Overclock.net › Forums › Intel › Intel CPUs › 5GHz and Beyond! - The Quest for a Stable Overclock
New Posts  All Forums:Forum Nav:

5GHz and Beyond! - The Quest for a Stable Overclock - Page 3

post #21 of 45
Although SB isnt as complicated in that the cpu overclock isnt associated with the memory, theres still things like PLL voltage being too high or too low that can cause instability.

Its just from personally experience, i found i saved myself alot of headaches from starting from scratch and writing it all down, so that I could point out where I was going wrong.
post #22 of 45
Thread Starter 
Quote:
Originally Posted by $ilent View Post
Although SB isnt as complicated in that the cpu overclock isnt associated with the memory, theres still things like PLL voltage being too high or too low that can cause instability.

Its just from personally experience, i found i saved myself alot of headaches from starting from scratch and writing it all down, so that I could point out where I was going wrong.
I definitely agree about writing it all down. This is the first time I've done it digitally. Gotta love Google docs!

So, question, do you find that once you get a stable vcore, vccio, and pll at, say 4.5 GHz, does that help you in picking the vcore, vccio, and pll for a higher OC?
post #23 of 45
Well once I find my stable volts, I try to raise the multiplier. Once I get BSod 101 it needs more vcore, 124 means higher/lower PLL, and failed worker during p95 means more vcore.
post #24 of 45
Nick read the bsod / freezing on sandy? link in my sig, more specifiaclly the section about VCCIO and PLL, it's a lenthy read but one that could probably help you on your way. If you're recieving bsod 124 during stress testing then there should be a link in the thread labeled 'possible way to remove BSOD 124' or something, have a read on that also and you might find that the method may work for you too.

Hope that helps
it looks alive
(17 items)
 
Sandy Transformed
(69 photos)
 
CPUMotherboardGraphicsRAM
i5 2500K [5ghz 1.44v 24/7] Asus Maximus IV Extreme SLI GTX 460 @ 920/1840/4100 [1.11v] 4GB Dominator GT @ 2133mhz CL8 [1.65v] 
Hard DriveHard DriveCoolingOS
Intel X25-M G2 80GB 3x Samsung F3 500GB RAID 0 EK Supreme HF Full Copper Windows 7 Ultimate 64bit 
MonitorPowerCaseAudio
LG42SL9500 42" 1080p [Surround Soon] XFX XXX Edition 650w Xigmatec Elysium JVC HA-RX700 [foam mod] 
  hide details  
Reply
it looks alive
(17 items)
 
Sandy Transformed
(69 photos)
 
CPUMotherboardGraphicsRAM
i5 2500K [5ghz 1.44v 24/7] Asus Maximus IV Extreme SLI GTX 460 @ 920/1840/4100 [1.11v] 4GB Dominator GT @ 2133mhz CL8 [1.65v] 
Hard DriveHard DriveCoolingOS
Intel X25-M G2 80GB 3x Samsung F3 500GB RAID 0 EK Supreme HF Full Copper Windows 7 Ultimate 64bit 
MonitorPowerCaseAudio
LG42SL9500 42" 1080p [Surround Soon] XFX XXX Edition 650w Xigmatec Elysium JVC HA-RX700 [foam mod] 
  hide details  
Reply
post #25 of 45
Thread Starter 
Quote:
Originally Posted by munaim1 View Post
Nick read the bsod / freezing on sandy? link in my sig, more specifiaclly the section about VCCIO and PLL, it's a lenthy read but one that could probably help you on your way. If you're recieving bsod 124 during stress testing then there should be a link in the thread labeled 'possible way to remove BSOD 124' or something, have a read on that also and you might find that the method may work for you too.

Hope that helps
I've read through that (and actually all of your posts on OCing SB--You are amazing!), but to be honest it's all quite overwhelming and I'm not sure what I'm doing is correct.

So, my current strategy is as follows:

I've got my Vcore at 1.47, and I'm currently going through the PLL voltages like Silent recommended. I'm setting VCCIO at 1.1 volts for now unless anyone recommends otherwise.
post #26 of 45
Quote:
Originally Posted by Nick2253 View Post

How likely is it that OS or software can cause 124 BSODs? I just noticed that, in my resent restarts, uTorrent doesn't start up automatically any more. When I looked at its settings, the "start with Windows" box was disabled. I mean, this could have just been a fluke, but is it possible I'm getting OS/software corruption? I mean, I have no problem reinstalling Windows (it's not like I put a whole lot of time into the setup, and only installed a few basic programs), but if this could be causing me to artificially interpret an unstable OC when it is actually stable, then I'd be worried.
It's a possibility but in all likelihood not the case. This is why most people opt to OC after a fresh install of their OS (or on a non-primary OS). Just to be able to rule out software as the source of instability.

