The Magic 5GHz.

[Deleted]

I wanted to try out today the 5GHz/ Ive set up the multiplier to 50, and BCLK to 100. Voltage: 1.445.
Currently performing a Prime test.
Temperatures can go up to 81 degrees (core 2 mostly).

While I was at 4.8GHz, voltage 1.345, the temperatures didn't go higher than 66 degrees.

I'm using an i7 2600k btw, with a Noctua NH D-14 and HAF 932 case.

Any observations?

---------- Post added 2011-11-20 at 04:19 AM ----------

Core 1 and 4 didn't go higher than 73 and 76 respectively.

[Jervaise]

with my 2600k i can go up to 4.9 with no problems, cannot achieve 5.0 no matter how hard i try and tweak. don't know what that is about though...

[Deleted]

What you mean by "cannot achieve"? Temps go too high? BSODs?

[Kankipappa]

http://www.youtube.com/watch?v=zUc6znC848o

Legendary 5GHz in year 2003 :')

[xXNyteStalkerXx]

Ive seen your setup work with 5.0 but in all fairness whats the point you wont notice a huge difference and/or spends tons of money for something too possibly stop working.So too your question you can go 5.0 with your setup i just recommend not doing that.(Just make sure too watch your temps carefully and check your work with more then one program just in-case you get different readings.)

[Deleted]

http://www.youtube.com/watch?v=zUc6znC848o

Legendary 5GHz in year 2003 :')

Yeah, I know
For me, between 2003-2004 until today, was a big gap in computer updates, I just didn't took the time to read these things anymore, and now that I was "back", I was surprised to see that CPUs are still at this frequency

[Xuvial]

with my 2600k i can go up to 4.9 with no problems, cannot achieve 5.0 no matter how hard i try and tweak. don't know what that is about though...

It's called the multiplier wall, all SB chips have one to varying extents. There's a "soft" wall where after a certain point you'll suddenly require a huge increase in vcore while only jumping up 100-200mhz, this can occur anywhere between 4.4ghz and 4.9ghz from what I've seen. Getting a stable OC as close to the soft wall as possible is the best thing to do if you want to see a good balance of vcore, mhz and temps. For me it was 4.5ghz/1.31v but I decided to push to 4.8/1.4v anyway.
Then there's a "hard" wall where your chip will simply refuse to go any higher regardless of how much you increase vcore, unless you cross the 1.5v safe limit. Many chips see this at 4.9-5ghz.

@Dangel7, 5ghz/1.445v is pretty damn good, 4.8/1.345v is even more impressive. Are you doing this with PLL overvoltage enabled or disabled? Any other variables we should be aware of e.g. HT disabled/enabled?
In any case 4.8/1.345v is definitely where I'd leave it, it seems to be your sweet spot. That extra 200mhz isn't worth the increase in vcore and temps.

[Deleted]

My temperatures went as high as 86 degrees now, with the 2nd phase (I think), of the Prime test.

[BicycleMafioso]

Sandy Bridge are mostly voltage capped rather than heatcapped.
Meaning one isn't likely to push forward with watercooling.

[Deleted]

It's called the multiplier wall, all SB chips have one to varying extents. There's a "soft" wall where after a certain point you'll suddenly require a giant increase in voltage to get a stable OC, this can occur anywhere between 4.4ghz and 4.9ghz from what I've seen. Then there's a "hard" wall where your chip will simply refuse to go any higher regardless of how much you increase vcore, unless you cross the 1.5v safe limit. Many chips see this at 4.9-5ghz.

@Dangel7, 5ghz/1.445v is pretty damn good, 4.8/1.345v is even more impressive. Are you doing this with PLL overvoltage enabled or disabled? Any other variables we should be aware of?
In any case, 4.8/1.345v is definitely where I'd leave it in your case. That extra 200mhz isn't worth the increase in vcore and temps.

PLL overvoltage is enabled.

