2600k @ 4.8GHz with 1.48 vCore - Safe?

P

Pyrocles

Limp Gawd
Joined
Dec 29, 2006
Messages
196
After a few weeks of on-and-off trying, I finally got my i7 2600k CPU to complete 8 hours of error-free Prime95 blend tests at 4.8 Ghz. I had to use a vCore of 1.48, though. Is 1.48 too high? I've read that anything under 1.52 should be safe for 24/7 use. So am I OK as far as voltage?

Also, my temps averaged about 76 degrees Celsius during the 8-hour Prime95 run. The temps did spike up to 81 a couple of times, but rarely. Are these temperatures acceptable for 24/7 use? The CPU is being air-cooled using a Thermalright Silver Arrow.

Here are the settings I changed in my Asus P8P67 WS Revolution motherboard's BIOS:

Ratio = 48
CPU Voltage = 1.48
Load Line Calibration = High (50%)
CPU PLL Voltage = 1.95

Thanks for any advice!
 
Gonna bump this once.

Unless I get knowledgeable encouragement to run this 2600k CPU at 4.8 GHz with 1.48 vCore 24/7, I'll stick to my current nice and cool 4.5 GHz.
 
I think the better question is, "do you really need 4.8 instead of 4.5?"

The reason I say this is that i've seen *NOTHING* (besides cpu stress progs) use all 4cores (8 threads) on my 2600k @ 4.4 and, thus, I see no reason to add extra fuel to the flames.
 
same board as me, same CPU as me, same range on the OC, too.

I've got Load Line set at Extreme, PLL is the same.

I can do 4.5GHz all day at 1.40. Above that it starts taking a lot of voltage to stay stable. 1.45 (possibly higher) to go to 4.735GHz, and temps skyrocket under Prime95 Blended test. The ASUS AI monitor program (coninciding with RealTemp) report cores spiking occassionally to 85-87C on a CoolIT ECO ALC.

However, I have found, that for my particular setup, the more CPU speed I have, the better my TRI-SLI scales, particularly in 3D, so I wouldn't say that there's NO reason to do this high end OC, but I will say that for most it's not likely to matter.

Regardless of whether or not you're loading all of those cores (most games won't, save some flight and racing sims), the speed at which they are operating at will cause the GPUs to scale their performance, as well. With no GPU OC, and my CPU from going stock to 4.5GHz, in Just Cause 2 in surround 3D, I was able to go from 30-45FPS @ almost all medium settings to 45-60FPS.
 
You should stay under 1.36v if you can. 1.48v is not safe.

I'm lucky, since my i7 2600K is doing 4.8 at 1.34v. :)
 
Do you have PLL overvoltage turned on? If not, turn it on and see if that lets you drop the CPU voltage.1.48 seems like a lot of volts for 4.8
 
All you are ever going to get from this type of thread is anecdotal evidence. No one will have solid answers as to whether or not your situation is safe or how long it will last. The only solid answers will be the few examples where the chips have failed - AFAIK the only examples of failed chips involved 1.6+ vcore and extensive benchmarks.

Personally I have run 1.425 idle/1.44v under load on my 2500k @ 5ghz every day more or less since the chips came out. Shows no sign of degrading. My brother's set up is almost identical. There may be individuals on xtremesystems or other places that have personal experience with benching lots and lots of chips. I bought 6 2500ks when they first came out and tried to break one...I couldn't. I have set up several other units for customers, but none at high overclocks.

I have run 1.5v for 5.2 and 1.55 for 5.4 for a night at a time, and experienced no problems.

If you wanted the fastest possible core speed you probably should have gotten a 2500k instead.

A 2600k will hit those temperatures under load with the factory heatsink at stock speeds. Intel would not warranty and certify them for operation if those temperatures (or anything up to 100C) were not safe.

At the end of the day, if you are worried enough to post about it then turn it down.
 
It really depends on how long you intend to use it. If it needs to run 24/7 for 3+ years at high loads etc., turning it down will make it more likely to last that long. If it idles most of the time and you will be getting a new one in 1-2 years, don't sweat it.

The higher the temp and voltage, the shorter it will last, simple as that. And by last, I mean "how long it will be stable at a given frequency." Typically when a processor has been subjected to too much heat/voltage over a long period of time, it won't die completely, it will simply refuse to run stably at the overclocked speed, at which point you'll need to back down on the overclock for it to run with complete stability. Many of my semi-abused chips still work years later, I just have to run them at stock.
 
