Overclock.net › Forums › Intel › Intel CPUs › Haswell Overclocking Guide [With Statistics]
New Posts  All Forums:Forum Nav:

Haswell Overclocking Guide [With Statistics] - Page 1132

post #11311 of 19540
Quote:
I would say as long as the chip stays under 75-80c max on all cores and averages lower than 1.5-1.54v should be fine

TBH i doubt that would last two years - we're kinda speculating though. Not too many voltage dead chips.. 1.45 (which often requires 1.43 set in bios) is a safer bet
Insert Name Here
(14 items)
 
  
CPUMotherboardGraphicsRAM
6700k Asus Maximus VIII Hero Gigabyte Aorus Xtreme 1080ti Corsair LPX 2x8GB 3200c16 
Hard DriveHard DriveCoolingOS
Old Seagate HDD Samsung 850 EVO Thermalright Silver Arrow SB-E SE Windows 7 Home Premium 64 bit 
MonitorKeyboardPowerCase
Asus PG258Q (240hz + Gsync) WASDKeyboards.com v1 semi custom w/ mx browns, ... Superflower Golden Green HX550 Air540 
MouseMouse Pad
Logitech G Pro Qck+ 
  hide details  
Reply
Insert Name Here
(14 items)
 
  
CPUMotherboardGraphicsRAM
6700k Asus Maximus VIII Hero Gigabyte Aorus Xtreme 1080ti Corsair LPX 2x8GB 3200c16 
Hard DriveHard DriveCoolingOS
Old Seagate HDD Samsung 850 EVO Thermalright Silver Arrow SB-E SE Windows 7 Home Premium 64 bit 
MonitorKeyboardPowerCase
Asus PG258Q (240hz + Gsync) WASDKeyboards.com v1 semi custom w/ mx browns, ... Superflower Golden Green HX550 Air540 
MouseMouse Pad
Logitech G Pro Qck+ 
  hide details  
Reply
post #11312 of 19540
Quote:
Originally Posted by FractinJex View Post

...I usually don't like to talk about those kinds of voltages esp on this thread where folks are learning to overclock on haswell etc so I don't want them to stumble on this post and think its ok to run anywhere near 1.6 24/7 so plz don't!!!!! tongue.gif

Also not all boards should be pushed high on voltages...I hear people say all the time that mobos above 200$ range or whatever are a rip and no different than a lower mid range board capable of overclock...while this is true for mid range overclocks once you start pushing high voltage the chip is not the only hardware handling the voltages lol...high end boards have different phases/more and most of the time better capacitors and vrms allowing for cleaner more efficient voltages to ran,
Quote:
Originally Posted by FractinJex View Post

yeah with that loop and delid you can def. push the chip hard I would say as long as the chip stays under 75-80c max on all cores and averages lower than 1.5-1.54v should be fine with the possibility of having to run vrin around 2.1-2.2v

Id say she might reaquire a voltage bump after aboiut 2 years or so but even then it should last until you upgrade or whatever. theres a couple of guys running ivy/haswells around 1.5-1.54 ish that ive seen I don't recall their usernames...but I have a 4.9ghz 4770k in my buds rig that was mine and beat to hell that's running 1.51v and hasd been for about 5 months + tongue.gif

You def didn't spend coin on a nice water loop and delid your chip to take it easy on her devil.gif
btw grab the tuning plan if you haven't already for fracking 25$ they allow us t beat them up and appranetly they are accepting delided chips as well but not lapped.



It’s actually the voltage that shorten’s the life of the CPU.
In semiconductors the effect is called Electromigration, where electrons could jump (interconnect) through an insulator from one internal “trace” to another, thus creating an electrical bridge between two points that were supposed to be isolated from one another.
And with such a small lithography, the interconnects are even more probable. This is what all that “silicon lottery” is all about.

The resistance in the electrical flow is what creates the heat, and more heat means more resistance. That electrical flow is increased with frequency, so technically, you can push 1.5V through the chip for 20* years without problems..., as long as the frequency (load) stays low.

In other words, 1.5V (or whatever) at idle or at medium load is not a problem, but it becomes a problem at lull load/max frequency.
This is why prime/linpacks create so much more heat than x264, games, etc. even though the voltage and the frequency is the same, because there is more load (higher electrical flow).

