2500k overclock to 4.9 question

[snuffleslol]

Hello, i have overclocked my 2500k to 4.9ghz with
a voltage level of 1.35, everytime i go to try and stress test i get blue screened and my computer shuts off, am i needing to put more voltage through my cpu or less?
im getting idle temps of about 25c, and while prime 95 is testing i get about 70c but it wont let me run it for long.... so answers?

 

[dudewitbow]

im going to assume you are at the point where you cannot overclock any further, as it is unstable to do so. although some people can probably hit 5.0, you have to remember the fact that not all cpus are built exactly the same.

 

[Pinhedd]

Idle temps of 25 at 4.9Ghz? Are you playing from an arctic research outpost or something?

 

[willard]

You need more voltage, which is almost always the case when you bluescreen. 1.35v at 4.9 GHz is really low. You can go all the way up to 1.4v for 24/7 with no worries, and 1.45 if you want to throw caution to the wind. Higher still you're risking damaging your chip.

70c at 1.35v with a short Prime95 run means you're probably already approaching, or at, your thermal limit. I'd strongly recommend not using an OC for 24/7 if it get into the 80s, and Prime95 will cause your temps to increase gradually over about 20-30 minutes. If you just got a couple minutes in and crashed, those are not your max temps.

Increasing your voltage is just going to make things worse, as voltage is by far the largest contributor to heat. Increasing frequency by itself has a fairly small impact on heat produced. I'm guessing the voltage required to get your chip stable at 4.9 GHz with your cooler is going to put you firmly into the 80s.

 

[Guest]

Hello, i have overclocked my 2500k to 4.9ghz with
a voltage level of 1.35, everytime i go to try and stress test i get blue screened and my computer shuts off, am i needing to put more voltage through my cpu or less?
im getting idle temps of about 25c, and while prime 95 is testing i get about 70c but it wont let me run it for long.... so answers?

Idle temps of 25 at 4.9Ghz? Are you playing from an arctic research outpost or something?

personally i think that deserves best answer (ok, its a question)

so whats going on with your Load Line Calibration?

excuse me for being ignorant since i do not have a Z77/68 board. but if i understand it correctly, having it to high or low could either spend a voltage spike or not enough voltage to your cpu when first stressing it.

and if it turns out to be ok, then a little more juice while keeping the temps down might be what the doctor ordered. from what i understand up to 1.4 volts is acceptable.

 

[oxford373]

you definitely need more voltage
1.4 v is the best voltage for 4.9GHZ, or try 1.35v with 4.5 GHZ which its better for 24/7
i had this issue before, every time i tried to make overclocking with low voltage i got blue screen with some games.

 

[willard]

1.4 v is the best voltage for 4.9GHZ

You can't possibly know that. Every chip is different, and will need a different voltage to hit certain overclocks, not to mention a lot of chips can't get stable at 4.9 GHz anyway.

The best voltage for any clock is the lowest voltage at which it is stable. That might be 1.4v, 1.45v, or 1.375v, or the voltage your chip needs for 4.9 GHz may be unsafe. We don't know, because your chip is different from every other chip. The only way to find the right voltage is to thoroughly test it yourself.

 

[bigcyco1]

You can't possibly know that. Every chip is different, and will need a different voltage to hit certain overclocks, not to mention a lot of chips can't get stable at 4.9 GHz anyway.

The best voltage for any clock is the lowest voltage at which it is stable. That might be 1.4v, 1.45v, or 1.375v, or the voltage your chip needs for 4.9 GHz may be unsafe. We don't know, because your chip is different from every other chip. The only way to find the right voltage is to thoroughly test it yourself.

+1

 

[snuffleslol]

oh yeh sorry the average temps idle were 45 my bad... i have actually changed it down to 4.8 because when i increased the voltage i started getting like 85*c running prime95 which i wasnt to happy about.... now, i have another question as im new to all of this, is there anything else apart from the bus or the multiplier and voltage i should worry about to get the most out of my overclock if so what? also does what is the difference between increasing the bus or multiplier? is there a certain way to know what ratio will make the higher overclock with the lower temps? any answers would be great im just trying to understand all this

 

[teh_gerbil]

What are you over clocking with as your cooler?

 

[snuffleslol]

hyper 212

 

[Deleted member 217926]

A Hyper 212 is not sufficient for overclocking past about 4.5Ghz. Your max temps when running Prime 95 should be in the low 70s C. Also 1.4v is cutting it close to being dangerous as far as reducing the life of your CPU. Unless you have a very good motherboard with at least 12 phase VRM you do not want to be at 1.4v. You really want to be more at 1.35v or less.

4.5Ghz is the sweet spot for these chips anyway and you will see little to no real world gains from going over that unless you are running 3 or more high end graphics cards.

