anything higher than 945Mhz and weird things happen in windows. i'm able to reach 1Ghz and get into windows and do a few things but when gaming it seems to lock. would i have better results if i bumped the voltage to 1.80/1.85 volts? i've never tried going past 1.75 volts as i'm not sure what the higher voltage could do to my CPU. my system specs are below.
is it risky if I bump my CPU voltage up from 1.75v to 1.80/1.85 volts???
- Thread starter Detoyminador_
- Start date
Raising the voltage wouldn't do much if you have efficient cooling. Granted that it will shorten your CPU's life a couple years, but having it last more than 5-8 years(instead of 10-15 on stock) years would definately be enough for any CPU before going obsolete. As long as you have an Alpha, golden Orb or any great cooler. Go for it if you need more stablity. Just don't go beyond the 1.95 volt barrier.
I am running the same exact CPU like yours ... P3 700e
here's my overclock and FSB result.
980 Mhz - 1.80 V
1030 Mhz - 1.85 V
1050 Mhz - 1.90 V
I am currently running at 1030. I have been running at 1050 for a while without a problem at all. I am using a GlobalWin heatsink. The CPU barely gets hot at all, even when it is under full load.
eRr
here's my overclock and FSB result.
980 Mhz - 1.80 V
1030 Mhz - 1.85 V
1050 Mhz - 1.90 V
I am currently running at 1030. I have been running at 1050 for a while without a problem at all. I am using a GlobalWin heatsink. The CPU barely gets hot at all, even when it is under full load.
eRr
I've a question,
If i run my 700Mhz Duron at 900Mhz and 1.55V which is default, would it shorten my CPU life?
If i run my 700Mhz Duron at 900Mhz and 1.55V which is default, would it shorten my CPU life?
If I am not mistaken, even though you are running it at the default 1.55v, you will still shorten the life expectancy of the processor by overclocking through electro-migration. It occurs when an electronic component is operated over it's maximum ratings regardless of temperature or voltage.
Hope I'm not being a complete ass here, but you are wrong,Technonut....well, at least not 100% correct.
Let me explain:
Electromigration is an effect that occurs when an extremely dense electron flow knocks off atoms within the wire and moves them, leaving a gap at one end and high stress at the other. In a chip, the formation of such a void will cause an open circuit and result in a failure. At the other end, the increase of stresses can cause fracture of the insulator around the wire and shorting
What this amounts to is the fact that when an electrical current passes through a conductor some of the metal atoms is swept along with the flow of electrons.
There have been a lot of scientific papers published on the effects of electromigration, although nearly all of them focus on forced electromigration for chip test purposes. There is almost no literature or documentation on the effects of electromigration neither during normal chip use nor overclocked chips, but when reading about the forced electromigration one can clearly see the similarities with overclocked chips: The higher temperature, the higher possibility for electromigration to occur.
The maximum operating frequency is proportional to (Vth-V) 1,25/V, where we assume Vth is 0,6V. Between 1V and 3V, the operating frequency is approximately proportional to the supply voltage, meaning that if you have a CPU that does 850MHz at 1,5V you will most likely make it run at about 1,0GHz to 1,13GHz when you increase its core voltage to 2,0V.
Increasing the core voltage automatically means higher wattage output of the chip; doubling the voltage means doubling the frequency, but it also increase the total wattage output by about 800%. If a CPU that originally emit 25W it will at double voltage and speed now radiate 200W of heat!
As most CPU designs nowadays are moving over to the copper interconnect technology because of speed and price considerations, this change over also has a hidden bonus. Research has shown that the dual Damascene Cu has a much higher resistance to electromigration than the previously utilized aluminum interconnect-technology.
The higher the temperature and voltage within the conductor, the faster the metal atoms will move, and the faster the chip will fail due to electromigration. There is not much we can do about this, as there is really only one factor we can change ? the temperature.
Now, over to the original question:
Yes, an increase of voltage will help, and yes, better cooling will both prevent electromigration, increase stability of your chip and most likely let you run your chip at a higher speed than earlier. I would say 1,8-1,9V is safe with your cooling, but you might consider getting a TEC for it.
Hope this helps!
Duckman
Let me explain:
Electromigration is an effect that occurs when an extremely dense electron flow knocks off atoms within the wire and moves them, leaving a gap at one end and high stress at the other. In a chip, the formation of such a void will cause an open circuit and result in a failure. At the other end, the increase of stresses can cause fracture of the insulator around the wire and shorting
What this amounts to is the fact that when an electrical current passes through a conductor some of the metal atoms is swept along with the flow of electrons.
