It's a common problem with VIA chipset, use riva tuner to fix it.

OK... Do i have to download a specific version of Riva Tuner? Is it something obvious to find or some kind of hidden feature...?

Btw: I tried to activate 4x with a registry entry. But it had no effect.

I have a QDI Advance 10T with a VIA Apollo Pro 133T chipset. I use VIA 4in1 v4.43 on XP and AGP 4X works without issues with my ATI Radeon X1950 Pro and my old Geforce 6800GT.

One thing though, you will have to extract the 4in1 installer and install the drivers manually through the Windows device manager, as for some reason, at least on XP, the installer doesn't replace the Microsoft drivers with VIA ones. You might try to set the installer to run in Windows 2000 compatibility mode, but I don't know if this workaround works.

If you are using Windows 98, the installer should work without issues.

I really don't understand how people during the 586 and 686 generations could put up with VIA and their endless problems. Somebody seriously needs to compile a list of the bazillion different VIA chipsets, their various revisions and the problems they have, and the tools needed to "fix" them. Did PCChips ever make a board with a VIA chipset? That would have been a match made in heaven.

I guess it's mostly because everything had tons of bugs back then.
Another example: Who designes a chipset with more PCI deviders than AGP ones and then releases CPUs to be run on officially not supported settings? 😉 It worked well enough for enough people. 😁

Back in the early- to mid-2000s, I used a motherboard with a VIA KX133 chipset and GeForce GPUs. The AGP bus ran at 2X, even though both the motherboard and the video cards were 2X/4X. I decided to leave it at 2X because the performance penalty was supposedly less than 3% and I was trying to make that system stable.

The nVidia driver release notes said,

VIA KX133 and 694X Chipsets With AGP 2X On Athlon motherboards with the VIA KX133 or 694X chipset, such the ASUS K7V motherboard […]
Show full quote

VIA KX133 and 694X Chipsets With AGP 2X
On Athlon motherboards with the VIA KX133 or 694X chipset, such the ASUS K7V motherboard, NVIDIA drivers default to AGP 2X mode to work around insufficient drive strength on one of the signals.
• On Windows 9x systems, the registry key
HKEY_LOCAL_MACHINE\Software\NVIDIA Corporation\Global\System\
EnableVia4X
can be created to force NVIDIA drivers to use AGP 4X transfers.
• On Windows NT 4.0 and Windows 2000 systems, the registry key is
HKLM\System\CurrentControlSet\Services\nv4\DeviceN\ EnableVia4X
where the N in DeviceN is the system-determined number indicating the current NVIDIA device. This number is normally 0.

These registry keys should only be used if there is reason to believe that the motherboard has the appropriate drive strength.

Your motherboard has a VIA Apollo Pro 133T chipset though. I suppose nVidia could've blacklisted 4X for your motherboard's chipset too.

Yeah there was another thread a few days ago about this issue. It's NVIDIA forcing 2x in their driver to alleviate issues with the VIA AGP 4x chipsets. It's a combination of chipset issues, motherboard quality, BIOS defaults, and AGP driver problems (VIA released tons of AGP driver "updates"). I think the behavior even varies by OS. You can try forcing 4x with Rivatuner but it is quite possible you will get freezes, crashing or perhaps graphics issues.

There are other examples of this. NVidia forcing AGP 1x on AMD 750 boards. ATI forcing AGP 1x on nForce1. Etc.

Anonymous Coward wrote:

I really don't understand how people during the 586 and 686 generations could put up with VIA and their endless problems.

The best board I own is an Abit VH6T with a VIA Apollo Pro 133T and I don't have ANY problems with that thing. I tried Asus and FIC before and was never able to fix extreme performance problems on these things.

It's also very overclock-friendly and works well with my 1.4GHz Tualatin. So I wouldn't say VIA is bad at all, quite the opposite.

Anonymous Coward wrote:

I really don't understand how people during the 586 and 686 generations could put up with VIA and their endless problems. Somebody seriously needs to compile a list of the bazillion different VIA chipsets, their various revisions and the problems they have, and the tools needed to "fix" them. Did PCChips ever make a board with a VIA chipset? That would have been a match made in heaven.