But heat accelerates electromigration aswell. So it’s a circular problem.
Edited by angelotti - 3/20/14 at 2:37pm
post #11313 of 19540
So staying under 1.45V on the core even if pushing say 4.8ghz is probably safe for a few years?
760T
(13 items)
 
LGA 775
(13 items)
 
 
CPUMotherboardGraphicsRAM
i7-4770K Gigabyte GA-Z97X-Gaming G1 EVGA GTX 1070 FTW G.Skill (2x8GB) 
Hard DriveHard DriveHard DriveCooling
Samsung EVO Crucial C300 WD Caviar Black Corsair H110 
OSKeyboardPowerCase
Windows 7 HP Tt esports Poseidon Cherry MX Brown XFX XTR 550W Corsair 760T 
Mouse
Deathadder Chroma 
CPUMotherboardGraphicsRAM
E6850 ASUS P5N-E SLI eVga GTX570 2x2GB Crucial Ballistix 800MHz 
Hard DriveOSMonitorKeyboard
Crucial C300 128GB + WD Caviar Black 1TB Win 7 Home 64-bit Samsung SyncMaster 225BW Logitech 
PowerCaseMouseMouse Pad
PC Power and Cooling Silencer 610W Cooler Master elite Logitech G5 very large 
  hide details  
Reply
760T
(13 items)
 
LGA 775
(13 items)
 
 
CPUMotherboardGraphicsRAM
i7-4770K Gigabyte GA-Z97X-Gaming G1 EVGA GTX 1070 FTW G.Skill (2x8GB) 
Hard DriveHard DriveHard DriveCooling
Samsung EVO Crucial C300 WD Caviar Black Corsair H110 
OSKeyboardPowerCase
Windows 7 HP Tt esports Poseidon Cherry MX Brown XFX XTR 550W Corsair 760T 
Mouse
Deathadder Chroma 
CPUMotherboardGraphicsRAM
E6850 ASUS P5N-E SLI eVga GTX570 2x2GB Crucial Ballistix 800MHz 
Hard DriveOSMonitorKeyboard
Crucial C300 128GB + WD Caviar Black 1TB Win 7 Home 64-bit Samsung SyncMaster 225BW Logitech 
PowerCaseMouseMouse Pad
PC Power and Cooling Silencer 610W Cooler Master elite Logitech G5 very large 
  hide details  
Reply
post #11314 of 19540
Quote:
Originally Posted by angelotti View Post


It’s actually the voltage that shorten’s the life of the CPU.
In semiconductors the effect is called Electromigration, where electrons could jump (interconnect) through an isolator from one internal “trace” to another, thus creating an electrical bridge between two points that were supposed to be isolated from one another.
And with such a small lithography, the interconnects are even more probable. This is what all that “silicon lottery” is all about.

The resistance in the electrical flow is what creates the heat, and more heat means more resistance. That electrical flow is increased with frequency, so technically, you can push 1.5V through the chip for 20* years without problems..., as long as the frequency (load) stays low.

In other words, 1.5V (or whatever) at idle or at medium load is not a problem, but it becomes a problem at lull load/max frequency.
This is why prime/linpacks create so much more heat than x264, games, etc. even though the voltage and the frequency is the same, because there is more load (higher electrical flow).

But heat accelerates electromigration aswell. So it’s a circular problem.

ADD

tongue.gif
Edited by FractinJex - 3/20/14 at 11:05am
Haswell Rig
(16 items)
 
  
CPUMotherboardGraphicsRAM
i7 4770k 4.9 MSI Mpower z87 Nvidia GTX 770 -SLI- Gskill 8GB 2400Mhz 
Hard DriveHard DriveOptical DriveCooling
2x Samsung 840 Pro SSD's 2x Seagate 2TB LG Blue Ray Custom loop 360/240 Rad EK Block 
OSMonitorKeyboardPower
Win 8.1 64b Qnix 2560x1440 96hz Corsair k70 Corsair AX860 
CaseMouseMouse PadAudio
Corsair Air 540 Razer Deathadder Razer Ironclad Creative X-FI - Logitech THX 2.1 
  hide details  
Reply
Haswell Rig
(16 items)
 
  
CPUMotherboardGraphicsRAM
i7 4770k 4.9 MSI Mpower z87 Nvidia GTX 770 -SLI- Gskill 8GB 2400Mhz 
Hard DriveHard DriveOptical DriveCooling
2x Samsung 840 Pro SSD's 2x Seagate 2TB LG Blue Ray Custom loop 360/240 Rad EK Block 
OSMonitorKeyboardPower
Win 8.1 64b Qnix 2560x1440 96hz Corsair k70 Corsair AX860 
CaseMouseMouse PadAudio
Corsair Air 540 Razer Deathadder Razer Ironclad Creative X-FI - Logitech THX 2.1 
  hide details  
Reply
post #11315 of 19540
Quote:
Originally Posted by MaKe OuT View Post

So staying under 1.45V on the core even if pushing say 4.8ghz is probably safe for a few years?