After all of the Clarksdales that dropped like flies and the reports of degraded Sandybridge chips I wouldn't run my chip at that voltage for 24/7.
 
2600K at 4.8Ghz with 1.48v is perfectly fine. Concerning your 4.5Ghz at 1.40v, have you tried 1.350v? My 2600K with Maximus IV Extreme-Z is completely stable at 4.6Ghz with 1.350v. See if you can get yours to run stable at 1.350v first.
 
Grrag275 said:
Id like to see those reports
Click to expand...

You don't have to look hard to hear these types of things about Clarksdale. Here is a good thread about Sandybridge. I've heard similar reports here. You can believe what you want.

All that I know is that I wouldn't go running a higher voltage on a 32nm cpu than most people ran on their 45nm cpus. Whatever you choose to do is up to you.

Intel's absolute max is 1.52 without loadline. Even that seems high to me.
 
BababooeyHTJ said:
You don't have to look hard to hear these types of things about Clarksdale. Here is a good thread about Sandybridge. I've heard similar reports here. You can believe what you want.

All that I know is that I wouldn't go running a higher voltage on a 32nm cpu than most people ran on their 45nm cpus. Whatever you choose to do is up to you.

Intel's absolute max is 1.52 without loadline. Even that seems high to me.
Click to expand...

I don't see anyone/significant number of people (I only read 3 pages) in that thread citing degradation... I just see a bunch of people running extreme voltage and not having degradation?
 
Thats just one thread. There are others out there. It is something to be aware of before giving advice. Like I said I think that pumping more core voltage to a 32nm cpu than most people feed their 45nm cpus is bad advice. Do whatever you want with your hardware.
 
I was having trouble with voltage at 4.5ghz on my 2600k. Tried everything from 1.3-1.45 and nothing was prime stable, but Windows and gaming was 100% OK. Tried putting everything back to auto, lowering the secondary CPU voltages, and set the CPU vcore to 1.3 and it is finally prime stable. Load line Calibration on my motherboard makes the vcore go crazy high (read: 1.6v+).
 
It's not stable at 4.5 GHz and 1.350v. I MIGHT be able to get it stable by cranking all the relevant BIOS options to Extreme. Some of those Extreme settings (like Extreme LLC) seem to affect heat more than the vCore does, although they sometimes do allow for a lower vCore setting. Right now, I'm just using LLC on High.

JCNiest5 said:
2600K at 4.8Ghz with 1.48v is perfectly fine. Concerning your 4.5Ghz at 1.40v, have you tried 1.350v? My 2600K with Maximus IV Extreme-Z is completely stable at 4.6Ghz with 1.350v. See if you can get yours to run stable at 1.350v first.
Click to expand...
 
QuiteSufficient said:
All you are ever going to get from this type of thread is anecdotal evidence. No one will have solid answers as to whether or not your situation is safe or how long it will last. The only solid answers will be the few examples where the chips have failed - AFAIK the only examples of failed chips involved 1.6+ vcore and extensive benchmarks.

Personally I have run 1.425 idle/1.44v under load on my 2500k @ 5ghz every day more or less since the chips came out. Shows no sign of degrading. My brother's set up is almost identical. There may be individuals on xtremesystems or other places that have personal experience with benching lots and lots of chips. I bought 6 2500ks when they first came out and tried to break one...I couldn't. I have set up several other units for customers, but none at high overclocks.

I have run 1.5v for 5.2 and 1.55 for 5.4 for a night at a time, and experienced no problems.

If you wanted the fastest possible core speed you probably should have gotten a 2500k instead.

A 2600k will hit those temperatures under load with the factory heatsink at stock speeds. Intel would not warranty and certify them for operation if those temperatures (or anything up to 100C) were not safe.

At the end of the day, if you are worried enough to post about it then turn it down.
Click to expand...

Damn you have a nice chip! I need your volts to hit 4.8 under load. -_-
 
That seems pretty high for 4.8. I can intelburntest/prime95 at 48x with 1.4v (maybe a bit less) and 46x at 1.35. You may want to tweak/play around with some settings.
 
You must log in or register to reply here.
Back
Top