---------- Post added 2011-11-20 at 04:37 AM ----------

Was reading about somebody today, that got as far as 5.9GHz, with 1.6 voltage

[Xuvial]

Ouch, with PLL enabled at those voltages you could see micro-spikes of 1.5v+, just something to keep in mind. 86c is pretty damn hot (though not critical), personally I steer clear of any OC that pushes temps beyond 75c during blend testing. Intel Burn could probably push your chip to 90c+.
Settle for your 4.8ghz OC brah 5ghz is overrated.

[Jervaise]

I'm currently using it with 4.7 1.330 temps don't go higher than 70, using v6gt which was a promotion product available (it's quite terribad actually)

[Deleted]

Ouch, with PLL enabled at those voltages you could see micro-spikes of 1.5v+, just something to keep in mind. 86c is pretty damn hot (though not critical), personally I steer clear of any OC that pushes temps beyond 75c during blend testing. Intel Burn could probably push your chip to 90c+.
Settle for your 4.8ghz OC brah 5ghz is overrated.

Yeah, but it keeps it more stable.
As for the temperatures, and high frequency, I agree, 200 MHz for 20 degree diferrence is just not worth it, already pulled it back to 4.8GHz
Just wanted to see how far I go with it and be more or less stable.

---------- Post added 2011-11-20 at 05:00 AM ----------

I'm currently using it with 4.7 1.330 temps don't go higher than 70, using v6gt which was a promotion product available (it's quite terribad actually)

Impressive numbers

[Jervaise]

slightly off topic,

1) could anyone suggest me a software to monitor my voltage live, cause CPU-Z sees it as 1.001 which is not correct. I want to be able to properly share my numbers if needed
2) i just realized I can't use Gigabyte Touch Bios due to a NET framework error. I don't use the windows bios anyway, still it bugs me not being able to fix it. Any suggestions?

cheers

[Deleted]

slightly off topic,

1) could anyone suggest me a software to monitor my voltage live, cause CPU-Z sees it as 1.001 which is not correct. I want to be able to properly share my numbers if needed
2) i just realized I can't use Gigabyte Touch Bios due to a NET framework error. I don't use the windows bios anyway, still it bugs me not being able to fix it. Any suggestions?

cheers

Windows BIOS? how can you get there? (I also noticed in the regular BIOS some settings that stated "can be changed in OS").

[Deleted]

http://www.youtube.com/watch?v=zUc6znC848o

Legendary 5GHz in year 2003 :')

Quite funny watching that vid tbh, the lapping of the no2 pot was awefull, and the application of thermal paste could have been better too.. i guess with lno2 temperatures aren't an issue. Plyed with dry ice on a p4 myself back in the day. pentium d920 i thin kit was, managed to get it to 5.1ghz, must have been a good batch though, it was running @ 4.4ghz on air for most of it's life ^_^

Back on topic thoguh - in what way is 5ghz not achievable? Just not happy with the temps? or are you actually getting crashes? Limited by motherboard perhaps?

[Jervaise]

Windows BIOS? how can you get there? (I also noticed in the regular BIOS some settings that stated "can be changed in OS").

ah it's GIGABYTE's Touch Bios software, specific to the Z68 series motherboards if I'm correct sorry for confusion

[Deleted]

ah it's GIGABYTE's Touch Bios software, specific to the Z68 series motherboards if I'm correct sorry for confusion

Using an ASUS z68 motherboard.
Still I would guess you need to install a software in windows, to get to controll the BIOS from it?
Also, I could not manage to install the Asus TurboV application for some strange reason.
I tried more downloads, 1 of them stated that I don't have a z68 MOBO (I do ><), and the other one kept asking me for windows permission to install, then after I confirmed them all, nothing else happent...

[DeltrusDisc]

Any reason you increased it by a whopping .1v? You should increase in .005v or .01v increments...

[Deleted]

Any reason you increased it by a whopping .1v? You should increase in .005v or .01v increments...

Me? Where did I say that?

Edit: Oh, you say that because you noticed the raise from 1.345 to 1.445?
No, I obviously didn't whopped by 0.1v that was a pure coincidence