That 1.5V was just a random number, for example purpose.
If you “won” the “silicon lottery” (which is a awful term, since is sorta loosely used and only half true), then the isolators will work as intended even at higher loads/frequencies.
The bsods happen when there is an interconnect wich result in random errors.
But even so, for that to be true, you will have to stay at medium-high loads or under (not full) for most of the time. If you benchmark/fold all day, degradation will occur.
Degradation happens on all chips at all voltages (incl stock) only much much slower.

Quote:
Originally Posted by FractinJex View Post

.......not sure why you copy and pasted that to me lol...im well aware of voltage effecting lifespan lol ive been overcloking and doing ln2/dice for years also biggrin.gif...

You got it wrong, i was agreeing with you, especially on the fact that the quality of other components (MB) is as important as the cpu.
The reason i linked to you, was to let other see that my post was a further discussion of yours (a bit more explanation)
post #11316 of 19540
lol my bad bro I reread im sorry im hopped up on coffee/add meds and a green leafy med lolssss also my coworker has irritated me today!!
Haswell Rig
(16 items)
 
  
CPUMotherboardGraphicsRAM
i7 4770k 4.9 MSI Mpower z87 Nvidia GTX 770 -SLI- Gskill 8GB 2400Mhz 
Hard DriveHard DriveOptical DriveCooling
2x Samsung 840 Pro SSD's 2x Seagate 2TB LG Blue Ray Custom loop 360/240 Rad EK Block 
OSMonitorKeyboardPower
Win 8.1 64b Qnix 2560x1440 96hz Corsair k70 Corsair AX860 
CaseMouseMouse PadAudio
Corsair Air 540 Razer Deathadder Razer Ironclad Creative X-FI - Logitech THX 2.1 
  hide details  
Reply
Haswell Rig
(16 items)
 
  
CPUMotherboardGraphicsRAM
i7 4770k 4.9 MSI Mpower z87 Nvidia GTX 770 -SLI- Gskill 8GB 2400Mhz 
Hard DriveHard DriveOptical DriveCooling
2x Samsung 840 Pro SSD's 2x Seagate 2TB LG Blue Ray Custom loop 360/240 Rad EK Block 
OSMonitorKeyboardPower
Win 8.1 64b Qnix 2560x1440 96hz Corsair k70 Corsair AX860 
CaseMouseMouse PadAudio
Corsair Air 540 Razer Deathadder Razer Ironclad Creative X-FI - Logitech THX 2.1 
  hide details  
Reply
post #11317 of 19540
Punch him!

...just kidding, keep your cool stay above it.
post #11318 of 19540
I was looking through the chart on the first page and thought it might be interesting to see the data a little more visually:

For the sake of simplicity I have this narrowed to 4770K only.

post #11319 of 19540
Quote:
Originally Posted by angelotti View Post


It’s actually the voltage that shorten’s the life of the CPU.
In semiconductors the effect is called Electromigration, where electrons could jump (interconnect) through an isolator from one internal “trace” to another, thus creating an electrical bridge between two points that were supposed to be isolated from one another.
And with such a small lithography, the interconnects are even more probable. This is what all that “silicon lottery” is all about.

The resistance in the electrical flow is what creates the heat, and more heat means more resistance. That electrical flow is increased with frequency, so technically, you can push 1.5V through the chip for 20* years without problems..., as long as the frequency (load) stays low.

In other words, 1.5V (or whatever) at idle or at medium load is not a problem, but it becomes a problem at lull load/max frequency.
This is why prime/linpacks create so much more heat than x264, games, etc. even though the voltage and the frequency is the same, because there is more load (higher electrical flow).

But heat accelerates electromigration aswell. So it’s a circular problem.
Quote:
Originally Posted by angelotti View Post


It’s actually the voltage that shorten’s the life of the CPU.
In semiconductors the effect is called Electromigration, where electrons could jump (interconnect) through an isolator from one internal “trace” to another, thus creating an electrical bridge between two points that were supposed to be isolated from one another.
And with such a small lithography, the interconnects are even more probable. This is what all that “silicon lottery” is all about.

