Why cannot overclock more?

[someonewhoknowsalittle]

Have a Core 2 Duo e4300. It's OC'ed to 3.2. The idle temp is only about 30 to 33. And the loaded temp is only about 50-52. I don't understand why I cannot raise the OC beyond 3.2 since the loaded temp is so far below the tdp and Tcase of this cpu. No matter if I use 7, 8 or 9 multiplier, and no matter what voltages I use, I cannot get a stable OC beyond 3.2. The NB is being cooled by a Mini-Kaze too. What is preventing the cpu from going faster than 3.2 GHz if it's still running well below its max temp? I thought people can OC their cpu's until their cpu's reach or almost reach the max temp they can sustain and heat they can dissipate. Is that not true?

GA-965P-DS3 Rev. 3.3
Corsair XMS2 1066 ram 4 GB
Antec New TP-750

 

[someonewhoknowsalittle]

No replies so might as well close this thread.

 

[lowjack989]

psycho answered it for you..

 

[someonewhoknowsalittle]

psycho answered it for you..

Yes, it was a 'psycho' answer. Not helpful at all. Neither was yours. If you have nothing useful or helpful to say, don't talk in a forum where people are asking for help. No one needs or appreciates a smart aleck remark.

 

[lowjack989]

Whao...now back the hell up fool...with remarks like that no one will help you....good luck

 

[lowjack989]

your MOBO has reached its potential...so there in lies your problem....

 

[Leaps-from-Shadows]

Overclocking is never guaranteed. Sometimes you get a bum CPU that won't overclock worth a darn. Other times you get an Olympic gold-medal winner. And yes, sometimes the CPU is limited by the other hardware in the system. There's no way to know unless you have other hardware handy to test it on.

You've almost doubled the speed of that CPU and you're not happy? Seriously? C'mon now...

 

[lowjack989]

You've hit the wall.......[:bilbat:3]

 

[lowjack989]

psycho can be harsh at times

 

[someonewhoknowsalittle]

psycho can be harsh at times

That's nicer than your last remark.

I was reacting to the harshness of psycho's remark calling my motherboard "crap".

As you determined, that remark was harsh and not helpful.

I'm not a fool.

I was just asking for help and I didn't get any. I didn't expect any---no one here owes me any help---but it would have been nice to get some. When I've thought of some way to help others here, I've written what I could to help them. I didn't just write that their MB or cpu or whatever is "crap". That's childish. I'm sure you would think the same if you had asked the question.

 

[someonewhoknowsalittle]

You've hit the wall.......[:bilbat:3]

With a 7x multiplier, I can get a FSB of more than 400, so I don't think the MB is limited by a FSB wall. However at 8x or 9x multiplier, the FSB has not exceeded 400, so maybe there is a wall of 400 for certain multipliers. Don't know why that is because the temp stays so much lower than TDP of the cpu.

 

[someonewhoknowsalittle]

Whao...now back the hell up fool...with remarks like that no one will help you....good luck

With all due respect, I don't think so. Someone who just tells you what you have is "crap" is not being helpful. No one would think that person is being helpful. It is a smart aleck or dumb thing to say. So why quibble about it?

I don't answer questions when I don't think I can be helpful. So why should psycho? We're all adults here, I think and hope.

The question is still why can't someone get a better OC on a core 2 duo if the cpu's temp under load conditions is much lower than the TDP of the cpu, and the FSB of the cpu is not hitting a wall, because it can be higher (in this case than 400) under lower multipliers?

I don't know why. I don't expect anyone else to know why. If they do, and want to be helpful to a fellow Tom's Hardware member, I would appreciate it. If they don't want to help, they don't have to. But don't go telling me or anyone else that their computer components are "crap" and expect to not offend them. And when people get offended, they often let the offender know that they don't appreciate it. It's only right. If you give someone, someone who never hurt you in any way. a shot, expect to get a shot back.

I've said my piece. Peace.

 

[someonewhoknowsalittle]

Overclocking is never guaranteed. Sometimes you get a bum CPU that won't overclock worth a darn. Other times you get an Olympic gold-medal winner. And yes, sometimes the CPU is limited by the other hardware in the system. There's no way to know unless you have other hardware handy to test it on.

