Overclock.net › Forums › Intel › Intel CPUs › [Official] The Ivy Bridge Stable / Suicide Club **Guides, Voltages, Temps & BIOS Templates** Inc SPREADSHEET
New Posts  All Forums:Forum Nav:

[Official] The Ivy Bridge Stable / Suicide Club **Guides, Voltages, Temps & BIOS Templates** Inc... - Page 46

post #451 of 3380
5/17/12 at 1:37am
Quote:
Originally Posted by Nihilo View Post

I was using those as an example of why LinX isn't a good test to consider your self stable on. It makes my processor run at lower volts meaning it's not pushing it as hard as P95.

I'm not sure about this. Yes Prime95 is good to test stability but it doesn't seem to need as much voltage as IBT - I am not using any LLC and am using offset. I had set the offset so that when running Prime95 voltage dropped to 1.12v at 4.2ghz and this passed a 22 hour run. However when running IBT or x264 the voltage dropped further to 1.112v causing whea warnings to appear. Then the penny dropped that voltage decreases proportionately to load when not using any LLC. So I upped the voltage so that when running IBT / x264 the voltage did not get lower than 1.12v. No more whea warnings.

I think Prime95 should be used as one part of a stability testing strategy, not solely.
Alpha II
(15 items)
  		  
CPUMotherboardRAMHard Drive
Intel 3770K Asus P8Z77-V Pro Corsair 8GB DDR3 1600MHz Vengeance Crucial M4 SSD 
Hard DriveOptical DriveCoolingOS
Samsung F3 LG GGW-H20L Thermalright Archon Windows 7 Ultimate 64-Bit 
PowerCase
Seasonic S12II-430W Coolermaster CM 690 ll 
View all
  hide details  
Reply
Alpha II
(15 items)
  		  
CPUMotherboardRAMHard Drive
Intel 3770K Asus P8Z77-V Pro Corsair 8GB DDR3 1600MHz Vengeance Crucial M4 SSD 
Hard DriveOptical DriveCoolingOS
Samsung F3 LG GGW-H20L Thermalright Archon Windows 7 Ultimate 64-Bit 
PowerCase
Seasonic S12II-430W Coolermaster CM 690 ll 
View all
  hide details  
Reply

Gear mentioned in this thread:

post #452 of 3380
5/17/12 at 1:44am
i was doing some testing yesterday.
It might be just me, dumb luck or what not, but the CPU PLL does affect stability at higher clocks.
At stock it might be stable at 1.5v but towards 4.6ghz i need 1.75v I'll try different runs today and see if it makes a difference at 4.8ghz
My system
(15 items)
  		  
CPUMotherboardGraphicsRAM
Intel i7 3770k 5.0 ghz 1.41v 78c Asus Maximus V Gene bios 1408 Gigabyte Nvidia GTX 670  Corsair XMS3 1600 2x8gb 
Hard DriveCoolingOSMonitor
Sandisk extreme 120 Noctua NH-D14 Windows 7 sp1 Dell u3011 
MonitorKeyboardPowerCase
Asus VG248QE Cooler Master storm Quick fire pro Brown SeaSonic 860w Plat Silverstone Temjin TJ08-e 
Mouse
Logitech performance MX 
View all
  hide details  
Reply
My system
(15 items)
  		  
CPUMotherboardGraphicsRAM
Intel i7 3770k 5.0 ghz 1.41v 78c Asus Maximus V Gene bios 1408 Gigabyte Nvidia GTX 670  Corsair XMS3 1600 2x8gb 
Hard DriveCoolingOSMonitor
Sandisk extreme 120 Noctua NH-D14 Windows 7 sp1 Dell u3011 
MonitorKeyboardPowerCase
Asus VG248QE Cooler Master storm Quick fire pro Brown SeaSonic 860w Plat Silverstone Temjin TJ08-e 
Mouse
Logitech performance MX 
View all
  hide details  
Reply
post #453 of 3380
5/17/12 at 4:25am
Quote:
Originally Posted by samwiches View Post

Yes, no errors that day.
I've noticed something weird for some time now, that often after around 12 hours the results.txt file will show splits in the time stamps. Could be I guess it's just some workers lagging behind the others, as I can look at the on-screen data and see nothing mucked up. So maybe not weird after all, dunno.
Merlin
(17 items)
  		  
