wrote on Yesterday, 05:08:

Well, this is a weird one. […]
Show full quote

Well, this is a weird one.

The attachment 20251111_000108.jpg is no longer available

The attachment 20251111_000114.jpg is no longer available

This PNY FX 5200 card clearly has all of the RAM locations populated, and is clearly marked as "128MB" on the sticker, yet the BIOS screen at startup says it has 64MB, and various programs in Windows agree. With 8 8Mx16bit chips it should also have a 128bit bus, but programs say that it is 64bit.

What's the deal here? I can't find any physical damage on it. It's actually in very good condition.

I've tested a lot of cards but this is an unusual issue to find, especially on a card with no visible damage. In testing the card works perfectly, other than performing like a 64bit card.

I noticed on some other PNY 128MB cards they utilize only four chips along the right, while the other spaces at the top are unpopulated. Maybe there was a manufacturing mistake and so only four chips are being utilized on yours?

The BFG here also looks pretty identical layout wise: https://www.ebay.com/itm/116848847477

Just my silly theory I dunno.

Ozzuneoj wrote on Yesterday, 05:08:

Well, this is a weird one. […]
Show full quote

Well, this is a weird one.

The attachment 20251111_000108.jpg is no longer available

The attachment 20251111_000114.jpg is no longer available

This PNY FX 5200 card clearly has all of the RAM locations populated, and is clearly marked as "128MB" on the sticker, yet the BIOS screen at startup says it has 64MB, and various programs in Windows agree. With 8 8Mx16bit chips it should also have a 128bit bus, but programs say that it is 64bit.

What's the deal here? I can't find any physical damage on it. It's actually in very good condition.

I've tested a lot of cards but this is an unusual issue to find, especially on a card with no visible damage. In testing the card works perfectly, other than performing like a 64bit card.

Perhaps a wrong BIOS?

sunkindly wrote on Yesterday, 05:52:

I noticed on some other PNY 128MB cards they utilize only four chips along the right, while the other spaces at the top are unpo […]
Show full quote

Ozzuneoj wrote on Yesterday, 05:08:

Well, this is a weird one. […]
Show full quote

Well, this is a weird one.

The attachment 20251111_000108.jpg is no longer available

The attachment 20251111_000114.jpg is no longer available

This PNY FX 5200 card clearly has all of the RAM locations populated, and is clearly marked as "128MB" on the sticker, yet the BIOS screen at startup says it has 64MB, and various programs in Windows agree. With 8 8Mx16bit chips it should also have a 128bit bus, but programs say that it is 64bit.

What's the deal here? I can't find any physical damage on it. It's actually in very good condition.

I've tested a lot of cards but this is an unusual issue to find, especially on a card with no visible damage. In testing the card works perfectly, other than performing like a 64bit card.

I noticed on some other PNY 128MB cards they utilize only four chips along the right, while the other spaces at the top are unpopulated. Maybe there was a manufacturing mistake and so only four chips are being utilized on yours?

The BFG here also looks pretty identical layout wise: https://www.ebay.com/itm/116848847477

Just my silly theory I dunno.

That's the most likely situation, in my opinion. If I can find another identical card that is reporting 128MB correctly I could check to see if there are any resistors in different locations. It's possible that just one or two resistors switches the card from 64bit to 128bit. I don't believe I've ever come across a 64MB FX5200 before, so it would make sense that only half the memory chips are being used on this one.

tehsiggi wrote on Yesterday, 05:56:

Perhaps a wrong BIOS?

That's possible too. I can try dumping the BIOS and checking to see if it specifically says it is 64MB or if it changes based on which memory configuration is being used (with resistor placement, as mentioned above).