It was part for the course back then. We were all just aware of it and didn't think anything of it.
Back in those days it was all about meticulously setting up your system, tweaking settings, getting everything running perfectly, and then writing down / backing up those settings.

And for stuff that needed different settings, we used config.sys and autoexec.bat menus.

Even Windows had different hardware profiles that you could use... Not something I ever did.

I'm pretty sure PC-Chips had some socket 7 boards with Via chipsets. Those things were pretty horrid back then.... But that was mostly just PC-Chips during that period.

Obviously I already have 4x transfer rate activated. Because I have made this registry entry. Riva Tuner shows it's activated. Well as described above I can't realize any difference.

One situation with AGP that I know of offhand that would show a difference between AGP 2x and 4x is ATI's Truform emulation on Radeon 9500+ cards. It is extremely heavy on AGP transfers. Though it is also extremely demanding of the CPU and you are too slow there for AGP to matter. 😀

It's also possible that Glide wrappers would be bottlenecked by AGP 2x. You really want AGP 8x there for best results. Or PCIe.

cyclone3d wrote:

It was part for the course back then.

Only for AMD CPU users. Intel chipsets were mostly solid. The flip side was if you wanted their chipset, you had to tolerate other Intel BS like the Pee4 and Rambust fiasco (and then their half assed "fix" with the SDRAM bolted onto an RDRAM chipset because intel was losing sales to AMD and needed a last minute workaround).
Then you had the leaking capacitor problem, which affected all electronics.

The best board I own is an Abit VH6T with a VIA Apollo Pro 133T and I don't have ANY problems with that thing. I tried Asus and FIC before and was never able to fix extreme performance problems on these things.

This is why some kind of list should be compiled...to figure out which VIA based boards are not crap.

I could never tolerate p4, I ran a dual PIII until core 2 came out and skipped p4 altogether.

Same, only I had a single Tualeron 1.2. Retired it in 2006. I couldn't tolerate P4, but I also couldn't tolerate AMD's crappy platform.

Anonymous Coward wrote:

Same, only I had a single Tualeron 1.2. Retired it in 2006. I couldn't tolerate P4, but I also couldn't tolerate AMD's crappy platform.

Yout loss. AthlonXP was Awesome with a capital A.

Yesterday I realized something else, that confuses me a bit. Although my Mainboard has an Apollo pro 133t chipset, Everest tells me it is a 133a. I know that the t is a modified a to realize support for tualatins. I have a 1,4ghz tualatin built in. And Everest tells me that the board supports them. So I think it is something that the device tells the program and Everest is not able to interpret it otherwise than as a 133a. Because everything runs fine I think I can ignore that?!

appiah4 wrote:

Anonymous Coward wrote:

Same, only I had a single Tualeron 1.2. Retired it in 2006. I couldn't tolerate P4, but I also couldn't tolerate AMD's crappy platform.

Yout loss. AthlonXP was Awesome with a capital A.

CPU was awesome. The majority of via chipset boards were shit though. I'm sure if you got a good board without a bugged chipset it was great, but I always got the buggy ones and got tired of constantly being screwed so played it safe and stuck with what I had because it worked. Just for the record, Tualatins were quite awesome too (though purposely held back by intel), and almost nobody had them.

Because they cost a friggin mint. 😀 I bought a Tualatin-capable PIII board and installed an old 733 (or something like that) until "the Tualatins come down in price." Which, I guess they have now. But I moved on to a Northwood P4 with DDR before the Tualatins made a compelling price/performance argument.

Never liked Athlons. The chipsets were too hot-rodded for my taste (fast, and crashed even faster), and the design of the chips was questionable. Tiny exposed die. A heatsink retention mechanism that led to "screwing" up your brand new motherboard from trying to latch the retaining clips onto the socket. Thermal margins that required a high-speed fan and really secure HS, else ... smoke. Nah.

Everything working fine... As always expected... 😎
Thx for helpful replies.