CPUMotherboardGraphicsGraphics
4770K Gigabyte GA-Z87X-UD5H PowerColor AX7950 3GBD5-2DHPP PowerColor AX7950 3GBD5-2DH 
RAMHard DriveHard DriveHard Drive
Kingston HyperX Predator Series 16GB KHX18C9T2K... SAMSUNG 830 Series MZ-7PC128N/AM x 3 Seagate Barracuda ST2000DM001 2TB 7200 RPM Western Digital WD Black WD1002FAEX 1TB 7200 RPM 
CoolingCoolingCoolingCooling
Raystorm cpu block Alphacool NexXxoS UT60 Full Copper 480mm  Alphacool NexXxoS UT60 Full Copper 480mm  Swiftech MCP35x 
CoolingOSMonitorPower
Aqua Computer Aquaero 5 Pro Windows 7 Professional Samsung SMS27A350H SeaSonic Platinum 1000 
Case
Case Labs SM8 
View all
  hide details  
Reply
Merlin
(17 items)
  		  
CPUMotherboardGraphicsGraphics
4770K Gigabyte GA-Z87X-UD5H PowerColor AX7950 3GBD5-2DHPP PowerColor AX7950 3GBD5-2DH 
RAMHard DriveHard DriveHard Drive
Kingston HyperX Predator Series 16GB KHX18C9T2K... SAMSUNG 830 Series MZ-7PC128N/AM x 3 Seagate Barracuda ST2000DM001 2TB 7200 RPM Western Digital WD Black WD1002FAEX 1TB 7200 RPM 
CoolingCoolingCoolingCooling
Raystorm cpu block Alphacool NexXxoS UT60 Full Copper 480mm  Alphacool NexXxoS UT60 Full Copper 480mm  Swiftech MCP35x 
CoolingOSMonitorPower
Aqua Computer Aquaero 5 Pro Windows 7 Professional Samsung SMS27A350H SeaSonic Platinum 1000 
Case
Case Labs SM8 
View all
  hide details  
Reply
post #454 of 3380
5/17/12 at 5:14am
Quote:
Originally Posted by iSeries View Post

I'm not sure about this. Yes Prime95 is good to test stability but it doesn't seem to need as much voltage as IBT - I am not using any LLC and am using offset. I had set the offset so that when running Prime95 voltage dropped to 1.12v at 4.2ghz and this passed a 22 hour run. However when running IBT or x264 the voltage dropped further to 1.112v causing whea warnings to appear. Then the penny dropped that voltage decreases proportionately to load when not using any LLC. So I upped the voltage so that when running IBT / x264 the voltage did not get lower than 1.12v. No more whea warnings.
I think Prime95 should be used as one part of a stability testing strategy, not solely.

This is my point. IBT doesn't push the CPU as hard as Prime95 or something because it's not allowing the CPU to get the voltage it needs to run at that speed. It's almost like it's limiting the CPU, this is why I say you can't run IBT on 10 passes at standard settings and say it's stable. In my case, I can pass P95 running 4.5GHz @ 1.24v, but I can also pass IBT @ 1.224v. I would trust the one that pushes the CPU more, but yes I agree, it should be one of many tests ran.
Edited by Nihilo - 5/17/12 at 6:36am
Ex Nihilo
(12 items)
  		  
CPUMotherboardGraphicsRAM
I7-3770K @ 4.5GHz Asus P8Z77-V Pro GTX 560ti x 2 SLI G.Skill RipjawS X 16GB (2 x 8GB) 
Hard DriveHard DriveOptical DriveCooling
Seagate 2TB 6.0GB/s Crucial M4 128GB SSD LG Blu-Ray Burner Corsair H100 
OSPowerCase
Windows 7 Professional Corsair TW850W V2 Corsair 500r 
View all
  hide details  
Reply
Ex Nihilo
(12 items)
  		  
CPUMotherboardGraphicsRAM
I7-3770K @ 4.5GHz Asus P8Z77-V Pro GTX 560ti x 2 SLI G.Skill RipjawS X 16GB (2 x 8GB) 
Hard DriveHard DriveOptical DriveCooling
Seagate 2TB 6.0GB/s Crucial M4 128GB SSD LG Blu-Ray Burner Corsair H100 
OSPowerCase
Windows 7 Professional Corsair TW850W V2 Corsair 500r 
View all
  hide details  
Reply
post #455 of 3380
5/17/12 at 6:43am
Quote:
Originally Posted by Nihilo View Post

This is my point. IBT doesn't push the CPU as hard as Prime95 or something because it's now allowing the CPU to get the voltage it needs to run at that speed. It's almost like it's limiting the CPU, this is why I say you can't run IBT on 10 passes at standard settings and say it's stable.

