GeorgeMan wrote on 2024-11-29, 07:56:

Unfortunately, this motherboard is unstable at 83MHz FSB with any RAM and any CPU

Have a look here https://theretroweb.com/chip/documentation/vt … ee821910227.pdf.

The northbridge is able to do 66/75 MHz FSB only (unless you maybe have a newer/better revision of it).

To be sure, you can disable the onboard-cache (L2-cache) in BIOS and test 83 MHZ FSB again - sometimes the cache, or the cache-timings are problematical.

In my eyes, this board is not good for any overclocking attempts, and this VIA-chipset VIA580VPX is not the best choice.

Thank you both for the suggestions, I'm happy with the 300MHz since this board is really not meant to work with K6-2 and Tillamook and there's no need to push it harder IMO.
After all, if I want more performance, there are dirt cheap Pentium IIIs and Celerons out there. 😀
https://www.anandtech.com/show/72/17

Sphere478 wrote on 2024-12-03, 04:56:

GeorgeMan wrote on 2024-11-29, 07:56:

Hi, tried my Tillamook on a Shuttle HOT-579 which has jumper settings to set 2,1v lowest voltage. So I set it to 2.1v, measured […]
Show full quote

Hi, tried my Tillamook on a Shuttle HOT-579 which has jumper settings to set 2,1v lowest voltage. So I set it to 2.1v, measured the output of the power transistor just to make sure I won't fry it, installed and voila.
Bios said Tillamook-MMX and then froze.

So I did the yellow and blue mods (not the green for the forced voltage detection, not the red for the L2 and not the orange), installed it back and it just worked at 4x66=266 😁
Unfortunately, this motherboard is unstable at 83MHz FSB with any RAM and any CPU (tried PC133 SDRam at the slowest settings and a K6-2 500 which works on another board at 83x6), so I ended up at 75x4=300MHz. That way it completed all of the benchmarks just fine.

It a nice little boost, considering that I didn't expect miracles, just an uplifted P-MMX. It of course loses to K6-2 due to raw MHz power of the latter.

Nice!

Maybe disabling onboard hardware like IDE and using external card would help?

gonzo wrote on 2024-12-06, 08:28:

Have a look here https://theretroweb.com/chip/documentation/vt … ee821910227.pdf. […]
Show full quote

GeorgeMan wrote on 2024-11-29, 07:56:

Unfortunately, this motherboard is unstable at 83MHz FSB with any RAM and any CPU

Have a look here https://theretroweb.com/chip/documentation/vt … ee821910227.pdf.

The northbridge is able to do 66/75 MHz FSB only (unless you maybe have a newer/better revision of it).

To be sure, you can disable the onboard-cache (L2-cache) in BIOS and test 83 MHZ FSB again - sometimes the cache, or the cache-timings are problematical.

In my eyes, this board is not good for any overclocking attempts, and this VIA-chipset VIA580VPX is not the best choice.

LeFlash wrote on 2018-08-16, 15:08:

Ok guys, another one. This time: 112MHz x 3.5 => 392MHz […]
Show full quote

Ok guys, another one.
This time: 112MHz x 3.5 => 392MHz

Quake Timedemo over 80fps, even with 640x480 over 20fps!

Board is a FIC VA503+ this time - only board i had on hand with >100MHz

The attachment 20180816_170037_resized.jpg is no longer available

The attachment 20180816_170105_resized.jpg is no longer available

The attachment 20180816_170206_resized.jpg is no longer available

I have this exact motherboard with a modded Tillamook 166mhz CPU overclocked to 450mhz stable. For the life of me I cannot get anywhere close to your Phil's Computer Lab Quake dos bench mark test results. Can you please give me a few pointers here on how you accomplished this? Thank you.

Interesting such extreme O'C result. I would focus on 300MHz Tilly because of 4x multi but as I can see even this low clocked 166MHz can perform much better than 233MHz I had. So I would change my focus and try to look also for low clocked Tilly if I'll find it somewhere... Max. multi of 166MHz is 3,5?

Ad poor Quake perf.
Just try all available FSB steps and note the results. Higher FSB doesn't always give better results, depends on chipset, dividers, other magic, etc...

I've also got a 166mhz chip that I can run at 400mhz (100mhz x 4.0 @ 2.3V) stable according to 3 hours of Prime95 on an Asus P5A-B. So at least on my board the 4.0x multiplier works fine on a 166mhz chip.

mumblyomod wrote on 2025-02-19, 21:24:

I've also got a 166mhz chip that I can run at 400mhz (100mhz x 4.0 @ 2.3V) stable according to 3 hours of Prime95 on an Asus P5A-B. So at least on my board the 4.0x multiplier works fine on a 166mhz chip.

Have you tried bumping up the front side buss up to 112mhz. That is where mine is at and it gives me the 448mhz on the CPU.

