1.1a - 66fsb burnin and high/low voltage tests...

[6502kid]

I can't find Underclockers.com anywhere, so I am posting this here. :p

Burnin and Low fsb tests at core voltages from 1.80 - 1.20

Test system:

Tualatin Celeron 1100A/256/100/1.475 - SL5ZE - Philippine
Lapped CoolerMaster HS with 6w fan from a Dragon Orb3 on it.

Gigabyte 60XET-C. F10 bios. Mod with HS on southbridge, .07a fan on northbridge

Enlight 7237 case with side vents taped, and holes in front
enlarged for better airflow. 4.3w 80mm Sunon & HD cooler for intake.
1.8w 80mm and PCI slot cooler for exhaust. 2coolPC blower inside.
Enermax 350 PSU.

Gainward GF2 Ti-450 @ 280/472. WD 15g hdd, SBLive, 3comNIC

All tests ran with case covers on. Room temp 19c.
Used Prime95, UT, QIII, Burnin test and other programs for load temps.

Prior to my chipset cooling and case mods, at default speed and voltage, I
was getting 27-32c temps. At 135fsb 1.75v I was getting 29c idle, 36c load.

I decided to run a burnin test at 66fsb, 1.80 core voltage.
The XET bios shows my pci/agp at 22/44 when I changed to 66fsb, and
reports memory clock at 66mhz on post, but loads windows fine.
Sandra 2002 Pro reports 33/66 and memory running at 99mhz. (1.5mult)

Let it burn in like this for about 48hours. Lots of UT and quake.
Runs fine. No errors encountered. Just a hair slower than my 800EB.

Temps running 66fsb at various voltages:

At 1.800v - idle 25c, load 29c

1.475v - idle 23c, load 24c

1.200v - idle 22c, load 23c

Been running at 1.2v for 3 hours now without a problem.
Still score 1881 in 3dmark 2001 bench.
(Was 4200-4500 or something at 135fsb I think...)

I bet this thing would run with even less volts, but I am too
lazy to open my case and move the clear CMOS jumper, so I ain't
going to try anything lower.

Now I will see if this thing will go any higher than the 137fsb
I was getting before all this.

Here is a pic or 2 at the low fsb & voltage.

[6502kid]

CPU bench. (I bet you could run one of these at 66fsb, 1.2v
or lower, without a CPU fan. I will let you try that. :D )

[oldfart]

The XET is a tough one for underclocking/burn in. You can't use a lower PCI divider than default. For instance, my PIII-S is 133 FSB 1/4 PCI by default. I wanted to do a 66 MHz FSB underclock burn in, but cant. 66 MHz/4 = 16.5 PCI! The lowest I was able to go to and still boot is 105 FSB/26 PCI. Any lower, and I'll hang when trying to load Win2K. You can go lower since you have a 100 MHz 1/3 CPU.

If you want to really heat it up, use CPUburn (http://users.ev1.net/~redelm/). It will heat up the CPU more than Prime95 or a game such as UT, Q3 etc.

[ol' man]

Originally posted by 6502kid
I can't find Underclockers.com anywhere, so I am posting this here. :p

Burnin and Low fsb tests at core voltages from 1.80 - 1.20

Test system:

Tualatin Celeron 1100A/256/100/1.475 - SL5ZE - Philippine
Lapped CoolerMaster HS with 6w fan from a Dragon Orb3 on it.

Gigabyte 60XET-C. F10 bios. Mod with HS on southbridge, .07a fan on northbridge

Enlight 7237 case with side vents taped, and holes in front
enlarged for better airflow. 4.3w 80mm Sunon & HD cooler for intake.
1.8w 80mm and PCI slot cooler for exhaust. 2coolPC blower inside.
Enermax 350 PSU.

Gainward GF2 Ti-450 @ 280/472. WD 15g hdd, SBLive, 3comNIC

All tests ran with case covers on. Room temp 19c.
Used Prime95, UT, QIII, Burnin test and other programs for load temps.

Prior to my chipset cooling and case mods, at default speed and voltage, I
was getting 27-32c temps. At 135fsb 1.75v I was getting 29c idle, 36c load.

I decided to run a burnin test at 66fsb, 1.80 core voltage.
The XET bios shows my pci/agp at 22/44 when I changed to 66fsb, and
reports memory clock at 66mhz on post, but loads windows fine.
Sandra 2002 Pro reports 33/66 and memory running at 99mhz. (1.5mult)

Let it burn in like this for about 48hours. Lots of UT and quake.
Runs fine. No errors encountered. Just a hair slower than my 800EB.

Temps running 66fsb at various voltages:

At 1.800v - idle 25c, load 29c

1.475v - idle 23c, load 24c

1.200v - idle 22c, load 23c

Been running at 1.2v for 3 hours now without a problem.
Still score 1881 in 3dmark 2001 bench.
(Was 4200-4500 or something at 135fsb I think...)

I bet this thing would run with even less volts, but I am too
lazy to open my case and move the clear CMOS jumper, so I ain't
going to try anything lower.

Now I will see if this thing will go any higher than the 137fsb
I was getting before all this.

Here is a pic or 2 at the low fsb & voltage.

You could do it at default hurtz, I have!

You don't remember when I ran my cel-t 1.2 @ 1.2GHz 1.3v for 6 hours without a fan on the HS.