Sorry, what I'm trying to say is, when no LLC is used voltage decreases as CPU load increases (vdroop in other words). When running IBT on my machine the voltage drooped to 1.112v, under Prime it only drooped to 1.12v. This tells me that when running IBT the load on the CPU is higher than when running Prime95, which would also explain why Prime95 runs a good 7-10 degrees cooler than IBT, despite apparently running at a higher voltage.
Edited by iSeries - 5/17/12 at 6:46am
Alpha II
(15 items)
  		  
CPUMotherboardRAMHard Drive
Intel 3770K Asus P8Z77-V Pro Corsair 8GB DDR3 1600MHz Vengeance Crucial M4 SSD 
Hard DriveOptical DriveCoolingOS
Samsung F3 LG GGW-H20L Thermalright Archon Windows 7 Ultimate 64-Bit 
PowerCase
Seasonic S12II-430W Coolermaster CM 690 ll 
View all
  hide details  
Reply
Alpha II
(15 items)
  		  
CPUMotherboardRAMHard Drive
Intel 3770K Asus P8Z77-V Pro Corsair 8GB DDR3 1600MHz Vengeance Crucial M4 SSD 
Hard DriveOptical DriveCoolingOS
Samsung F3 LG GGW-H20L Thermalright Archon Windows 7 Ultimate 64-Bit 
PowerCase
Seasonic S12II-430W Coolermaster CM 690 ll 
View all
  hide details  
Reply
post #456 of 3380
5/17/12 at 7:03am
Quote:
Originally Posted by iSeries View Post

Sorry, what I'm trying to say is, when no LLC is used voltage decreases as CPU load increases (vdroop in other words). When running IBT on my machine the voltage drooped to 1.112v, under Prime it only drooped to 1.12v. This tells me that when running IBT the load on the CPU is higher than when running Prime95, which would also explain why Prime95 runs a good 7-10 degrees cooler than IBT, despite apparently running at a higher voltage.

Actually, Prime pushes the CPU harder as it gets deeper into the tests. At one point, my temps were equal to what IBT would've put them at. It's just that P95 has more tests to go through and they get harder and harder on the CPU as it gets longer into the test.
Ex Nihilo
(12 items)
  		  
CPUMotherboardGraphicsRAM
I7-3770K @ 4.5GHz Asus P8Z77-V Pro GTX 560ti x 2 SLI G.Skill RipjawS X 16GB (2 x 8GB) 
Hard DriveHard DriveOptical DriveCooling
Seagate 2TB 6.0GB/s Crucial M4 128GB SSD LG Blu-Ray Burner Corsair H100 
OSPowerCase
Windows 7 Professional Corsair TW850W V2 Corsair 500r 
View all
  hide details  
Reply
Ex Nihilo
(12 items)
  		  
CPUMotherboardGraphicsRAM
I7-3770K @ 4.5GHz Asus P8Z77-V Pro GTX 560ti x 2 SLI G.Skill RipjawS X 16GB (2 x 8GB) 
Hard DriveHard DriveOptical DriveCooling
Seagate 2TB 6.0GB/s Crucial M4 128GB SSD LG Blu-Ray Burner Corsair H100 
OSPowerCase
Windows 7 Professional Corsair TW850W V2 Corsair 500r 
View all
  hide details  
Reply
post #457 of 3380
5/17/12 at 7:15am
I may have to disagree. Yes, at some points during the 22 hours I ran Prime temps approached what I was getting with IBT, but on average over the run temps were 7-10 degrees cooler. As my voltage when running Prime did not go as low as when running IBT (or even x264) this shows me that processor load is not as high (higher load=more vdroop),

I don't think running Prime95 for 22 hours proves anything at all, as evidenced by the whea errors in my event viewer when running an x264 encode due to not enough voltage. CPU errors during encoding=visual corruption of encode. I had considered my PC stable after the 22 hour Prime run, but this was proved not to be the case. Bumped up the voltage a notch, and no more whea errors.

