OC'in a Celeron 2.4

ropey said:

You could buy a 2.4C Pentium4 and overclock it well on that board tho.

Click to expand...

Exactly how? That board will be doing excellently to reach the 2.4c's stock 2.4GHz. To OC it well takes a fsb in excess of 250MHz, a FSB clock speed level clearly impossible with the i845pe chipset.

larva said:

Exactly how? That board will be doing excellently to reach the 2.4c's stock 2.4GHz. To OC it well takes a fsb in excess of 250MHz, a FSB clock speed level clearly impossible with the i845pe chipset.

Click to expand...

The term "well" is certainly not only contextual but relative as well. Just as the term "a few" to one person might be "two" and to another could be "six or seven". Both views have been taken to discussion by those better adapted to discussive argument than you or I thus in my opinion your interpretation of "Well" and mine might easily be different. So I put forward that your statement regarding the term well and your definition can not be argued as a correct one and mine incorrect. Since of course this board taken to 210 - 215+ fsb is in my opinion well done for the chipset. Now on a i865 this might not be termed well but as you can see, your term "well" changes with the environment and (no pun intended) well it should. How one can take such a term and apply a 250mhz fsb to it is beyond me at any rate but that is just my view on this.

Regarding your view that this board can only make the 200MHz X 4 fsb. Well, you are incorrect in this supposition of yours as this board will allow for an increase of up to 255MHz via 1MHz increments. I have personally seen this board and chipset take further than 200MHz fsb when pci/agp locks are in place. Please regard page 43 of 95 in the manual download for these specifications should you wish. Thus this board, with the proper care, consideration, processor and other parts can make the higher bus frequencies. There is both PCI lock and AGP lock as well.

Hopefully this clarifires my statements.

Now, back to the thread originator.

Shelnutt2 said:

here's my result at the moment

Vcore idle: 1.55v
Vcore underload: 1.52v
Bios Vcore: 1.6v
Idle temp: 113-115/45-46 F/C
Load temp: 117/47 F/C

FSB:115

Overall speed: 24x115=2759MHz/2.759 GHz

Click to expand...

Have you tested for stability? FSB and Temps are fine.


R

ropey said:

You will not be able to change the Multiplier on any Intel Processor regardless of the board unless you have an ES(Engineering Sample) that comes multiplier unlocked. All other Intel processors are locked.

Click to expand...

Not to nitpick but on the 3.4 prescott (s478) you can change the multi to 17 or 14. I think this is the case as well with some of the EE procs.

Shentx said:

Not to nitpick but on the 3.4 prescott (s478) you can change the multi to 17 or 14. I think this is the case as well with some of the EE procs.

Click to expand...

One of the new features of certain ASUS motherboards is that in the bios there is a new choice specified "CPU LOCK FREE" and there is sometimes (haphazardly) a multiplier lowering from 18 down to 14 and when the specified function is disabled the CPU defaults to original 18. The same with other SpeedStep processors other than a different high and low setting. An ability to access the high and low multi but not a Multiplier Unlocked Processor.
Unfortunately these are the only settings so they are not really unlocked, rather there is two tier choice.
This choice is limited in addressing only 3.2 to 4.0 Prescotts (LGA775 & 478). Nothing for Northwoods thus certainly for this board there will be no chance to work with multipliers.

This is NOT a multiplier unlock such as AMD FX's or ES samples. This is a bug accessed from the new speedstep technology from Intel that Asus has taken advantage of by adding it to certain bios including the p4c800e-dlx. This will be corrected in future chips you can be sure although the speedstep will allow for algorithmic raising and dropping of multi's.

If they could actually unlock the multi on Intel chips why would they limit the choice to high or low and to Prescotts only? If they could actually unlock the multi then one would think they would allow more multi's and include the Northwood into the equation.

So my statement stands even though there is a CPU function that has been found and optimized for by ASUS and one supposed EE 840 shipping as an unlocked Multi. There is at this time NO method to unlock the multiplier on an Intel Processor other than Engineering Samples and that "supposed" unlocked 840 EE.

I looked in the Enhanced Speedstep specification sheet again and read that "the top frequency of the processor cannot be exceeded." Whether this means the processor monitors the actual clock setting and restricts automatic increases of the multiplier when this speed is exceeded remains to be seen. The processor may also simply assume a default FSB and increase to the default multiplier, and not exceeding the maximum default multiplier.

Also read this: "Voltage/Frequency selection will be software controlled by writing to processor MSR’s (Model Specific Registers) thus eliminating chipset dependency."

To me, this reads as the processor cannot act on it's own to change multplier and voltage settings. Software has to command it to by programming in the BIOS.

