Sad... those cards are so precious and rare... hopefully if the issue doesn't lie in the main chip itself it can be solved?

anunnak wrote on 2026-03-23, 23:50:

Thank you all for the assistance! :) […]
Show full quote

Thank you all for the assistance! 😀

I have come to the conclusion that the GPU has a defective 3D component and only the 2D works. I have tried the following:

1. re-write the BIOS
2. toggled off: external cache, video bios shadow/Video BIOS Cacheable
3. Tried on i430TX, i815, VIA MVP3 (same results)
4. Checked the card circuits.
5. Installed a Verite 1000E from Creative in the same systems - works perfectly.

Once i put back the Verite 1000L-P and install a driver other then the Win95 OSR the system freezes when loading the desktop. After i reset the system sometimes it even looses video signal which might mean the chip is going to a bad state. I have to shut down the PC a couple of times and wait.
3D mark 99 crashes when i hit benchmark (with the OEM win 95 driver).
So i guess when the card is trying to initialize the acceleration/features it crashes.

It is much more likely that it is a broken solder joint or bad memory than that the V1000 chip is defective.

If you don't have a microscope or a good magnifier it may be tough to diagnose this, but i would get a sewing needle and check every leg of the V1000 to see if any come free from the solder pad if you apply a tiny amount of side to side pressure (right at the tip where it connects to the pad). If you're experienced with hot air soldering then reflowing the entire chip is an option, but it's probably overkill in this case. It is most likely just one or a few solder joints that could use a bit of flux and heat to get it working again.

I have also checked the pins from the chip and memory, all ok. Unless there is a defect in one of the memory chips somehow. I will not throw it away. Will try to replace all the memory chips but the fact that after i install the new driver and the system freezes i get no more image in reset tells me that the chip might go to a bad state mode. 🙁(

You could try to apply selective hot air or freezer spray on particular chips if you find some sensitivity of one...

vGut 1.0.4 - by Chris Lundie

vGut has only been tested on Windows 95, because Windows NT
does not yet support the required RRedline drivers.

** Rendition is not responsible for
this utility and will not support it! **

The Verite General Utility (vGut) changes various registry settings
for Rendition video drivers, and reports the total & available video
memory on your board. Please refer to your driver instructions for
more details on what these settings do.

All settings should take effect immediately, although you may have
to quit and restart your game/video/etc. to see a difference.

OpenGL Settings
---------------
Some of these settings may not apply to the miniGL.

-Per-polygon mipmapping
Smooths out objects as they get more distant. This may introduce
a performance penalty. In theory, it's supposed to improve
performance, though.

-Triple buffering
May improve performance because it avoids locking into your
monitor's refresh rate. However, it uses more video memory
and can be a problem for 4 megabyte boards.

-Stereo line doubling
Improves performance when you use stereo goggles.

Bilinear Filter
---------------
This setting can affect how video clips are scaled. Normally, you
should leave it at "Auto". However, you may want to try "Always On"
if video clips look blocky when you enlarge them.

Video Memory
------------
When you click "Refresh" vGut will attempt to determine the total and
free amounts of video memory on your board. If it doesn't work, it
will report "Unknown". (This might happen on V1000 boards.)

You must exit all OpenGL applications before the memory check will
run. This is because of a nasty conflict between RRedline and OpenGL
that can make your computer lock up. vGut will attempt to detect
running OpenGL apps and to warn you about this.

The end.

Hello guys! sorry to polute the thread again but it happened once more.

I manage to buy another Rendition Verite 1000L-P, this time the Canopus Total 3D, and i get the same issue. Once it tries to initialise the desktop i get a system freeze.
TEST system:

1. GA-BX6C slot 1
2. 256 mb RAM
3. Fresh Windows 98SE
4. Build-in driver (the one that the OS installed).
5. Tried with the Canopus driver, same result.

All the behaviour as in my previous posts with the Miro Crystal VRX. What are the chances this is the defective as well? or at all? They are actually based on the same card. I am not sure that i am doing wrong. The other 2 cards i have which are based on the 1000E chip work without flaw.

If anyone has a card like this and can assist. What am i doing wrong? 🙁

anunnak wrote on 2026-07-04, 23:28:

Hello guys! sorry to polute the thread again but it happened once more. […]
Show full quote

Hello guys! sorry to polute the thread again but it happened once more.

I manage to buy another Rendition Verite 1000L-P, this time the Canopus Total 3D, and i get the same issue. Once it tries to initialise the desktop i get a system freeze.
TEST system:

1. GA-BX6C slot 1
2. 256 mb RAM
3. Fresh Windows 98SE
4. Build-in driver (the one that the OS installed).
5. Tried with the Canopus driver, same result.

All the behaviour as in my previous posts with the Miro Crystal VRX. What are the chances this is the defective as well? or at all? They are actually based on the same card. I am not sure that i am doing wrong. The other 2 cards i have which are based on the 1000E chip work without flaw.

If anyone has a card like this and can assist. What am i doing wrong? 🙁

That is very strange. Am I correct in saying you have tried: Multiple V1000L-P cards, multiple motherboards (with Pentium MMX, PIII, etc.), multiple operating systems and multiple drivers?

Are the problems always exactly the same? It is important to rule out that the two cards may be having issues in slightly different situations.

Have you tried a different monitor? You mentioned the display losing signal after crashing sometimes, so it is worth a try.