I've done a lot of OC'ing on an OS jam packed full of different software (Utorrent included) and software has never been the cause of instability, it is always fixed by tweaking a BIOS setting.
Edited by Heazy - 10/31/11 at 6:29am
Work Horse 2.0
(13 items)
 
  
CPUMotherboardGraphicsRAM
i5 2500k @ 4.8ghz 1.410v ASRock P67 Extreme 4 Galaxy GeForce 660ti 3GB 8G G.Skill Sniper 
Hard DriveOSMonitorPower
Corsair 120GB SSD Windows 7 Ultimate x64 ASUS VG248QE 144hz RR Corsair 750W 
CaseMouse
Antec 1200 Logitech G5 
  hide details  
Reply
Work Horse 2.0
(13 items)
 
  
CPUMotherboardGraphicsRAM
i5 2500k @ 4.8ghz 1.410v ASRock P67 Extreme 4 Galaxy GeForce 660ti 3GB 8G G.Skill Sniper 
Hard DriveOSMonitorPower
Corsair 120GB SSD Windows 7 Ultimate x64 ASUS VG248QE 144hz RR Corsair 750W 
CaseMouse
Antec 1200 Logitech G5 
  hide details  
Reply
post #27 of 45
Thread Starter 
I finished by PLL iterations. So you know, I manually set VCORE to 1.47V and VCCIO to 1.1V for these runs.

In all cases, I was able to boot into Windows and start P95, however I inevitably got a 124 BSOD after running P95 on blend for the duration listed below (in hours).

1.875 @ 1.6
1.85 @ 0.9
1.825 @ 0.9
1.8 @ ?? <- I just realized I never ran this. I'll update when I do.
1.775 @ 0.7
1.75 @ 3.2
1.725 @ 0.05
1.7 @ 0.75
1.675 @ 2.9
1.65 @ 4.2
1.625 @ 0.9
1.6 @ 1.25
1.575 @ 0.3
1.55 @ 1.75
1.525 @ 2.75
1.5 @ 1

So, based on the instructions in munaim's other thread, I'm now going to pick the best PLL voltage (which in this case is 1.65V), and then iterate through the VCCIO voltages from 1.00V to 1.15V.

Should I be using smaller steps? Is there anything I'm missing?

As always, thanks so much for all the help!
post #28 of 45
Thread Starter 
Hey guys! while I'm in the process of iterating through VCCIO voltages, I've got a couple more questions:

1. What is the max VCCIO voltage?

2. Does RAM speed effect stability in conjunction with VCCIO? I mean, I'm currently underclocking my RAM, but eventually I'll want to bring it back to stock and/or beyond. Will I have to increase VCCIO when I increase the speed of my RAM?

3. When should I start increasing/decreasing VCORE? I mean, every test I've ever run with this processor has always ended in a 124 BSOD. Does this mean I just haven't found the right PLL/VCCIO sweet spot yet, or could it mean I've got too much/little VCORE?

4. In order to rule out software problems, I'm going to install Windows 7 on one of my spare Raptor HDDs, add just the chipset drivers, P95 and monitoring software, and start stressing with that. Should I update Windows 7 on that install, or could the updates from Windows cause a stability problem? Should I at least install SP1?

Thanks again for the help. And my God Munaim, you are quite right when you say (time and time again I might add) that OCing is all about patience. Do you have any tips for me to speed it up a little bit though ?
post #29 of 45
Quote:
Originally Posted by Nick2253 View Post
So, based on the instructions in munaim's other thread, I'm now going to pick the best PLL voltage (which in this case is 1.65V), and then iterate through the VCCIO voltages from 1.00V to 1.15V.

Should I be using smaller steps? Is there anything I'm missing?

As always, thanks so much for all the help!
Definitely use small steps, here's my own testing I did a couple hour's ago with my new chip:

Quote:
Originally Posted by munaim1 View Post
I found the best time to tweak those votlages is when I worker fails or you get a bsod 124 and nearing stability.

I'm currently doing the same for my new chip, it's crap compared to my old one but atleast it does 5ghz even though it takes 1.44v to get there.

Using Custom Prime blend 1344 first, here's my PLL and VCCIO testing method:

Offset 0.090+ 1.424 LOAD

PLL Auto (1.806v)
VCCIO Auto (1.058-1.072) - 1min 10s Prime froze.