There have been a lot of scientific papers published on the effects of electromigration, although nearly all of them focus on forced electromigration for chip test purposes. There is almost no literature or documentation on the effects of electromigration neither during normal chip use nor overclocked chips, but when reading about the forced electromigration one can clearly see the similarities with overclocked chips: The higher temperature, the higher possibility for electromigration to occur.
The maximum operating frequency is proportional to (Vth-V) 1,25/V, where we assume Vth is 0,6V. Between 1V and 3V, the operating frequency is approximately proportional to the supply voltage, meaning that if you have a CPU that does 850MHz at 1,5V you will most likely make it run at about 1,0GHz to 1,13GHz when you increase its core voltage to 2,0V.
Increasing the core voltage automatically means higher wattage output of the chip; doubling the voltage means doubling the frequency, but it also increase the total wattage output by about 800%. If a CPU that originally emit 25W it will at double voltage and speed now radiate 200W of heat!
As most CPU designs nowadays are moving over to the copper interconnect technology because of speed and price considerations, this change over also has a hidden bonus. Research has shown that the dual Damascene Cu has a much higher resistance to electromigration than the previously utilized aluminum interconnect-technology.
The higher the temperature and voltage within the conductor, the faster the metal atoms will move, and the faster the chip will fail due to electromigration. There is not much we can do about this, as there is really only one factor we can change ? the temperature.
Now, over to the original question:
Yes, an increase of voltage will help, and yes, better cooling will both prevent electromigration, increase stability of your chip and most likely let you run your chip at a higher speed than earlier. I would say 1,8-1,9V is safe with your cooling, but you might consider getting a TEC for it.
Hope this helps!
Duckman
Wow! Detailed explanation!
Duckie, i seem to have met you before. Did you ever frequent AGN forums?
Duckie, i seem to have met you before. Did you ever frequent AGN forums?
Well, I've been there, but under my real nickname "Duckman". It's a long time since last time though. I mostly hang out at HardWare Central where I'm...*cough*...admin in the forums...please don't tell Sander, he'll kill me
If you wanna know the whole story about electromigration, then have a look at the HWC mainpage in a few days...oh, goodness me...here I go spamming..shame on me...
Won't happen again
Duckman
If you wanna know the whole story about electromigration, then have a look at the HWC mainpage in a few days...oh, goodness me...here I go spamming..shame on me...
Won't happen again
Duckman
I did say "If I'm not mistaken" The way I understood the process, the processor could still be well under it's maximum operating temperature, and still suffer the effects of electro-migration. This may not be the case with copper interconnects, but what of aluminum? You stated that the copper has much higher resistance against electro migration compared to aluminum, but the aluminum is still subject to the process nonetheless. I have to do some digging, but I remember that research from IBM revealed that temperature was not a factor. Again, I may be mistaken, and stand corrected if I am.
EDIT: Seems like a good reason to upgrade to a T-Bird with the copper interconnects. Now, just have to to convince my Wife to alter the budget....
The way I understood the process, the processor could still be well under it's maximum operating temperature, and still suffer the effects of electro-migration. This may not be the case with copper interconnects, but what of aluminum? You stated that the copper has much higher resistance against electro migration compared to aluminum, but the aluminum is still subject to the process nonetheless. I have to do some digging, but I remember that research from IBM revealed that temperature was not a factor. Again, I may be mistaken, and stand corrected if I am.EDIT: Seems like a good reason to upgrade to a T-Bird with the copper interconnects. Now, just have to to convince my Wife to alter the budget....
Well, having spent the last few weeks reading about this subject, I know some of it.
Electromigration will occur on a chip, sooner or later. How fast this will happen is determined by several factors:
-The interconnect material ie aluminum, copper or other experimental alloys
-The operating voltage
-The operating temperature
-The size of the core
Now, the voltage, temperature, core width and frequency are directly related to eachother, and this effect will increase as the cores shrink.
IBM reports clearly shows that aluminum is much more suceptible than copper, and this is one of the big advantages to Cu CPUs. As I said, there are almost no literature or documentation on the effects of electromigration neither during normal chip use nor overclocked chips, or at least not available to me and thus the severity of this problem/how much this will shorten a chips lifetime is not known.
Edited to say: I can't say more about this until the article is uploaded at HWC......sorry.
Duckman
Electromigration will occur on a chip, sooner or later. How fast this will happen is determined by several factors:
-The interconnect material ie aluminum, copper or other experimental alloys
-The operating voltage
-The operating temperature
-The size of the core
Now, the voltage, temperature, core width and frequency are directly related to eachother, and this effect will increase as the cores shrink.
IBM reports clearly shows that aluminum is much more suceptible than copper, and this is one of the big advantages to Cu CPUs. As I said, there are almost no literature or documentation on the effects of electromigration neither during normal chip use nor overclocked chips, or at least not available to me and thus the severity of this problem/how much this will shorten a chips lifetime is not known.