HI! Yes, tried all of that. I have 2 monitors, one Iiyama and one Fujitsu, same issue. I just can't figure out what i am doing wrong. 🙁(

EDIT: Does anyone have either of these cards and can make a tutorial for me? or at least let me know a configuration that worked for him?

Thanks 🙁

Hi all, I am an author of Rendition Verite driver for NetBSD. The driver was recently committed into the repository. Currently it supports full 2D acceleration on V2x00 cards, utilizing the original v20002d.uc. 3D is a work in progress (I already got GLQuake running on a private branch, but there's still a lot to do). If anyone has reasonably priced V1000 card for sale, let me know - V1000 is different enough from V2200 that I can't add support for it unless I have the card. I am also looking Technical Specifications of V1000, a document that explains register-level programming of the chip (I have the V2200 edition).

I've noticed that I can't reply to DMs because I am too fresh user - my location is Poland and if you're dropping me DM, leave your email too.

First release of Rendition Verite OpenGL driver:
https://c0ff33.net/cgit/verite-gl/about/

Together with recently committed NetBSD veritefb driver, this provides a complete solution to running OpenGL 1.x workloads on Rendition Verite. It is the first open source Verite accelerated driver ever (it uses the original v2000gl.uc though).

Just curious, when I run
git clone https://c0ff33.net/cgit/verite-gl
I got just empty repo with only one .git subdir - is is correct?

\verite-gl>git status
On branch master
Your branch is up to date with 'origin/master'.

Changes to be committed:
(use "git restore --staged <file>..." to unstage)
deleted: .gitignore
deleted: LICENSE
deleted: Makefile
deleted: Makefile.DJ
deleted: Makefile.mgw
deleted: Mesa.dsw
deleted: README.md
deleted: bin/installmesa
deleted: bin/mklib
...

RayeR wrote on 2026-07-20, 17:50:

Just curious, when I run git clone https://c0ff33.net/cgit/verite-gl I got just empty repo with only one .git subdir - is is cor […]
Show full quote

Just curious, when I run
git clone https://c0ff33.net/cgit/verite-gl
I got just empty repo with only one .git subdir - is is correct?

\verite-gl>git status
On branch master
Your branch is up to date with 'origin/master'.

Changes to be committed:
(use "git restore --staged <file>..." to unstage)
deleted: .gitignore
deleted: LICENSE
deleted: Makefile
deleted: Makefile.DJ
deleted: Makefile.mgw
deleted: Mesa.dsw
deleted: README.md
deleted: bin/installmesa
deleted: bin/mklib
...

That is incorrect and it seems like the contents of checkout disappeared from your local directory (that's why it suggests to use git restore).

Correct looks like :

1$ git clone https://c0ff33.net/cgit/verite-gl
2Cloning into 'verite-gl'...
3$ cd verite-gl 
4$ ls -l
5total 80
6-rw-r--r--    1 rkujawa  staff    146 Jul 21 14:27 LICENSE
7-rw-r--r--    1 rkujawa  staff  11474 Jul 21 14:27 Makefile
8-rw-r--r--    1 rkujawa  staff   2436 Jul 21 14:27 Makefile.DJ
9-rw-r--r--    1 rkujawa  staff   2482 Jul 21 14:27 Makefile.mgw
10-rw-r--r--    1 rkujawa  staff   3484 Jul 21 14:27 Mesa.dsw
11-rw-r--r--    1 rkujawa  staff   3242 Jul 21 14:27 README.md
12drwxr-xr-x    4 rkujawa  staff    128 Jul 21 14:27 bin
13drwxr-xr-x   72 rkujawa  staff   2304 Jul 21 14:27 configs
14-rw-r--r--    1 rkujawa  staff    551 Jul 21 14:27 descrip.mms
15drwxr-xr-x  103 rkujawa  staff   3296 Jul 21 14:27 docs
16drwxr-xr-x    7 rkujawa  staff    224 Jul 21 14:27 examples
17drwxr-xr-x    3 rkujawa  staff     96 Jul 21 14:27 include
18-rw-r--r--    1 rkujawa  staff    457 Jul 21 14:27 mms-config.
19drwxr-xr-x    4 rkujawa  staff    128 Jul 21 14:27 progs
20drwxr-xr-x    7 rkujawa  staff    224 Jul 21 14:27 src
21drwxr-xr-x    5 rkujawa  staff    160 Jul 21 14:27 vms

No idea how it could disappear just after checkout, maybe antivirus but it would delete binaries not text files...

RayeR wrote on 2026-07-21, 13:01:

No idea how it could disappear just after checkout, maybe antivirus but it would delete binaries not text files...

No idea either, but it seems the repository itself is just fine (ran a fresh clone to test), so I suspect it is something in your environment that is causing a problem. If you're cloning on Windows, you may want to try that on WSL or Linux container, in case it is some kind of problem with file names (there are files that come from Mesa, which end with just dot, not sure if that is legal on Windows).

OK, I tried it from a linux machine instead W11 and it cloned. But naming files with a dot at the end is generally a bad idea I thing...
I also tried to tar it and unpack on windows and all files unpacked, just those ending with dot had removed dot at end. I don't know why git client didn't even try, I watched the target dir during cloning and just .git appeared there. Maybe you could write a little note about it to web/readme.md... 😀