Now lets see what the PLL voltage and VCCIO can do:

First PLL:
1.50000v - 44sec BSOD 124
1.50625v - 49sec Worker failed
1.51250v - 46sec Worker failed


Offset 0.095+ 1.432 LOAD
Back to PLL:
1.52500v - 45sec BSOD 124
1.53125v - 41sec Worker failed
1.53750v - 1min 02sec BSOD 124
1.54375v - 47sec BSOD 124
1.55000v - 53sec Prime froze then BSOD 9C
1.55620v - 1min 02sec BSOD 124
1.56250v - 58sec Worker failed
1.56875v - 47sec BSOD 124

Offset 0.100+ 1.440 LOAD
Again continuing with PLL:
1.56875v - 1min 17sec BSOD 124
1.57500v - 55sec BSOD 124
1.58125v - 50sec Worker Failed
1.58750v - 12min 39sec BSOD 101
1.59375v - 1min 02sec Worker Failed
1.60000v - 9min 52sec Worker Failed
1.60625v - 53sec Worker Failed

That's where I'm currently at but as you can see, even after I increased the vcore few times it wasn't making a difference but with the right PLL prime does last a lot longer compared to the other values. I managed to get over 10mins with 1.58750v where nearly all of them where below a minute. I might just keep it there and mess around with the vccio.

I think I'll update this as I go along. Hopefully I get this chip stable at 5ghz with less than 1.45v.




Quote:
Originally Posted by Nick2253 View Post
Hey guys! while I'm in the process of iterating through VCCIO voltages, I've got a couple more questions:

1. What is the max VCCIO voltage?

2. Does RAM speed effect stability in conjunction with VCCIO? I mean, I'm currently underclocking my RAM, but eventually I'll want to bring it back to stock and/or beyond. Will I have to increase VCCIO when I increase the speed of my RAM?

3. When should I start increasing/decreasing VCORE? I mean, every test I've ever run with this processor has always ended in a 124 BSOD. Does this mean I just haven't found the right PLL/VCCIO sweet spot yet, or could it mean I've got too much/little VCORE?

4. In order to rule out software problems, I'm going to install Windows 7 on one of my spare Raptor HDDs, add just the chipset drivers, P95 and monitoring software, and start stressing with that. Should I update Windows 7 on that install, or could the updates from Windows cause a stability problem? Should I at least install SP1?

Thanks again for the help. And my God Munaim, you are quite right when you say (time and time again I might add) that OCing is all about patience. Do you have any tips for me to speed it up a little bit though ?

1. Max Vccio recommended for 24/7 usage is 1.2v, however I've not really seen anyone using over 1.15/1.16v.

2. RAM should be at default stock settings, when overclockign RAM, VCCIO can help with stability, however when you're concentrating on the cpu clock I recommend leaving the RAM at stock (timings and voltage - Just set it to XMP).

3. Once you take the time to find the PLL and VCCIO that actually makes a differnence you'll know what to do. You'll find a pattern and it certainly helps when you write it down.

4. Sounds like a very good idea!!! Don't see a problem with installing SP1.

Oh yes, patience and pen and paper can do wonder's!!!! I'm still going through my settings with this new mobo and chip, but it's actually going okay. Really doesn't take long
it looks alive
(17 items)
 
Sandy Transformed
(69 photos)
 
CPUMotherboardGraphicsRAM
i5 2500K [5ghz 1.44v 24/7] Asus Maximus IV Extreme SLI GTX 460 @ 920/1840/4100 [1.11v] 4GB Dominator GT @ 2133mhz CL8 [1.65v] 
Hard DriveHard DriveCoolingOS
Intel X25-M G2 80GB 3x Samsung F3 500GB RAID 0 EK Supreme HF Full Copper Windows 7 Ultimate 64bit 
MonitorPowerCaseAudio
LG42SL9500 42" 1080p [Surround Soon] XFX XXX Edition 650w Xigmatec Elysium JVC HA-RX700 [foam mod] 
  hide details  
Reply
it looks alive
(17 items)
 
Sandy Transformed
(69 photos)
 
CPUMotherboardGraphicsRAM
i5 2500K [5ghz 1.44v 24/7] Asus Maximus IV Extreme SLI GTX 460 @ 920/1840/4100 [1.11v] 4GB Dominator GT @ 2133mhz CL8 [1.65v] 
Hard DriveHard DriveCoolingOS
Intel X25-M G2 80GB 3x Samsung F3 500GB RAID 0 EK Supreme HF Full Copper Windows 7 Ultimate 64bit 
MonitorPowerCaseAudio
LG42SL9500 42" 1080p [Surround Soon] XFX XXX Edition 650w Xigmatec Elysium JVC HA-RX700 [foam mod] 
  hide details  
Reply
post #30 of 45
Thread Starter 
Once again, thank you so much! SB is so different from everything else I've done that I'm kinda going crazy.

So, couple questions:

1. Why did you increase your offset during your PLL runs?

2. Is my "goal" to get either a worker failure or a 101 BSOD? Up until now its been constant 124s.

3. I'm using the default blend. Should I instead be using 1344? Will that speed up my testing?

4. Is my strategy of underclocking my RAM to get a better OC a good one, or should I just be running it at stock right now?

5. Is there a way to monitor PLL and VCCIO voltages from within Windows itself?

Again, thanks so much. You are a life saver!
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Intel CPUs
Overclock.net › Forums › Intel › Intel CPUs › 5GHz and Beyond! - The Quest for a Stable Overclock