You've almost doubled the speed of that CPU and you're not happy? Seriously? C'mon now...

I appreciate the thought and effort you put into your answer.

I am happy with the overclock. It's quite good and the cpu performs well for my uses. I just never understood why it couldn't overclock higher considering the FSB was not hitting a wall at 400 at 7x, only at 8x and 9x, and the temp wasn't anywhere near the TDP. That not knowing why bugged me.

Anyways, thanks for trying to help.

 

[lowjack989]

The RAM could be a limiting factor .....

 

[jsc]

Have a Core 2 Duo e4300. It's OC'ed to 3.2.

That is 356 MHz X 9 or about a 75% OC - way better than average.

The idle temp is only about 30 to 33. And the loaded temp is only about 50-52. I don't understand why I cannot raise the OC beyond 3.2 since the loaded temp is so far below the tdp and Tcase of this cpu. No matter if I use 7, 8 or 9 multiplier, and no matter what voltages I use, I cannot get a stable OC beyond 3.2. The NB is being cooled by a Mini-Kaze too. What is preventing the cpu from going faster than 3.2 GHz if it's still running well below its max temp? I thought people can OC their cpu's until their cpu's reach or almost reach the max temp they can sustain and heat they can dissipate. Is that not true?

Not always. I had an E5200 that I just couldn't push past 3.78 GHz at about 1.35 volts with load temps at about 55 C. I ran it up to 1.6 volts. No go.

GA-965P-DS3 Rev. 3.3
Corsair XMS2 1066 ram 4 GB
Antec New TP-750

I see two things:

1. If your memory settings are on Auto, you are overclocking the memory beyond its limits. This is a very common mistake when trying to overclock Core2 processors.

2. You may simply have reached the limits of the core. After a certain point, even if you have not reached the motherboard FSB limits or the CPU voltage or thermal limits, you just cannot push the core any farther.

Have you stress tested the system with something like Prime95? If not, your system probably is not stable once you put a load on the CPU.

 

[jsc]

Oops. Double posted.

 

[megamanx00]

Cuz motherboard no go faster

*grunt*

 

[someonewhoknowsalittle]

That is 356 MHz X 9 or about a 75% OC - way better than average.


Not always. I had an E5200 that I just couldn't push past 3.78 GHz at about 1.35 volts with load temps at about 55 C. I ran it up to 1.6 volts. No go.



I see two things:

1. If your memory settings are on Auto, you are overclocking the memory beyond its limits. This is a very common mistake when trying to overclock Core2 processors.

2. You may simply have reached the limits of the core. After a certain point, even if you have not reached the motherboard FSB limits or the CPU voltage or thermal limits, you just cannot push the core any farther.

Have you stress tested the system with something like Prime95? If not, your system probably is not stable once you put a load on the CPU.

Thanks for your answer.

1.Memory settings are on manual and memory not overclocked beyond its limits. FSB: DRAM ratio at optimal 1:1 and FSB at 400, Memory running at 800 MHz is running at its rated limit. Memory is at +0.2v. Won't work at that speed at 0.1v or 0.3v. I've run the memory over 950 mhz with no problem if I dialled back the latencies to 5-5-5-15.

2. I think you've hit on it here. The system is Prime95 stable at 3.2 GHz with 8x multiple or 9x multiple and max temp between 52-53 with small FFTs, but further voltage increases doesn't increase the cpu OC at all, only increases the cpu temp. The Intel Load test can take it up to almost 60.

I can go up to 1.425v for cpu which gives about 1.392v after vdroop. Vdroop is usually pretty high for this MB. But even if I go to 1.45v or 1.46v or so, it won't be stable over 3.2 GHz, and I don't want to burn out the cpu to push it more than that.

So, I'm going to retire this unit to simple web surfing soon for the family. It's been a great workhorse chip for 5 years or so. I'll buy probably a 2500k unless Bulldozer comes out and surprises me. I don't know whether to get an Asus or Gigabyte MB, but that will be another question after I do some research.

Thanks for your willingness to help.

 

[someonewhoknowsalittle]

Best answer selected by someonewhoknowsalittle.