Take some time to educate yourself about overclocking. It's trial and error and no 2 chips are ever the same. What is the same is the safe voltages and temperature. Go over at your own risk. You really need a custom water loop for the speeds you are trying to achieve though.

 

[bigcyco1]

hyper 212

get a better cooler for one :lol: those are o.k. for mild OC 4.5GHZ though some do it higher it's really not smart without a custom loop even these would be better than your current heatsink nothing else is up to par with these listed below as far as heatsinks go CPU Cooler:THERMALRIGHT SILVER ARROW DUAL140CM ROUND FAN HEATSINK
http://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=0CF4QFjAA&url=http%3A%2F%2Fwww.amazon.com%2FTHERMALRIGHT-SILVER-ARROW-DUAL14CM-HEATSINK%2Fdp%2FB003XRAKLQ&ei=_JPhT5fBNOaw2QW7pdjmCw&usg=AFQjCNHI-MZYKU2ahV-tLWOMvRskh3fg6Q&sig2=B0mUpnS6kbDfAS79npKCCQ

CPU Cooler:Noctua 6 Dual Heatpipe with 140mm/120mm Dual SSO Bearing Fans CPU Cooler NH-D14 - Retail
http://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=0CE4QFjAA&url=http%3A%2F%2Fwww.tigerdirect.com%2Fapplications%2FSearchTools%2Fitem-details.asp%3FEdpNo%3D5592441%26CatId%3D798&ei=cgrsT5HLF4bq2AW6iv24AQ&usg=AFQjCNHKBvrH9_b_UrPsdhyQG0Y0ZObpEg&sig2=Jh_62P5cvlu3g4Ue08VQ0g


PHANTEKS PH-TC14PE_BL 5 x?8mm Dual Heat-Pipes Dual 140mm Premium Fans and Quiet CPU Cooler with Patented P.A.T.S Coating
by Phanteks
http://www.amazon.com/PHANTEKS-PH-TC14PE_BL-Heat-Pipes-Premium-Patented/dp/B005OQJVP0/ref=pd_cp_pc_1/178-7594657-4819959

 

[amuffin]

For anything above 1.35V a custom watercooling loop is recommended.

 

[bigcyco1]

I agree with amuffin i use XSPC Rasa 750 RS240 on my main rig i have i5-2500K @ 4.8GHZ 1.36v 24-7 my temps are 61C on full load i use the NH-D14 on my secondary rig i have i7-2600K @4.5GHZ 1.32v 24-7 been that way around 8months now i recommend you get a better cooler to start with if you plan on OC past 4.5GHZ

 

[willard]

Actually, a Hyper 212 can pull off the higher overclocks if you replace the fan. I'm using one myself, with a pair of Cougars in push/pull.

At 4.9 GHz (took about 1.42v I think) it peaked at around 73c in Prime95. 4.7 GHz (my 24/7 clock) at just shy of 1.4v peaks right at 70. Anywhere north of 5.0 GHz takes a ton of voltage and gets me to 80c fast, so I'm staying well below that point until I build a decent water loop.

 

[Deleted member 217926]

Anyone thinking about approaching 1.4v and above with 32nm Sandy Bridge processors should read up on and understand electromigraton. Voltage is a measure of force. According to hardware testing right here at Tom's you can reduce the lifespan of your CPU to months or even weeks. Just understand what you are doing and why before you do it.

 

[willard]

According to hardware testing right here at Tom's you can reduce the lifespan of your CPU to months or even weeks.

I've seen this claimed in a number of places (including the article here from last year), but I'm skeptical. Before I go any farther, I need to say that I won't disagree, 1.4v is the absolute max you should use for 24/7, and you will get electromigration above that point. It may not kill your chip, but it will slowly damage it and this will lead to a failure eventually.

Basically, it comes down to numbers of dead chips versus how many are still running at those voltages. When I got my chip I did a thorough search to collect opinions on maximum safe voltage. I found a grand total of zero reports of people killing their chips by putting too much voltage through them. I did find quite a few people running much higher than 1.4v, however. I've even seen some people on water running at 1.5v-1.6v 24/7, which to me is insanity, but it illustrates the point.

If 1.4v-1.45v really killed chips in "weeks to months" as Tom's claimed, there would be a ton of reports out there of people killing their chips since there are tons of people out there running their chips at those voltages and higher. The reports just aren't there. To add my own experience to the mix, I've been running at or above 1.4v 99% of the time since I got this chip in April. I work my chip a lot harder than the average person does as well, spending several hours each day encoding or transcoding video. Still works fine, still has the same max overclock (I check every few weeks to see if I'm damaging it). According to Tom's, my chip should be at death's door by now.

Electromigration is bad, but as always, heat is the real killer. Keep your voltages sane and temps down, and your chip will be fine. You are not giving your chip a weeks long lifespan just by going up to 1.4v, and probably not even for 1.45v.

