Dialing FSB back to 262 made the rig stable enough to complete 3DM 2003 - it never finished at 263. Stock GPU clocks. Could've moved the window a little.

After installing EasyTune as recommended earlier on, I also tried playing around with CineBench R10 to see what temperatures arise. Just shy of 70 degree C with the copper cored Core2 cooler.

My LGA backside bracket needs some minor adjustments, then I'll be able to put on a Zalman CNPS9000 full copper, heat pipe cooler. Maybe I'll just mount it with the backside bracket as it is now 🤣

H3nrik V! wrote on 2023-12-09, 15:20:

Dialing FSB back to 262 made the rig stable enough to complete 3DM 2003 - it never finished at 263. Stock GPU clocks.

Interesting that the highest I can get on any motherboard with my Presler is 262FSB.

My memory will go higher, and the chipset will go higher (with a Core 2 Duo or Pentium Dual-Core). The Netbursts just won’t though.

Well, apparently, setting fan speed to at least 90% in EasyTune, helped my buy enough time before crashing, at 263 FSB. Need to put on that Zalman now 🤣

[Edit] Set PCI-E speed to 103 - just for fun 😉 111 MHz PCI-E didn't boot.

Ah, mine won’t break 262 and it’s not a temperature issue 🙁

Doesn’t look like 4-way CrossfireX is enough to compensate for the overhead of Windows 7 vs XP. A touch higher than on XP, but definitely not a whole HD5970s worth!

Nice and quiet though… 7 fans at 100%

Scores updated. Ill see what i can get with my HD487 x2 crossfire setup.

Update: So apparently my second HD 4870 X2 doesnt work now.

I kindly asked the system to chill ... The Zalman CNPS9900 helped it do that.

It made it stable enough to run 3DMark 2003 at 4.3 GHz (269 FSB) and 3DMark 2001 at 4.35 GHz (272 FSB). 4.3 Actually seemed a bit faster in 2001 as well.

I also made a run in 2000 - very unimpressive 🤣 But actually even slower when not setting processor affinity to only use one core.

PCI-Express at 103 in all tests.

I did also find out why the GTX285 doesn't wanna overclock at all ... The heat sink gets crazily warm. May want to consider a dust-removal operation ...

waterbeesje wrote on 2023-12-09, 12:11:

Its regular Kingston ddr-3200 zo that could make sense. I'll see what happens when I set it to 333MHz. Also, I may swap it with some other ram I have laying around.

It does not have it's own fan, but the Zalman flower is breezing over it (CPU at 30-40'c) I guess their providers enough cooling.

All right, it's definitely the CPU.
DDR to 333MHz at 3-3-3-8 timing and putting an extra fan behind the motherboard to provide extra airflow, some bios setting set to 'standard' instead of 'turbo' and no success. 202MHz at 1,625v still makes 3D mark 03 crache to bsod.

Stock CPU speed, DDR at 400MHz, 2-2-2-5 timing, 'turbo' setting and AGP at 72MHz rock solid.

3Dmark 2001 at 21227 and 3Dmark 2003 at 12786 isn't bad at all; I'm gonna let this CPU be happy at stock speed 😀
Guess I'm gonna have to put in a different CPU to get into the list :')

waterbeesje wrote on 2023-12-10, 18:25:

All right, it's definitely the CPU. DDR to 333MHz at 3-3-3-8 timing and putting an extra fan behind the motherboard to provide […]
Show full quote

waterbeesje wrote on 2023-12-09, 12:11:

Its regular Kingston ddr-3200 zo that could make sense. I'll see what happens when I set it to 333MHz. Also, I may swap it with some other ram I have laying around.

It does not have it's own fan, but the Zalman flower is breezing over it (CPU at 30-40'c) I guess their providers enough cooling.

All right, it's definitely the CPU.
DDR to 333MHz at 3-3-3-8 timing and putting an extra fan behind the motherboard to provide extra airflow, some bios setting set to 'standard' instead of 'turbo' and no success. 202MHz at 1,625v still makes 3D mark 03 crache to bsod.

Stock CPU speed, DDR at 400MHz, 2-2-2-5 timing, 'turbo' setting and AGP at 72MHz rock solid.