I've tried bumping it up to 105mhz @ 2.3V but it wasn't stable. I'm not convinced it's the CPU that can't take more speed.
I suspect the issue is with my motherboard because even a with K6-2+ it has occasional crashes with a 112mhz fsb, and with the Tillamook the bios detects a vcore error even though the voltage it detects looks correct. I'm planning on testing the caps sometime soon but I haven't gotten around to desoldering them yet.

mumblyomod wrote on 2025-02-20, 03:34:

I've tried bumping it up to 105mhz @ 2.3V but it wasn't stable. I'm not convinced it's the CPU that can't take more speed.
I suspect the issue is with my motherboard because even a with K6-2+ it has occasional crashes with a 112mhz fsb, and with the Tillamook the bios detects a vcore error even though the voltage it detects looks correct. I'm planning on testing the caps sometime soon but I haven't gotten around to desoldering them yet.

I think I have finally figured it out. I removed the 128 meg stick of PC100 ram and switched to a 128meg stick of Infineon PC133 CAS2.0 ram then went into my FIC 503+ (v1.2 motherboard) BIOS and set anything to do with performance to the TURBO setting. Next I booted into DOS 6.22 (instead of Windows 98SE) and ran the original DOS version of the Quake Demo. (See pictures below of the results) and then finally the really huge difference made in boosting the results came when I

Nice, so it seems cache doesn't help much but RAM speed..

RayeR wrote on 2025-02-20, 19:29:

Nice, so it seems cache doesn't help much but RAM speed..

Sorry, I did not realize that I had not finished my last statement in the previous post until it was to late and I could not find a way to edit the post once I submitted it. What I was going to say was that the huge difference in the frame rate performance from the Quake Demo came when I reduced the screen size to the smallest window before running the Demo. Essentially, I cheated the Demo results which makes the results of 106.8 FPS null and void. 59 - 60 FPS is the maximum I can get with my current setup.

My results from 3 years ago:

Ignore the reported 100 MHz FSB, the system was running at 115 MHz, which was a max tested stable frequency with L2 active on my GA-5AA. CPU voltage: 2.4V

Results for Quake:

1Intel Pentium Embedded MMX 266 MHz SL2Z4 @ 460 MHz, 2.4V
2
3Quake v1.08 (DOS)
4timedemo demo3 (-nosound -nocdaudio)
5
6320x200 - 91.4 fps
7640x480 - 28.2 fps

For a comparison, my other system (GA-5AX):

1K6-III+ @ 550 MHz, 1.9V
2Quake v1.08 (DOS)
3timedemo demo3 (-nosound -nocdaudio)
4
5320x200 - 99.9 fps
6640x480 - 38.7 fps

And yet another (VIA VPX motherboard):

1Pentium MMX @ 225 MHz
2Quake v1.06 (DOS)
3
4timedemo demo3
5320x200 - 50.5 fps
6640x480 - 17.7 fps

Especially for 640x480 you can improve your score with bios tweaking. Settings like Primary Frame Buffer, Graphics Aperture Size can move score a lot.

Any new thoughts on how to possibly activate 4.5 and 5x on ppga or film chip variants? Maybe through software?

Socket 7 Tillamook 300 (4.5x)

this thread is so long that i can't read it all through, so i'll be short:
has everyone come to conclusion on getting tillamook running with l2 cache yet? is it dependent on chipset, bios, or board circuit design?
also, to keep it safe from risk of burning up, the board has better have adjustable Vio, rather than the default 3.3/3.5 volt.

noshutdown wrote on 2025-07-30, 06:21:

this thread is so long that i can't read it all through, so i'll be short:
has everyone come to conclusion on getting tillamook running with l2 cache yet? is it dependent on chipset, bios, or board circuit design?
also, to keep it safe from risk of burning up, the board has better have adjustable Vio, rather than the default 3.3/3.5 volt.

The required and optional hardware mods are this image posted by Sphere478, somewhere in this topic, but also here:
Tillamook- Aiming For The Stars!

Vio seems not be problem, just take care about vcore, ideally measure voltage with a DMM before you plug CPU in the socket.

RayeR wrote on 2025-07-30, 13:23:

Vio seems not be problem, just take care about vcore, ideally measure voltage with a DMM before you plug CPU in the socket.

Vio may be less relevant than Vcc, but still listed in "absolute voltage limit that will cause your processor to blow up" section of the datasheet.

The reality is that most users runs VIO at 3,3-3,5V without problems. Only very few MBs can set VIO lower. I was afraid a bit about it too but I fried my Tilly by excessive Vcore (bad manipulation with VRM feedback during having CPU in the socket, bad idea, I know...)

noshutdown wrote on 2025-07-30, 06:21:

this thread is so long that i can't read it all through, so i'll be short:
has everyone come to conclusion on getting tillamook running with l2 cache yet? is it dependent on chipset, bios, or board circuit design?
also, to keep it safe from risk of burning up, the board has better have adjustable Vio, rather than the default 3.3/3.5 volt.

Most people have been successful with it

Scroll back a few pages and until you see the most up-to-date image it’s like a JPEG that we uploaded it shows where to put the resistors

Alternatively, if you look in my signature, there are some PCB projects that provide the mods

Yeah, as far as anyone can seem to tell VIO seems to be no problem. Just run it at the 3.3 V but the core should be 1.9 to 2.1 V.

Edit
Re: Tillamook 266MHz and working L2 cache?