Overclock.net › Forums › Intel › Intel CPUs › i7 2600(Non-K) overclock, possible HW Info 64 bug?
New Posts  All Forums:Forum Nav:

i7 2600(Non-K) overclock, possible HW Info 64 bug?

post #1 of 7
Thread Starter 
I have recently had a rather successful overclock on my old i7 2600 simply by enabling and disabling a couple things and increasing the base clock to 106, which has proven pretty stable(after I turned the EPU power save option off) and increased my clock speed from 3.4 Ghz to 3.57 Ghz.

This resulted in an improvement in Cinebench from 563 to 652 and Passmark CPU score from 8.5k to 9.2k, which rather impressed me since the processor is a locked version and everywhere seems to say it cant be overclocked, not to mention I'm on the stock cooler and haven't even put new paste on and the cpu still runs cool(no higher than 70c, although admittedly it's kinda cold where I am like 15-20c ambient).

So the main reason I am posting here is to ask about something I'm seeing in HW Info 64 v5.33.
Sensors are showing my cpu cores running at 4028Mhz when not in use, with occasional drops of 1 core at a time going down to just under 1700Mhz. Under load all cores run at just over 3700Mhz consistently.

So the bios and benchmarks say my CPU is running at 3.57 Ghz, while HW Info shows it running at ~3.7-4.0 Ghz.

Is this just the boost clock putting in work? Anybody else have any experience with the 2600 non k?
post #2 of 7
You can see the Turbo Boost table for your 2600 here (click on the text "2nd Generation Intel® Core™ i7 Desktop Processors" for the table to fold out):

http://www.intel.com/content/www/us/en/support/processors/000005523.html

The four values in the table are used depending on how many cores are sleeping and running at any given moment. The highest frequency is 3.8 GHz, which will be used when just one single core is running and the other three cores are sleeping.

That 3.8 GHz means it's a multiplier 38 with the normal 100 MHz BCLK. You've set your system to 106, and then you get this:

38 * 106 MHz = 4028 MHz

I'd guess that's where those strange speeds you saw came from.

The possible values according to that Turbo Boost table from Intel's website would be this here:

3710 = 35 * 106
3816 = 36 * 106
3922 = 37 * 106
4028 = 38 * 106
post #3 of 7
Thread Starter 
Thanks man, this clears up my confusions. I had seen values of 3.5 and 3.8 for the boost clock but didn't know which was right as I'd not seen it running at 3.8 and didn't realise that was for single core boost, and that also explains why it goes down instantly when i run a bench.

Do you know what the risks are like of trying a higher bclk OC? If I tick it up by one at a time until it no longer passes a stability test (x264) is there much risk of permanent damage? I realise I'm also overclocking my ram and adjusting other things beside the cpu clock with this method, how risky is that? Did I just get lucky to even get this kind of OC?
post #4 of 7
The main thing to worry about with overclocking in general is just that you might get corrupted data in files. That's not permanent, but it could be pretty annoying. I had to reinstall Windows after experimenting and crashing a lot. If you have to work on important documents or something, I'd only do it after your experiments are done and you are confident in your new settings running stable.

I've seen people a bit more worried about BCLK changes compared to multiplier because it also changes the way the CPU talks to other parts on the motherboard, like the graphics card in the PCI-E slot or the RAM. The worry is then that something bad might happen in those other parts. I don't know about anything ever actually breaking for anyone.

Does your motherboard allow changing the multiplier and changing voltage? You could add that to your BCLK changes. The highest Turbo Boost multipliers the CPU itself accepts is +4 added to the four stock values if I remember that correctly. This would mean you could set Turbo Boost to 39, 39, 39, 39 and would get a flat 3.9GHz (with "flat" I mean no changes depending on number of sleeping cores). With your 106 MHz BCLK, you'd get 4134 MHz.
post #5 of 7
Thread Starter 
No my mobo doesn't allow anything other than a bclk change. Cheap mobo and non K processor, I'm happy to have gotten anything. Thanks for the help I'll give another few % a shot.
post #6 of 7
If you are getting 106 mhz on your 2600 your doing great, mine and many more won't pass 104 mhz without blue screening, and that's with the k version, just keep track of what voltage your cpu is using .
post #7 of 7
Thread Starter 
Yeah today I replaced the thermal paste on my stock cooler and struggled to even get it back to 106 lol. Now that I have it's managing stress tests just fine, max voltage is 1.266V, I think the cpu is rated at 1.25V. Maxed out at 82 Celsius and 77W. Good to know I got lucky.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Intel CPUs
Overclock.net › Forums › Intel › Intel CPUs › i7 2600(Non-K) overclock, possible HW Info 64 bug?