SB Overclocking Thread

Page 9 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

kashwashwa

Member
Sep 13, 2006
90
0
66
In search for some SB OC results I came across this OC article from bittech:
http://www.bit-tech.net/hardware/cpus/2011/01/07/how-to-overclock-the-intel-core-i5-2500k/2

it seems to suggest slight tweeking in non vcore factors to stabilize the OC like PLL etc. For those of you looking to squeeze more out of SB, check the article out, has some useful suggestions to try.

I tried using their Gigabyte suggestions, but found it is no more stable, and just runs about 5-10 degrees hotter overall. I wouldn't recommend these suggestions.

So far, I'm 100% stable with i5-2500k @ 4.2Ghz with 1.325v set in BIOS (in Win7, it's telling me voltage is 1.35 though), on stock cooler

4.4Ghz gives me a BSOD either right after OS loads, or I fire up something that fully uses CPU.

Hopefully it'll do better once my Coolermaster 212 plus gets here.
 
Last edited:

SuPrEIVIE

Platinum Member
Aug 21, 2003
2,538
0
0
Nope, go ahead, it be fine



Noted the bios still is somewhat messy.
in regard, my gigabyte x58 took a year to get a good bios.


Yeah i coulndt find an answer to that at all for sb dont get why reviewers dont test it out.

Does memory speed matter on sb for OC? since baseclk is not used it would always be asynch anyway? according to the bit tech review however they mentioned anything over 1600 rate is not worth it.

Also so if igo single 4gb is it ok to later on get one 2gb stick for 6gb in all provided that the difference btw each other is just ram storage?
 

pm

Elite Member Mobile Devices
Jan 25, 2000
7,419
22
81
I bought my 2600K this morning. I just need to choose a motherboard now and decide if I want to upgrade my memory.
 

Hogan773

Senior member
Nov 2, 2010
599
0
0
It is looking like 4.4 to 4.6 is the norm....


And just to be clear, this is ALL FOUR CORES MAXED right? If its only 1 core, then its not much better than the Intel Limited OC which could take 1 core up to 4.2 anyway even on a non-K chip.

I'm just trying to determine how much pop I'll get in OCing and therefore how much value I place on it in deciding between P67/OC/GPU and H67/no OC/No GPU
 

Markfw

Moderator Emeritus, Elite Member
May 16, 2002
25,752
14,783
136
And just to be clear, this is ALL FOUR CORES MAXED right? If its only 1 core, then its not much better than the Intel Limited OC which could take 1 core up to 4.2 anyway even on a non-K chip.

I'm just trying to determine how much pop I'll get in OCing and therefore how much value I place on it in deciding between P67/OC/GPU and H67/no OC/No GPU

Well, mine certainly is all 4 WITH HT. At least 10 hours @ 100% load stable.
 

coffeejunkee

Golden Member
Jul 31, 2010
1,153
0
0
So my Vcore as logged by HWmonitor is 1.41 max and I think that's high (most people do). It normally sits at 1.35 in load but can spike up, and at idle its 0.928 or something. But, I went to my BIOS and switched my Vcore from Auto to 1.35. The problem with this is that I would still like it to lower to sub-1V when idle.

Is there a way to allow it to lower but force it to a 1.35 max?

There's a couple of things about vcore:

- Vdrop: every mobo has vdrop. The vcore you see in cpu-z will always be a little lower than what you specified in the bios/uefi. Some mobo's have worse vdrop than others.

- Vdroop: this is an intended mechanism as specified by Intel. Simply put, it adjusts the vcore under load in order to prevent damage by short voltage spikes that might occur when switching to load state. In cpu-z you will see the vcore fluctuate during load.

Vdroop makes overclocking harder. The difference between the lowest and highest vcore value will grow when you increase vcore in bios. Lets say at stock, vcore under load is between 1.2 and 1.1V. If you increase vcore it might become 1.4 and 1.2V. So the difference between the highest and the lowest value has increased from 0.1V to 0.2V.