This shows that ASUS has found a way to control the low and top frequencies via speedstep and not unlocked the chips multiplier. Shut down speedstep and these functions are NOT accessible via muliplier.

Hopefully I can now get back to the task at hand which is of course, to help this person gain as maximally stable an overclock as possible and learn some of the ins and outs of overclocking rather than discussing semantics and esoteric hardware dispositions.

R

ropey said:

Have you tested for stability? FSB and Temps are fine.
R

Click to expand...

Glade temps are good....yes I've tested stabilty adn well Pi works fine but when finding prime numbers after 7 minutes of a CPU stress test error, 1 minute of memory error and 10 minutes of the first option. At first I was getting error's quicker so I then bumped up the vcore buy .025 volts. Should I bump then up agian? I haven't run into any problem yet buy running other apps...but I'm not sure about raising the vcore much higher?

What is the voltage for your dram set at? If you can keep your load temps at 50c then you can up your voltage to ~1.65. If your load temps start moving past 50c then you need (in my opinion) to get better cooling and should drop voltage until this is done.

Edit: Rather than create new posts, you can just edit previous ones.

ropey said:

What is the voltage for your dram set at?

Click to expand...

Soyo HW monitor say 2.47...but let me check the bios.


Edit:

Bios is set to Default 2.5v...has the option to go to 2.6/2.7/2.8 volts.

I would raise the dimm voltage to 2.6v and test again. If the same errors occur I would then drop the dimm voltage and raise the vcore .1v and test again all the while keeping an eye on the temps. This is the fun of overclocking so have at 'er. You can not hurt the processor with this method, the most that will happen is the system will freeze and a cmos reset will be necessary. Rather than open the box for cmos reset you can shut the system down (completely off) and back on while holding the <Insert> key down. You will get a system start and then tap the bios entry key <Del> or whichever key allows bios entry. And back at it again.

ropey said:

I would raise the dimm voltage to 2.6v and test again. If the same errors occur I would then drop the dimm voltage and raise the vcore .1v and test again all the while keeping an eye on the temps. This is the fun of overclocking so have at 'er. You can not hurt the processor with this method, the most that will happen is the system will freeze and a cmos reset will be necessary. Rather than open the box for cmos reset you can shut the system down (completely off) and back on while holding the <Insert> key down. You will get a system start and then tap the bios entry key <Del> or whichever key allows bios entry. And back at it again.

Click to expand...

Thanks....good to know I can't hurt the processor...now I'm not gonna worry about raising the volts...just watch the temps.

ok I get your point....



Anyway here's my lastest.

raised vcore to 1.625(really 1.55/1.53 uder load)
Tempature
Idle: 45 celsios
Load: 47/48 (bounces)
Memory volts are 2.6v

Shelnutt2 said:

Thanks....good to know I can't hurt the processor...now I'm not gonna worry about raising the volts...just watch the temps.

Click to expand...

I did not say that you can be indiscriminate with voltage raises as you can endanger the life length of the processor going over levels that have been ascertained by those who have gone before in this testing procedure.

The general rule of thumb is no higher than 1.7v on this processor for testing purposes and 1.65 for longer term use and this is with the above mentioned temps. I personally never even take these processors past 1.65 for testing and 1.6 for longevity. .1v raises are more than enough for testing and I would not if I were you move past 1.65v. Others may give you their advice and views on this though. This is only my experience and choice.

R

ok I get your point....



Anyway here's my lastest.

raised vcore to 1.625(really 1.55/1.53 uder load)
Tempature
Idle: 45 celsios
Load: 47/48 (bounces)
Memory volts are 2.6v

Prime95 ran for an 1.5 hours with out errors(stoped by me).
Pi 16million digits in 12 minutes.

Looking good. Congratulations.

R

ropey said:

The term "well" is certainly not only contextual but relative as well. Just as the term "a few" to one person might be "two" and to another could be "six or seven". Both views have been taken to discussion by those better adapted to discussive argument than you or I thus in my opinion your interpretation of "Well" and mine might easily be different. So I put forward that your statement regarding the term well and your definition can not be argued as a correct one and mine incorrect.

Click to expand...

Believe it or not, I know damn well what 'well' means. Your use of the term was in error, not mine.

ropey said:

Since of course this board taken to 210 - 215+ fsb is in my opinion well done for the chipset.

Click to expand...

The problem here is you weren't tallking about OC'ing the chipset 'well', however you construe the term's meaning to support your (bad) advice. Your recommendation was to OC a 2.4c well, and that is beyond the capability of any i854pe board.

ropey said:

Now on a i865 this might not be termed well but as you can see, your term "well" changes with the environment and (no pun intended) well it should. How one can take such a term and apply a 250mhz fsb to it is beyond me at any rate but that is just my view on this.