3Dmark 2001 at 21227 and 3Dmark 2003 at 12786 isn't bad at all; I'm gonna let this CPU be happy at stock speed 😀
Guess I'm gonna have to put in a different CPU to get into the list :')

I assume you’ve had the motherboard higher than that with a different CPU? Otherwise it could be the board not the CPU.

If you’ve got a multimeter and know the voltage test points, it might be worth testing the board under load. Failing capacitors could easily be making the rails droop.

I received and tested the P4 670 (3.8Ghz 2MB).
Works well... but i only have two motherboards for it

Asus P5Q Pro Turbo.... Which seems to not like this CPU well. I can only boot once. Then I have to clear the bios. It works just fine with the QX6850 that i generally use with it ...

And a 4CoreDual-Sata2... which is not a top performer.

I will try my best but before that i still have a couple of things to test with my Willamette and with the Dual s604 system.

paradigital wrote on 2023-12-10, 18:33:

waterbeesje wrote on 2023-12-10, 18:25:

All right, it's definitely the CPU. DDR to 333MHz at 3-3-3-8 timing and putting an extra fan behind the motherboard to provide […]
Show full quote

waterbeesje wrote on 2023-12-09, 12:11:

Its regular Kingston ddr-3200 zo that could make sense. I'll see what happens when I set it to 333MHz. Also, I may swap it with some other ram I have laying around.

It does not have it's own fan, but the Zalman flower is breezing over it (CPU at 30-40'c) I guess their providers enough cooling.

All right, it's definitely the CPU.
DDR to 333MHz at 3-3-3-8 timing and putting an extra fan behind the motherboard to provide extra airflow, some bios setting set to 'standard' instead of 'turbo' and no success. 202MHz at 1,625v still makes 3D mark 03 crache to bsod.

Stock CPU speed, DDR at 400MHz, 2-2-2-5 timing, 'turbo' setting and AGP at 72MHz rock solid.

3Dmark 2001 at 21227 and 3Dmark 2003 at 12786 isn't bad at all; I'm gonna let this CPU be happy at stock speed 😀
Guess I'm gonna have to put in a different CPU to get into the list :')

I assume you’ve had the motherboard higher than that with a different CPU? Otherwise it could be the board not the CPU.

If you’ve got a multimeter and know the voltage test points, it might be worth testing the board under load. Failing capacitors could easily be making the rails droop.

I've tested with 2 known good power supplies (Aopen fsp and Delta) and all power rails including it's 12v rail at the P4 connector stays rock stable: varying from +5.11/5.13v and +11.93/11.96v (molex) and +12.04 / 12.08v.
The actual readings from the motherboard I haven't measured.

Now I've got another P4C800E but the non deluxe and a 3.00/800 CPU (and 2.80/800 and 2.60/800 that I'm gonna try. I also found a cooler with copper core and aluminum fins... Not as fancy af the Zalman flower, but it'll do the job 😀

Could still be tired capacitors in the board's vrm leading to fluctuations, both at the chipset as well as the cpu?

H3nrik V! wrote on 2023-12-10, 21:50:

Could still be tired capacitors in the board's vrm leading to fluctuations, both at the chipset as well as the cpu?

Exactly my thoughts. Unlikely to be a PSU directly supplied voltage, but vcore, memory voltage, chipset voltage etc.

here's a tip for everyone.
forceware drivers, control panel, change max pre-rendered frames from default (3) to 8.
helpful when working with slow cpu.

wirerogue wrote on 2023-12-11, 02:16:

here's a tip for everyone.
forceware drivers, control panel, change max pre-rendered frames from default (3) to 8.
helpful when working with slow cpu.

Unfortunately I did the above when I tried my GTX 295, still couldn’t hit your score though with identical drivers and settings. I think I also had forced the blurriest mipmaps possible using rivatuner.

Next thing I’m going to try is strapping the CPU to 1066 bus from 800. See if the FSB wall I’m hitting at 262 can be bypassed by getting the board to reset at the next strap (266).

I don't have any more to contribute with in this thread for now, so the 775 has been taken back in the box, making space for some slot-1 tinkering. But I'll probably revisit this, when I receive some of the parts I have at CPUShack's - or if I decide to borrow the GTX770 from my daily driver, just for the fun of it 🤣

Yay! Seems my P4 3.0E is willing to OC quite a bit more than my 3.4E! It even surprises it MHz-wise!