Edited to say: I can't say more about this until the article is uploaded at HWC......sorry.
Duckman
um....to sum all that up, heat is the real enemy (at least the one you can control), so just keep it cool and that voltage is fine.
Exactly!
(dam, here I go, typing 'till I get blisters on my fingers, and you use one single line to say what I didn on a whole page)
Duckman
(dam, here I go, typing 'till I get blisters on my fingers, and you use one single line to say what I didn on a whole page
)Duckman
You guys running your P3 CuMine at higher voltages... what CPU core temperatures do you get when idling, and when the CPU is at full load (RC5, SETI, etc.)?
Mean time to failure for EM increases with voltage since the current density rises. Improved cooling will help with this, but will not fix the problem since the fact that current has increased can't be changed. In any case, nowadays aluminum interconnects aren't really even aluminum any more and their susceptibility to EM has dropped to the point where the failure mechanism on silicon usually isn't EM. Compared to NFET hot-electron gate ionization and PFET Bias Temperature Instability the failure mechanisms of which increase squared to increased voltage, EM is not the major issue any more.
But I guess the point that I really disagree is the conclusion that improving cooling will completely erradicate the problems associated with increasing voltage. For EM, increased current density means a bigger problem for a given constant T and increasing voltage leads to higher current. For Hot-E, voltage is the main culprit. For PMOS BTI, voltage is the main culprit. Cooling will help, but it doesn't eliminate the problem. There's a reason why Intel (and AMD, and all other manufacturers) don't run their components at 2V and the reason is that the reliability falls off sharply with increased voltage. As a process matures usually the voltage can be increased slightly and manufacturers take advantage of this, but this is a small effect and it's gradual.
Engineers at semiconductor companies are well aware of the fact that if you increase the voltage of a processor, the transistors switch faster and signal to noise ratios improve, and so the voltage of a chip is set to as high a value as possible without compromising reliability. This is a very competitive business and if it were possible to improve bin split (frequency) by increasing the voltage spec, then companies would take advantage of it, but Intel (and AMD, and everyone else) set the spec to as high a value as they can without unacceptable reliability issues.
But I guess the point that I really disagree is the conclusion that improving cooling will completely erradicate the problems associated with increasing voltage. For EM, increased current density means a bigger problem for a given constant T and increasing voltage leads to higher current. For Hot-E, voltage is the main culprit. For PMOS BTI, voltage is the main culprit. Cooling will help, but it doesn't eliminate the problem. There's a reason why Intel (and AMD, and all other manufacturers) don't run their components at 2V and the reason is that the reliability falls off sharply with increased voltage. As a process matures usually the voltage can be increased slightly and manufacturers take advantage of this, but this is a small effect and it's gradual.
Engineers at semiconductor companies are well aware of the fact that if you increase the voltage of a processor, the transistors switch faster and signal to noise ratios improve, and so the voltage of a chip is set to as high a value as possible without compromising reliability. This is a very competitive business and if it were possible to improve bin split (frequency) by increasing the voltage spec, then companies would take advantage of it, but Intel (and AMD, and everyone else) set the spec to as high a value as they can without unacceptable reliability issues.
i've bumped it up to 980Mhz 1.80 volts and so far so good, not 1 lock-up at all. i've used quake3 timedemo for a number of times, 3dmark2000 looped, grandprix 3 30+min, quake2 looping 30+min and not 1 lock-up. my temps are at idle CPU=32c SYSTEM=27c, at full load my CPU=42-43c, i think these are fairly good temps. now if i have no dramas with this speed for a while, i'll be going for the 1Ghz mark, the highest i'll take the voltage to is 1.85v though.
Eug
Lifer
- Mar 11, 2000
- 24,165
- 1,809
- 126
With your PEP66 I wouldn't worry about 1.85. I understand the wisdom of the experts above, but I figure you're the type who isn't going to keep that chip for more than a couple of years, and just the fact that you bought a PEP66 partially indicates that you understand the risks of o'clocking. I note also that the PIII 1.13 chips shipped at 1.8 V as well. (But then again, they were recalled. )
Not really long enough to say, but I had been running 1.9 V for several months with a PEP66, before drive problems (PCI bus 38.3) started to spook me. (Now I'm running a more reasonable FSB to appease my drives, and consequently only need 1.8 V.)
)Not really long enough to say, but I had been running 1.9 V for several months with a PEP66, before drive problems (PCI bus 38.3) started to spook me. (Now I'm running a more reasonable FSB to appease my drives, and consequently only need 1.8 V.)
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes + WCL Discussion Threads
- Started by Tigerick
- Replies: 23K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
-
Facebook
Twitter