Sign in with
Sign up | Sign in
Your question

Overclocking E7200 "Natural Overclock?"

Last response: in Overclocking
Share
June 21, 2008 12:46:52 PM

I just build my own PC with the help of people in the new build forum and was wondering about overclocking the PCU slightly.

This is my current setup:
Memory - http://www.ebuyer.com/product/139662
Processor - http://www.ebuyer.com/product/145258
Cooling - http://www.ebuyer.com/product/105994
Motherboard - http://www.ebuyer.com/product/143176
500W PSU

I was told that the E7200 can reach 3.16GHz as a "natural overclock".
Does "natural overclock" mean i can just go into the bios and shift the FSB to 333 and everything will be fine?

Im aware the chip can go up to 4GHz+ but i dont really want any of the messing around that overclocking comes with, id just like a safe option of getting the most from the chip.

Any advice would be priceless! thanks.

June 21, 2008 1:22:39 PM

All depends on what you mean by natural overclock? Overclocking isnt promised on any CPU and you could have an E7200 that could just make the cut. If you mean natural in no voltage change you at best would have a 20% chance. If you mean natural in not breaking the 10% voltage increase rule you would have a 99% chance.

I suggest you set your FSB to 333 at stock voltage and see if it boots. If it doesn't reset and increment the voltage by 1%. IE if stock is 1.2v then 1.21. If it does boot but isn't stable increment by smaller numbers. You want the least amount of voltage as possible while keep the system stable.
June 21, 2008 6:40:51 PM

Well i shifted the frequency to 333 and it automatically changed my RAM speed to 1000MHz so i tried to boot and it failed.

I then reduced the RAM back to 800MHz and its fine now.

I set the voltage control to automatic and it says that its now 1.216V when it was only 1.5~ before.

Also CPU-Z tells me that the multiplier of the CPU is changing between 6 and 9.5 fairly often, is that right?
Related resources
June 21, 2008 6:49:45 PM

Sorry,

I set the voltage control to automatic and it says that its now 1.216V when it was only 1.12~ before.
June 21, 2008 7:02:24 PM

The change from 6-9.5 is normal and can be disabled in Bios. Its Intels Speedstep technology to save power while not under load.
June 21, 2008 7:31:27 PM

Post your idle temps. Just load Real Temp and watch the temperature settle for 2-4 minutes, it should be <40. Better yet, <30.
June 21, 2008 7:33:56 PM

you might be able to do 4Ghz on it. but i think lapping both will be better.
June 22, 2008 4:14:49 PM

Its at 3.16Ghz now and running fine, i was playing around last night and it peaked at 49*C while playing old games, havent tried any stress tests yet tho.

Anyone got any good suggestions on stress test software?
June 22, 2008 4:17:11 PM

Prime95.

Run the blend test first for about 12hours. Then move on to the small FFTs for about 8hours. These are the bare minimum for everyday use, but I found out running the blend test for 24hours gave me absolute stability.
June 22, 2008 4:35:40 PM

run small ffts for 10 hours at least and dont forget to tick error checking.
June 22, 2008 8:21:56 PM

Well i just set my case fan to Medium speed and it idles at 37*C which i think is pretty good.

Il give those stress tests a go but im not sure about 24 hours :o 
Is that really needed? i was just thinking about using something that would max out the CPU load for an hour or so just to check it wont overheat.
June 22, 2008 8:23:55 PM

I meant to say 27*C idle, but i cant edit it.
June 22, 2008 9:08:36 PM

do a screenie with CPU-Z,coretemp,realtemp,hwmonitor and of course prime on small ffts:) 
June 23, 2008 12:04:51 AM

There was a constant 10*C difference in temp between realtemp and coretemp, somethings not ok there but i got no errors during 15min of the test and the temp seemed to level off.

Il run a longer test tomorrow.

June 23, 2008 12:12:28 AM

sorry im not into 45nm. i wish the temp is Realtemp but i somehow prefer the "safer" reading of the coretemp. i think the temp different you better query a person called Lupiron.PM him along with screenshot.:) 
June 23, 2008 12:42:03 AM

I would run the longer test and see how it goes.
June 24, 2008 8:30:13 PM

rich rich said:
There was a constant 10*C difference in temp between realtemp and coretemp, somethings not ok there but i got no errors during 15min of the test and the temp seemed to level off.

Il run a longer test tomorrow.

]http://img113.imageshack.us/img113/2607/screen22yt7.th.jpg


The difference is because of the value for the processor's TjMax configured in Real Temp and Core Temp. It has not been formally documented (externally) by Intel but many believe it to be 95C. Some of the older versions of programs have it as high as 105C, which many believe is too high to be correct.

Edit: In other words, RealTemp is currently considered more accurate at defaults. You can configure the TjMax

Core 2 Quad and Duo Temperature Guide
http://www.tomshardware.com/forum/221745-29-core-quad-temperature-guide
June 24, 2008 11:44:03 PM

So the cooler readings were the correct ones, should be able to chill beers in this thing if i turn the fans on max :p 
June 25, 2008 1:00:26 AM

Or you can fry eggs if you remove the fans.
June 26, 2008 10:02:15 AM

*burp*

Now where's my cheeseburger?
June 26, 2008 11:18:44 AM

how many "mugz" of coke you want?
July 22, 2008 10:45:57 PM

Hey I have an Asus P5N-D 750i with a E7200 2gb of Patriot DDR2DDR2 SDRAM 800mhz RAM. Asus has a program with the mobo called AI gear 3 or something and you can overclock from there and they have presets that allow inciments of 5, 10, 15, and 20% I went with 10 then 20 and my CPU is running at 2.78ghz at about 42 degrees C. All I have changed is applying the 20% setting is it safe to do this without changing anything else???
!