Asus P4C800E Deluxe
P4 3.0E Northwood HT @ 3.45GHz (fsb 230, 1.55V)
Nvidia 5700, 128MB, AGP 80MHz
2x 512MB Twinmos DDR400 2.5-3-3-6 @ 460 2.5-3-4-6 (auto V)
Random Seagate IDE HDD, one with rubber bumper
SATA dvd drive
Chieftech HPC-350-202 PSU
Intel CPU cooler with copper core, and a big bad fancy blower covering the whole setup

This pulls the 3Dmark 2001 scooter to 13596 and 3Dmark 2003 to 3222. Feels like the GPU is bottlenecking the system.

Stock of got 12890 / 3195... Hardly worth the OC looking at the scores but I had fun 😁

waterbeesje wrote on 2023-12-12, 13:20:

Yay! Seems my P4 3.0E is willing to OC quite a bit more than my 3.4E! It even surprises it MHz-wise! […]
Show full quote

Yay! Seems my P4 3.0E is willing to OC quite a bit more than my 3.4E! It even surprises it MHz-wise!

Asus P4C800E Deluxe
P4 3.0E Northwood HT @ 3.45GHz (fsb 230, 1.55V)
Nvidia 5700, 128MB, AGP 80MHz
2x 512MB Twinmos DDR400 2.5-3-3-6 @ 460 2.5-3-4-6 (auto V)
Random Seagate IDE HDD, one with rubber bumper
SATA dvd drive
Chieftech HPC-350-202 PSU
Intel CPU cooler with copper core, and a big bad fancy blower covering the whole setup
IMG_20231212_141533.jpg

This pulls the 3Dmark 2001 scooter to 13596 and 3Dmark 2003 to 3222. Feels like the GPU is bottlenecking the system.
IMG_20231212_132836.jpg

Stock of got 12890 / 3195... Hardly worth the OC looking at the scores but I had fun 😁

Great, so you got proof, that it was actually the CPU. Did it miss any capacitors or any thing?

2. It's always worth the overclock 🤣

H3nrik V! wrote on 2023-12-12, 13:29:

waterbeesje wrote on 2023-12-12, 13:20:

Yay! Seems my P4 3.0E is willing to OC quite a bit more than my 3.4E! It even surprises it MHz-wise! […]
Show full quote

Yay! Seems my P4 3.0E is willing to OC quite a bit more than my 3.4E! It even surprises it MHz-wise!

Asus P4C800E Deluxe
P4 3.0E Northwood HT @ 3.45GHz (fsb 230, 1.55V)
Nvidia 5700, 128MB, AGP 80MHz
2x 512MB Twinmos DDR400 2.5-3-3-6 @ 460 2.5-3-4-6 (auto V)
Random Seagate IDE HDD, one with rubber bumper
SATA dvd drive
Chieftech HPC-350-202 PSU
Intel CPU cooler with copper core, and a big bad fancy blower covering the whole setup
IMG_20231212_141533.jpg

This pulls the 3Dmark 2001 scooter to 13596 and 3Dmark 2003 to 3222. Feels like the GPU is bottlenecking the system.
IMG_20231212_132836.jpg

Stock of got 12890 / 3195... Hardly worth the OC looking at the scores but I had fun 😁

Great, so you got proof, that it was actually the CPU. Did it miss any capacitors or any thing?

2. It's always worth the overclock 🤣

Visually there's not anything wrong with the CPU, so I guess it's a dud in a certain way :')
Had some other tests with other ram and GPU but no way the 3.4 went up... And as soon as I put in this one, it got 15% more MHz almost instantly, hardly any overvolt and just a tiny bit loosened ram timing.

And yet again I find stuff that I forgot I had. This is the reason I now maintain an asset tracking system.

Asus P5N-D, BIOS modified for LGA-771, and a load of 775/771 CPUs. No 771 Netburst Xeons though 🙁

paradigital wrote on 2023-12-12, 15:42:

And yet again I find stuff that I forgot I had. This is the reason I now maintain an asset tracking system.

Asus P5N-D, BIOS modified for LGA-771, and a load of 775/771 CPUs. No 771 Netburst Xeons though 🙁

I was running my 660 on an Asus P5N-D.