Maybe x264 is the best test for stability smile.gif
Edited by iSeries - 5/17/12 at 7:21am
Alpha II
(15 items)
  		  
CPUMotherboardRAMHard Drive
Intel 3770K Asus P8Z77-V Pro Corsair 8GB DDR3 1600MHz Vengeance Crucial M4 SSD 
Hard DriveOptical DriveCoolingOS
Samsung F3 LG GGW-H20L Thermalright Archon Windows 7 Ultimate 64-Bit 
PowerCase
Seasonic S12II-430W Coolermaster CM 690 ll 
View all
  hide details  
Reply
Alpha II
(15 items)
  		  
CPUMotherboardRAMHard Drive
Intel 3770K Asus P8Z77-V Pro Corsair 8GB DDR3 1600MHz Vengeance Crucial M4 SSD 
Hard DriveOptical DriveCoolingOS
Samsung F3 LG GGW-H20L Thermalright Archon Windows 7 Ultimate 64-Bit 
PowerCase
Seasonic S12II-430W Coolermaster CM 690 ll 
View all
  hide details  
Reply
post #458 of 3380
5/17/12 at 7:22am
Quote:
Originally Posted by iSeries View Post

I may have to disagree. Yes, at some points during the 22 hours I ran Prime temps approached what I was getting with IBT, but on average over the run temps were 7-10 degrees cooler. As my voltage when running Prime did not go as low as when running IBT (or even x264) this shows me that processor load is not as high (higher load=more vdroop),
I don't think running Prime95 for 22 hours proves anything at all, as evidenced by the whea errors in my event viewer when running an x264 encode due to not enough voltage. CPU errors during encoding=visual corruption of encode. I had considered my PC stable after the 22 hour Prime run, but this was proved not to be the case. Bumped up the voltage a notch, and no more whea errors.

I didn't say Prime is the say all end all test. I agreed that it's just one of many. My whole point a few posts back was that people need to stop saying they're stable after 10 passes of standard IBT is all. That's what my pics were trying to prove. I even said to others in another thread that just because you pass Prime doesn't mean you're completely stable and that they need to check for event viewer errors (specifically WHEA 19). I, personally, run 20 passes at least of LinX on very high setting, Prime for 22 hours, then check event viewer. Any errors, I try to correct and run again. OCCT is good one too to push the GPUs. Also, in my pics, that was at LLC 50% so I shouldn't be getting that much VDroop at all especially not .020v worth.
Edited by Nihilo - 5/17/12 at 7:25am
Ex Nihilo
(12 items)
  		  
CPUMotherboardGraphicsRAM
I7-3770K @ 4.5GHz Asus P8Z77-V Pro GTX 560ti x 2 SLI G.Skill RipjawS X 16GB (2 x 8GB) 
Hard DriveHard DriveOptical DriveCooling
Seagate 2TB 6.0GB/s Crucial M4 128GB SSD LG Blu-Ray Burner Corsair H100 
OSPowerCase
Windows 7 Professional Corsair TW850W V2 Corsair 500r 
View all
  hide details  
Reply
Ex Nihilo
(12 items)
  		  
CPUMotherboardGraphicsRAM
I7-3770K @ 4.5GHz Asus P8Z77-V Pro GTX 560ti x 2 SLI G.Skill RipjawS X 16GB (2 x 8GB) 
Hard DriveHard DriveOptical DriveCooling
Seagate 2TB 6.0GB/s Crucial M4 128GB SSD LG Blu-Ray Burner Corsair H100 
OSPowerCase
Windows 7 Professional Corsair TW850W V2 Corsair 500r 
View all
  hide details  
Reply
post #459 of 3380
5/17/12 at 7:22am
  • Joined: Jan 2011
  • Posts: 3,047
  • Rep: 142 (Unique: 114)
  • Reviews: 1
  • Trader Rating: 4
  • Folding Team Rank: 3,140
  •  
    Name Team Rank Points / Jun Work Units / Jun
    OCNApparition 3,140 133,255 / 0 29 / 0
     