The resistance in the electrical flow is what creates the heat, and more heat means more resistance. That electrical flow is increased with frequency, so technically, you can push 1.5V through the chip for 20* years without problems..., as long as the frequency (load) stays low.

In other words, 1.5V (or whatever) at idle or at medium load is not a problem, but it becomes a problem at lull load/max frequency.
This is why prime/linpacks create so much more heat than x264, games, etc. even though the voltage and the frequency is the same, because there is more load (higher electrical flow).

But heat accelerates electromigration aswell. So it’s a circular problem.

That is not quite how electomigration works. The electrons in the current flowing through a conductor lattice on the chip can occasionally gain enough energy to knock a metal lattice atom loose. The cumulative effect is the erosion of the conductor trace in the chip. The erosion can result in the conductor no longer conducting, and/or the eroded metal can accumulate such to cause a short to a nearby circuit element.

Electromigration depends quadratically on the current density (voltage) and exponentially on the temperature, so high temperature can accelerate electromigration significantly. Heat kills just as much.

http://en.wikipedia.org/wiki/Black's_equation

Electormigration occurs at any temperature, voltage and frequency. Operating the chip within the design parameters, you shouldn't see degradation due to electomigration for some time (but in principle it could be as short as the warranty period biggrin.gif ).

As you can imagine, the smaller the chip die process, the smaller the conductive traces, the more electormigration is an issue.

Voltage can kill a chip fast, but I think it is through processes other than electromigration.

-
Edited by GeneO - 3/20/14 at 11:42am
post #11320 of 19540
Quote:
Originally Posted by MNiceGuy View Post

I was looking through the chart on the first page and thought it might be interesting to see the data a little more visually:

For the sake of simplicity I have this narrowed to 4770K only.


it's interesting how the median VID actually drops from 44x to 45x. It looks like a concentration of chips at 44x that just couldn't make it to 45x, and ALSO need more voltage than average to maintain 44x.

The CPU doing 48x below 1.25v...is that JJ haha...Want... biggrin.gif
Edited by coelacanth - 3/20/14 at 11:33am
My System
(20 items)
 
  
CPUMotherboardGraphicsRAM
Intel i7-4770K @ 4.5GHz ASUS Maximus VI Hero Z87 Gigabyte GTX 980 Ti G1 Gaming 16GB (2 x 8GB) G.Skill Trident X DDR3-2400 
Hard DriveHard DriveHard DriveHard Drive
750GB Samsung 840EVO 256GB Crucial M4 150GB Velociraptor 1TB Samsung F1 
Optical DriveCoolingOSMonitor
Asus DRW-24B1ST Phanteks PH-TC14PE Win 10 64-bit 27" Qnix QX2710 96Hz 
KeyboardPowerCaseMouse
Cooler Master QuickFire Ultimate Cooler Master V1000 1000W 80+ Gold Fractal Design Arc Midi R2 Logitech G500s 
Mouse PadAudioOtherOther
Func Surface 1030 Asus Xonar Essence STX M-Audia AV 40 Polk PSW10 Sub 
  hide details  
Reply
My System
(20 items)
 
  
CPUMotherboardGraphicsRAM
Intel i7-4770K @ 4.5GHz ASUS Maximus VI Hero Z87 Gigabyte GTX 980 Ti G1 Gaming 16GB (2 x 8GB) G.Skill Trident X DDR3-2400 
Hard DriveHard DriveHard DriveHard Drive
750GB Samsung 840EVO 256GB Crucial M4 150GB Velociraptor 1TB Samsung F1 
Optical DriveCoolingOSMonitor
Asus DRW-24B1ST Phanteks PH-TC14PE Win 10 64-bit 27" Qnix QX2710 96Hz 
KeyboardPowerCaseMouse
Cooler Master QuickFire Ultimate Cooler Master V1000 1000W 80+ Gold Fractal Design Arc Midi R2 Logitech G500s 
Mouse PadAudioOtherOther
Func Surface 1030 Asus Xonar Essence STX M-Audia AV 40 Polk PSW10 Sub 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Intel CPUs
Overclock.net › Forums › Intel › Intel CPUs › Haswell Overclocking Guide [With Statistics]