Suppose your cpu at a certain frequency needs a minimum of 1.3V to remain stable. As you can see the highest value of 1.4V is plenty but the minimum of 1.2V is too low. So you will need to increase vcore more, which will make the difference even bigger. You might need 1.55V to ensure a minimum of 1.3V. My numbers are exaggerated to make the example more clear but this is what you are noticing in HWmonitor.

To make overclocking easier most mobo's have a Load Line Calibration or VDroop option. This option enables or disables the Vdroop mechanism. With LLC enabled you will see no more fluctuating vcore under load, or just very slightly.

There's some debate about wheter enabling LLC has harmful effects. I wouldn't be too worried about it since there are many overclockers using it and running their cpu's fine for years. The alternative isn't very attractive. With vdroop enabled you'll need to overcompensate on the vcore causing high spikes. On the other hand, if your oc stays stable within safe voltage margins with Vdroop enabled, I see no reason to disable it.

- C1E/EIST: both these technologies lower the vcore and multiplier in idle. The difference is that EIST can adjust the multiplier and vcore according to load (fex. only x25 with 1.1V under light loads while the max would be x30/1.2V), whereas C1E will always use the highest multiplier and vcore, light or heavy load.

Both C1E and EIST can only lower the vcore in idle if it is set to 'auto' or using an offset/dynamic mode. If you set the vcore to a fixed value EIST and C1E will not lower it in idle.

The problem with 'auto' is that mobo's generally apply way too much extra voltage when increasing the multiplier or bclk. Therefore auto isn't very usefull for overclocking.

Most mobo's (except MSI afaik) nowadays have a dynamic vcore option. You can lock the vcore to its standard values and specify an offset value, either positive or negative. Now C1E and EIST can lower the vcore in idle even when overclocked.

However, altough a nice feature, dynamic vcore is not completely necessary. There's a third power saving mechanism called C-states or C3/C6, which basically turns off the cpu even more than C1E. If you leave C-states enabled idle power consumption will be about the same regardless of lowered vcore or not.

It would be interesting to learn if C-states cause any instability at all when overclocking SB.
 
Last edited:

namtran512

Member
Jan 2, 2011
78
0
0
It is looking like 4.4 to 4.6 is the norm....



If you got lucky like me you can push past 5 on 1.35v, but it's not very stable. I could put it to 1.4 but I'm not gonna fry my chip for the sake of saying I broke the 5 ghz barrier




Most people should hit 4.5 on 1.3 volts easy on stock. With a better cooler you can go abit higher on 1.3 to 1.35 volts, but 1.35 is the highest I would go and that's pretty dangerous territory right there to be honest.



I know some people have been able to hit 5 ghz on air with a little bit of extra juice past 1.3, but that's an exception so far.
 

d33pblue

Senior member
Jul 2, 2003
225
1
81
So the best OC I've seen so far from this thread is a 2600k. It would make sense that the absolute best chips are 2600ks and not 2500ks. It might make sense for those on the fence to just go ahead and get the 2600k. You'll not only get HT and more cache, but you'll likely get more overclocking headroom as well.

I still haven't made up my mind yet as to which one I'll pick up
 

Makaveli

Diamond Member
Feb 8, 2002
4,760
1,159
136
It is looking like 4.4 to 4.6 is the norm....

This!

I've only seen a few chips hit 5Ghz its definitely not the norm.

for 24/7 Stable overclock I don't really see 5ghz being an option unless you want to fry your chip.

I think we will see 5ghz normal overclocks with Ivy bridge or maybe SB-E later this year.
 

Markfw

Moderator Emeritus, Elite Member
May 16, 2002
25,752
14,783
136
If you got lucky like me you can push past 5 on 1.35v, but it's not very stable. I could put it to 1.4 but I'm not gonna fry my chip for the sake of saying I broke the 5 ghz barrier




Most people should hit 4.5 on 1.3 volts easy on stock. With a better cooler you can go abit higher on 1.3 to 1.35 volts, but 1.35 is the highest I would go and that's pretty dangerous territory right there to be honest.



I know some people have been able to hit 5 ghz on air with a little bit of extra juice past 1.3, but that's an exception so far.