Click to expand...

You established the environment for the application of the term by (mis) selecting the 2.4c in your recommendation. In the context of a 2.4c, i845pe cannot OC well.

ropey said:

Regarding your view that this board can only make the 200MHz X 4 fsb. Well, you are incorrect in this supposition of yours as this board will allow for an increase of up to 255MHz via 1MHz increments.

Click to expand...

There is a seletion to set the multiplier, do you think it works just because it is there? Show me even a post of this, or any i845pe board running a 255fsb and then come back and talk this gibberish.

ropey said:

I have personally seen this board and chipset take further than 200MHz fsb when pci/agp locks are in place.

Click to expand...

Yeah, guess why you mentioned 210-215fsb above? Because that is very nearly the most this, or any other, i845pe board will do. That 133MHz design that is the i845pe does not have the ablility to run in excess of 220fsb at the untmost. This is a ****-poor OC for a 2.4c and to recommend someone go purchase one just for the priviledge of mis-applying it and achieving little result is just bad advice. And feel free to analyze the meaning of 'bad' all you want.

You have been hounding me for some time now and I am getting rather tired of your bullying methods.

YOU THINK YOUR OPINION IS THE ONLY ONE THAT MATTERS!!!

YOU SIR ARE AN OUT AND OUT BULLY. AND AS SUCH IMO A VERY POOR MODERATOR!!!!!!

EDIT: - I REMOVE MEMBERSHIP FROM THIS SITE -

Please continue treating members in the same way...

R

ropey said:

You have been hounding me for some time now and I am getting rather tired of your bullying methods.

YOU THINK YOUR OPINION IS THE ONLY ONE THAT MATTERS!!!

YOU SIR ARE AN OUT AND OUT BULLY. AND AS SUCH IMO A VERY POOR MODERATOR!!!!!!

EDIT: - I REMOVE MEMBERSHIP FROM THIS SITE -

Please continue treating members in the same way...

R

Click to expand...

No, my opinion is not the only one that matters, but it obvious you think yours is. But the point here is, this is not a matter of opinion. It cannot be logically argued that counseling someone to buy a 2.4c to put on their i845pe board so they can overclock it is good advice. That's just plain wrong.

You would buy a 1.8a or 2.4b for this application, I used to run a 1.8a at 3.42GHz on one. That is overclocked 'well'. 2.4c can achieve under most conditions a whopping 5% OC on i845pe, and at most 10%. Considering 2.4c always make 3GHz on a decent board (25% OC) and most made 3.4GHz (over 40% OC), this is clearly bad advice and I will say so.

And don't expect I am going to stand around and let you explain what 'well' means like I'm four years old. Like this mutates bad advice into good. You can think what you want about me or my moderation, but you can't expect I won't correct errors like this. There is plenty enough of error and misinformation on this forum, I will stop this particular bit of it, you can rest assured. If that makes me a bad guy, well, I'm the worst.

Reviving a dead thread.

Guys I need help. The only thing I've changed in this system was I added a 2.4B that I got from Elif Times, everything was working great (for about 2 weeks now) until today I went to add another CD drive to it because I had taken the burner out. I pluged it in, bios reconised it. But when I started to load windows I got the blue screen of death saying "Umountable_Boot_Drive" Anyone know what thsi is? Also It started making this clicking sound, kind of like when my old CRT would make a clicking sound to change resolutions. I do believe this clicking sound is not good. It is in interval of like 2 seconds, to me it sounds liek it is switching back and forth between somehting.

I've tried removing the cd drive still same thing happens.

I check Bios temps, everything looked good 45C processor. 25C Chipset. Help please.


Thanks,
Shelnutt2

What's making the clicking noise? Not your CD drive? It sounds like your main hard disk drive (that your OS is on) is dead, or almost dead. Does the hard drive get recognized by the BIOS?

If this is straying too far off OC topic somebody let me know.

Max0r said:

What's making the clicking noise? Not your CD drive? It sounds like your main hard disk drive (that your OS is on) is dead, or almost dead. Does the hard drive get recognized by the BIOS?

If this is straying too far off OC topic somebody let me know.

Click to expand...

I don't know what is makign the clicking noise, I've tried putting my ear up to it but I can't tell. The hard drive is reconised by the bios. I'll try a different hard drive later today.

Everything's at sock so your right on topic.

Yep hard drive was or is dieing, replaced it and it works fien now. No more clicking noises.