  • Select All Posts By This User
Any other settings people have found that make a difference when doing minor tweaking?
I have not had the time yet to do a long Prime run but for my 4.7 OC I am up to 1.288 V now from previous 1.280 and started raising PLL from 1.62 to 1.65.
I am experimenting with PLL differences, which I think is like micro tuning the cpu. Prime failed at 1.280 v after a few hours but I am wondering if I play with the PLL voltage enough if I could make it stable. Just need the time...
LLC is very high (%75) which gives me almost zero v-droop, +- 0.002.
Anyone change the CPU Frequency setting? Default is 300.
Rig 2.0
(16 items)
  		  
CPUMotherboardGraphicsRAM
i7 3770k Asus P8Z77-V EVGA GTX 470 SC (DEAD :( ) Samsung DDR 3 (2 DIMMS) MV-3V4G3D/US 
Hard DriveHard DriveOptical DriveCooling
Samsung 830 SSD WD Caviar Black Asus DVD/RW Xigmatek s-1283 HDT (Air Cooling) 
CoolingCoolingOSMonitor
Scythe Slipstream 1200RPM (x2) Antec 140mm + Antec 120mm + Xigmatek 120mm (x2) Win 7 64bit Acer S243HL bmii - 24" 
PowerCaseMouseAudio
Corsair TX750 Corsair Carbide 300R MX 518 Auzentek X-Fi Forte 7.1 
View all
  hide details  
Reply
Rig 2.0
(16 items)
  		  
CPUMotherboardGraphicsRAM
i7 3770k Asus P8Z77-V EVGA GTX 470 SC (DEAD :( ) Samsung DDR 3 (2 DIMMS) MV-3V4G3D/US 
Hard DriveHard DriveOptical DriveCooling
Samsung 830 SSD WD Caviar Black Asus DVD/RW Xigmatek s-1283 HDT (Air Cooling) 
CoolingCoolingOSMonitor
Scythe Slipstream 1200RPM (x2) Antec 140mm + Antec 120mm + Xigmatek 120mm (x2) Win 7 64bit Acer S243HL bmii - 24" 
PowerCaseMouseAudio
Corsair TX750 Corsair Carbide 300R MX 518 Auzentek X-Fi Forte 7.1 
View all
  hide details  
Reply
post #460 of 3380
5/17/12 at 7:24am
@ Nihilo and iSeries, are you guys talking about the same version of prime95? Cuz 27.7 gives much higher temps than 26.6. 26.6 is colder than IBT, but 27.7 is about the same.
My Rig
(14 items)
  		  
CPUMotherboardGraphicsRAM
Intel Core i7 3770k P8Z68-V GEN3 MSI GeForce GTX 670 Power Edition G. Skill Ripjaws X 
Hard DriveHard DriveOptical DriveCooling
Corsair Force Series 3 Seagate Barracuda ASUS DVD Burner Corsair H100 
OSMonitorMonitorPower
Microsoft Windows 7 Professional Samsung EX2220 21.5" 1920x1080 Acer 21.5" 1920x1080 Corsair HX750 
Case
NZXT Tempest 410 Elite 
View all
  hide details  
Reply
My Rig
(14 items)
  		  
CPUMotherboardGraphicsRAM
Intel Core i7 3770k P8Z68-V GEN3 MSI GeForce GTX 670 Power Edition G. Skill Ripjaws X 
Hard DriveHard DriveOptical DriveCooling
Corsair Force Series 3 Seagate Barracuda ASUS DVD Burner Corsair H100 
OSMonitorMonitorPower
Microsoft Windows 7 Professional Samsung EX2220 21.5" 1920x1080 Acer 21.5" 1920x1080 Corsair HX750 
Case
NZXT Tempest 410 Elite 
View all
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Intel CPUs
  • [Official] The Ivy Bridge Stable / Suicide Club **Guides, Voltages, Temps & BIOS Templates** Inc SPREADSHEET

Gear mentioned in this thread:

Overclock.net › Forums › Intel › Intel CPUs › [Official] The Ivy Bridge Stable / Suicide Club **Guides, Voltages, Temps & BIOS Templates** Inc SPREADSHEET