I can't even start loading windows at 4.5 and 1.37 vcore, so I would not go that far.
 

namtran512

Member
Jan 2, 2011
78
0
0
I can't even start loading windows at 4.5 and 1.37 vcore, so I would not go that far.



Depends on your chip, I got it running rock solid at 4.8 on 1.35 vcore and stressed it quite abit and it didn't crash. However, to be quite honest, the difference between say 4.5 ghz and 5 ghz is not big at all, and the only people who are going to see that difference is if they do some extremely heavy workloads with their CPUs.



Even 4 ghz is overkill for most practical usage, even intense gaming. I mean the numbers people are putting out on stock volts is quite hilarious already (4 ghz on stock volts for instance is already quite ridiculous). Anything past the 4.4-4.6 ghz mark is just icing on the cake to be honest, and the risk you run vs the realistic gains you get from that little bit of extra speed is not worth it.
 
Last edited:

996GT2

Diamond Member
Jun 23, 2005
5,212
0
76
I can't even start loading windows at 4.5 and 1.37 vcore, so I would not go that far.

Is that 1.37 in BIOS or 1.37 in Windows?

Mine's perfectly stable at 1.26V under load, but that actually is 1.32V in BIOS since I don't have LLC enabled.

Also, how do motherboards calculate the VCore when you use the "Auto" setting? Are there predefined VCore settings depending on speed?
 

Xed

Golden Member
Nov 15, 2003
1,453
0
71
4.2 at 1.232v so far. Will see how far I can take it in time =P

edit - 4.5 @ 1.264
 
Last edited:

Markfw

Moderator Emeritus, Elite Member
May 16, 2002
25,752
14,783
136
1.37 in bios. I can't even get the splash screen, let alone boot @ 4.5
 

Arkainium

Member
Sep 25, 2007
44
0
0
Most mobo's (except MSI afaik) nowadays have a dynamic vcore option. You can lock the vcore to its standard values and specify an offset value, either positive or negative. Now C1E and EIST can lower the vcore in idle even when overclocked.

How is this "standard" value determined? When I use the offset feature on my Asus P8P67 Pro it appears that the offset is applied to the value determined as when set to Auto.

Also, how do motherboards calculate the VCore when you use the "Auto" setting? Are there predefined VCore settings depending on speed?

I'm curious about this myself. Perhaps it has something to do with VID?
 

Dadofamunky

Platinum Member
Jan 4, 2005
2,184
0
0
Even 4 ghz is overkill for most practical usage, even intense gaming. I mean the numbers people are putting out on stock volts is quite hilarious already (4 ghz on stock volts for instance is already quite ridiculous). Anything past the 4.4-4.6 ghz mark is just icing on the cake to be honest, and the risk you run vs the realistic gains you get from that little bit of extra speed is not worth it.

The higher the mean performance gets, the less difference an OC actually makes. It's noticeable in benchies and in occasional intensive apps, but for the most part it's just a fun way to measure e-peen. That said, it's cool that Intel saw fit to release a couple of unlocked SB processors at reasonable prices. (And their margins are at record levels!) The last time OCing was this simple was in the days of the Athlon X2, when all you really had to do was tweak the multiplier. People forget that. (My wife right now runs an Athlon X2 3800+ 1.8 GHz stock OCd to 2.5 GHz. I built it for her out of parts in my old company's tossed-out leftovers 2 1/2 years ago and it still runs rock solid - it's so solid it's outlived itself.

First thing I'll be doing when I finally build my rig (likely next week) is run LinX and other benchies on stock config and then on OC config.
 

coffeejunkee

Golden Member
Jul 31, 2010
1,153
0
0
How is this "standard" value determined? When I use the offset feature on my Asus P8P67 Pro it appears that the offset is applied to the value determined as when set to Auto.



I'm curious about this myself. Perhaps it has something to do with VID?

That is how it is supposed to work. On stock settings (normal multiplier) 'auto' will use the voltages that are identified within your specific cpu, which are indeed read from the VID table. The offset feature allows you to lock the vcore to these values.

If you'd leave the vcore on 'auto' and up the multiplier it will increase the vcore, usually much more than necessary.
 
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |