can some one expalin the danger of overclocking to me

i understand it a bit but are there any other dangers then heat? i have a nforce 3 250 mobo it suposed to lock the agp and pci bus when i raisse my fsb.i overclcok my cpu by raseing fsb from 2000 to 240 leaving the mutliplier at 10 .my temps under extreme load are 47 c.i have my cpu voltage at less then the defught voltage.so i should be safe for years?i always oc then i get scared and put it back lol

If I understand you correctly you are raising your cpu clock speed from 2 GHz (200x10) to 2.4 GHz (240x10) without having to increase the default voltage settings on the cpu?

Your temperature data is incomplete so no conclusions can be drawn from the one data point alone. Ideally you would report idle temperatures at both of the cpu speeds and loaded temperatures at both of these speeds. It is also critical to know the ambient temperature in your testing location. You also didn't report what application you used to get "extreme load." I would use something like CPU Burn or Toast. Once you have reported all of the temperatures you can get a good idea of whether too much heat is being created to be reasonably safe for long-term operation of the processor.

Just because a processor looks like it is stable under "extreme load" doesn't mean that it is. I would recommend that you run Memtest and/or the latest version of 3DMark in a continuous loop for extended durations (even overnight). These programs are great at revealing weaknesses in your hardware - they will either lock up or shut down your machine if EVERYTHING isn't exactly right.

I would be very surprised if you could increase the FSB without having to significantly increase the voltage to your RAM DIMMs. If you didn't have to increase the voltage to your memory or your processor and you are using air cooling I would suspect that you may be underclocking the processor at 2 GHz?

Can you provide more detail of what processor you believe you have and what type of RAM you are using?

idle temps are around 32 c.running prime 95 it goes up to 44 c i can run super pi 1 meg in around 38 seconds.i did raise memory voltage to 2.6 vause cosair recomends that cpu is at 1.5 now.i found 3dmark test video card toi prefare prime 95 to test for stable memory and cpu.i can run prime 95 to i get bored with it i give up after 12 hours or so

There are several things that go with overclocking;

First is the heat. You are addressing that.

Second is the stability. As the clock speed goes up the rate at which the processor's transistors change state and the capacitors charge increases. This means less time for charging capacitors and less time for voltage to rise in the logic gates. If you are increasing the FSB, then all the logic and memory chips are impacted.

I find that I get rock solid performance by underclocking my AMD systems by 5%. Rock solid as in no programs ever crash. Ever. I figure that if I crash the system occasionally, the reboot time eats up the gains from overclocking. I run a linux system, by the way, and find that repetitively recompiling the kernel is a good stress test. It hammers the I/O, the memory and the processor. Someone once said if you can recompile the kernel 15 times in a row your system could be considered stable.

Hey Raggie! You are the PC guru....you testin us to see if we are doin our homework?
get ClockGen for your chipset. It will show you if your board does have the locks on or not. Just raise FSB (HTT) and watch the PCI speed. If it moves as you change HTT, then locks are off, and you risk possible HDD corruption if you push it.

raggie,
Ahhhh, the fun of overclocking...the thrill of victory and the agony of the crash. I have been overclocking for years and understand how and why processors are set as specific speeds. The nForce chipset is a good one in that it locks the PCI/AGP speeds to their stock levels. Here is the "safe" way to overclock.

The safest way to overclock your CPU is to UNDERVOLT your CPU! Let me explain. Let it run at the stock 2GHz and look at the stock voltage setting... for an Athlon it should be around 1.65V... drop the voltage to 1.60 and run benchmarks to see what happens. If they pass with no problems... keep dropping the voltage until something becomes unstable. Say it starts to flake out at 1.45V, this gives you an indicator how "good" the core is. I tried this on an Athlon XP Mobile processor (1.8GHz at 1.45V) and dropped the voltage down to 1.30 volts and it was stable. This gave me a strong indication that the processor had some speed to be tapped. If it crashes only a little below the stock voltage, you will most likely not have a lot of head room for overclocking at stock voltages. Take it as a warning.

Now to the overclocking part....

1. Don't increase the CPU voltage and keep pushing the speeds until the computer crashes. Watch your temps as it is good to use heat pipe style heat sinks and Artic Silver thermal goo. I use Thermalright XP90's on my A64 and it runs very cool and quiet with the 92mm Panaflo L series at 1000 RPM.

2. If using an AMD processor, drop the multiplier down to have the CPU run slower than stock. Raise the bus speed up to were it crashed earlier and see if it is stable. Hammer it hard with benchmarks, multiple ones with video (3D Mark) CPU burn in etc. If it is still stable 12 hours later and there are no heat problems... that lets you know what the limits of your processor are.

3. If your computer crashes at the higher bus speeds even with the CPU underclocked, the problem is your memory or motherboard northbridge crashing due to the higher bus speed. Drop your bus speed down at least 5MHz and see if you have stability. You can also boost the voltage to the memory or check to see what levels it can run.

4. Check the temps on your northbridge chip on the motherboard. Those puppies can get really HOT so you might need better cooling. If your memory is rated for very high speeds (Corsair memory rox!) then you might be having problems with the northbridge, southbridge or power FETs on your motherboard. Check those temps.

5. You can see if your motherboard is unstable at higher bus speeds due to heat by putting a big fan blowing on them and test to see if it becomes stable. I use Zalmans biggest passive heatsink on my Radeon200 chipset motherboard and added a heat sink to the southbridge. Presently, I am not overclocking... just like cool running components.

6. Now you have figured out what the max speed the processor can run at stock voltages. If you want, you can boost the voltage to the CPU a touch to see if your computer becomes stable (watch those temps!) if it does, you now really know what your max "safe" speed of your CPU is. Back it off 100 MHz or so for a buffer and proceed to the next step.

7. Heat testing. Crank the temps in your room to 90F (32C) and run a bunch of benchmarks again to see if heat is a factor... if it crashes... keep dropping your bus speed down until stable. Amazing how many people overclock in the winter and surprise! guess what happens in the summer.

8. Another way you can make a "safety buffer" is to undervolt your overclocked processor just a little bit. Does it still run fine under heat/stress testing? If so, you have a good buffer so return it to stock voltage or just leave it undervolted for cooler operation.

I overclocked my son's computer (Athlon XP Mobile 1.8GHz on a 133MHz bus) Since it was a laptop processor, I could set the bus speeds or multipliers to whatever I wish. It ran fine at stock speeds undervolted from stock 1.45 to 1.30 volts with no problems. Jumped the bus to 200 MHz with a 9 multiplier for the stock 1.8GHz but a much faster memory bus. No problems... moved the bus speed to 210MHz x 9 for 1.89MHz speed... no problems. Jumped it to 220MHz x 9 for 1.98GHz and it froze on boot up. Reset the BIOS and set it for 220x8 or underclocked 1.76GHz... it froze.

The crappy Kingmax memory could not handle the speed... boosted the voltage and it still froze. Dropped the speed to 215... no problem... 216... no problem... 217 it froze no matter the voltage. Cranked the fans on the motherboard chips and it did not matter... the memory had maxed out. Dropped it to 210MHz and ran 3 benchmarks constantly for 24 hours in a hot room... no problems. Bumped the multiplier to 10x210 for 2.1GHz for 2.1GHz and it crashed after an hour.... bumped the voltage to 1.50 volts and it ran fine for 24 hours. Figured it was a laptop processor and "stock" Athlon CPUs ran at 1.65V so I cranked it to 1.65 and bumped the multiplier up to 11X210 for 2.31GHz... it ran fine.

I went back and dropped the speeds down to 2.1GHz (210x10) and 1.50V and left it that way. Dropped the fan voltage to 5 volts (Vantec Stealth 80mm fan on a copper Thermalright heat sink) Let it cook at 75C for 24 hours (laptop processors are "allowed" to run hotter) Ran it for a week at 2.1GHz at 1.65 volts and my boys played Quake etc. Dropped the voltage down to a stock 1.45V and it ran nice and stable (want a fight? mention "burn-in" and its effect on voltages)

It has been running that way for 6 months now... 2.1GHz at stock 1.45V and has never crashed. My son's previous computer was a Intel Celeron 300A overclocked to 464MHz (max 511 MHz) and it burned in nicely at the stock 2.00 volts. Yep, it was built in 1998 and trashed in 2005 and ran constantly overclocked for almost 6 years. Yes, the heat sink was huge and I had good case airflow so I did not have any problems.

Have a new motherboard for my microATX home theater PC, it allows overclocking options and I will install it over the holidays. My Corsair CAS2 memory should be able to handle a 250MHz bus (500MHz DDR of dual channel) my plan is to change it from 1.8GHz 200x9 to 2.0GHz or 250x8. Giant heat sinks on everything to include passive Northbridge, passive Southbridge and FET heat sinks. Larged vented Artic Cooler Silencer on my video card, two Panaflo L 80mm case fans, one 60mm case fan (case fans adjustable with rheostat) hard drive Zalmann heat pipe cooler and large heat sinks on my Theater 550 TV card. Very quiet, cool and overclocked...

Now that I have caused you brain damage... undervolt your processor and see what you memory will do!

lots a good info temps never been a problem for me i was just wondering if the higher frenquecy cause problems my cpu runs cool.i dont have to go over defaught voltage to get it to 240X10 which is not a bad oc .but i think the memory is holding me back from going higher

temp 2 is my cpu temp this is underload running prime 95

Here's a video of what can happen if you overclock. Can you say toasted motherboard? http://video.google.com/videoplay?docid=5393904704265757054