Above all, don't look for those high overclocks if you aren't willing to accept the risk of potentially killing your chip. I fully accept that my computer may just stop working one day, and I'll be out a $250 chip. I won't be happy, but it won't kill me either. If you can't afford to replace it, don't do things that might kill it.

 

[Deleted member 217926]

^ This also depends on the quality of the voltage regulation on your motherboard. If you have 12 phase VRM or better you can get away with higher voltages for longer.

Orignal reference article.

http://www.tomshardware.com/reviews/automatic-overclock-motherboard-cpu,3048.html

 

[Pinhedd]

I've seen this claimed in a number of places (including the article here from last year), but I'm skeptical. Before I go any farther, I need to say that I won't disagree, 1.4v is the absolute max you should use for 24/7, and you will get electromigration above that point. It may not kill your chip, but it will slowly damage it and this will lead to a failure eventually.

Basically, it comes down to numbers of dead chips versus how many are still running at those voltages. When I got my chip I did a thorough search to collect opinions on maximum safe voltage. I found a grand total of zero reports of people killing their chips by putting too much voltage through them. I did find quite a few people running much higher than 1.4v, however. I've even seen some people on water running at 1.5v-1.6v 24/7, which to me is insanity, but it illustrates the point.

If 1.4v-1.45v really killed chips in "weeks to months" as Tom's claimed, there would be a ton of reports out there of people killing their chips since there are tons of people out there running their chips at those voltages and higher. The reports just aren't there. To add my own experience to the mix, I've been running at or above 1.4v 99% of the time since I got this chip in April. I work my chip a lot harder than the average person does as well, spending several hours each day encoding or transcoding video. Still works fine, still has the same max overclock (I check every few weeks to see if I'm damaging it). According to Tom's, my chip should be at death's door by now.

Electromigration is bad, but as always, heat is the real killer. Keep your voltages sane and temps down, and your chip will be fine. You are not giving your chip a weeks long lifespan just by going up to 1.4v, and probably not even for 1.45v.

Above all, don't look for those high overclocks if you aren't willing to accept the risk of potentially killing your chip. I fully accept that my computer may just stop working one day, and I'll be out a $250 chip. I won't be happy, but it won't kill me either. If you can't afford to replace it, don't do things that might kill it.

It's not just Tom's Hardware claiming that 1.4 volts is the max, it's also Intel. The datasheets for Sandybridge and Ivybridge list 1.4v as the absolute safe maximum.

 

[Deleted member 217926]

It's not just Tom's Hardware claiming that 1.4 volts is the max, it's also Intel. The datasheets for Sandybridge and Ivybridge list 1.4v as the absolute safe maximum.

The max VID as per Intel is 1.52v.

 

[willard]

This also depends on the quality of the voltage regulation on your motherboard.

Absolutely. If you're going to overclock, you want a good mobo. Using the Z77 Extreme6 myself, with very good results. Not the best voltage regulation, but good. Next build will be going with one of the top of the line $300 mobos for a really extreme overclock.

http://www.tomshardware.com/reviews/automatic-overclock-motherboard-cpu,3048.html

Yep, that's the article I was thinking of. They don't really provide any info on how many chips died, if any. They just said:

Those same processors die a fairly slow death at voltage levels between 1.40 V and 1.45 V (somewhere between weeks and months on our test benches).

We don't know what motherboards they were using in their test beds, what quality VRM they had, what temperatures they were running at, what load they were running at, how many chips were being tested, how many chips actually died (if any) etc.. It's just not enough info.

My guess is they were covering their ass. It's certainly possible you could kill your chip above 1.4v, and if they'd endorsed higher voltages it opens them up to dumbasses trying to sue them because "Tom's killed my chip!"

 

[Pinhedd]

The max VID as per Intel is 1.52v.

That's the max VID, not the maximum safe voltage. The VID is based on an 8 bit voltage divider and specifies the maximum capabilities, not the safe capabilities

 

[Deleted member 217926]

I agree with the CYA aspect of telling people to stay below 1.4 but it's not just Tom's. You pretty much see the same thing on all the review sites. Makes me think there must be more than a little truth to it.

Hell I can say for sure they last a year and a half at 4.5Ghz and 1.32v! Well this is technically my second motherboard due to the P67 chipset SNAFU.

 

[Pinhedd]

I agree with the CYA aspect of telling people to stay below 1.4 but it's not just Tom's. You pretty much see the same thing on all the review sites. Makes me think there must be more than a little truth to it.

Hell I can say for sure they last a year and a half at 4.5Ghz and 1.32v! Well this is technically my second motherboard due to the P67 chipset SNAFU.

I couldn't find the absolute max for The second generation LGA1155 processors but I did find the specification for the LGA2011 processors which are very similar

http://www.intel.com/content/www/us/en/processors/core/core-i7-lga-2011-datasheet-vol-1.html

Figure 7-8 on page 60