• A Great friend to the HardForum with a great kid that he is trying to get a scholorship to continue his schooling. Please give hime a vote! Only 24 hours left! Thanks.
    If you have an VOTE FOR KEENAN!

i5 2500k Problem?

R

r2yneth

n00b
Joined
Dec 31, 2011
Messages
9
Hey [H]ardForum, almost 3 month forum viewer and now this will be my first post.

I was wondering why this scenario below me happened:

-4.5GHz @ 1.3v BIOS
-Ran 12 hours of prime95 and then after I stopped it
-Highest temperature of 67 degrees C
-Idle temps back at 30-40 degrees C
-Stopped prime95 as I thought it was stable
-Load up SWTOR installation, instant BSOD

This is actually my first comp-build ever, using H60 w/ stock fan as my CPU cooler.

Any thoughts on what might have happened or any beginners tips?

Many thanks in advance :)
 
I'm not sure it's the processor related. I ran mine for about 40 minutes in prime 95 (4.8ghz @ 1.36v) and it froze the pc up. Didn't have any errors before the freeze in 95 though. Do you get this constantly when loading SWTOR?
 
Superfly3176 said:
I'm not sure it's the processor related. I ran mine for about 40 minutes in prime 95 (4.8ghz @ 1.36v) and it froze the pc up. Didn't have any errors before the freeze in 95 though. Do you get this constantly when loading SWTOR?
Click to expand...

Nope, just a one-time thing, only when I loaded up the SWTOR installation.

But I did that after stopping a 12 hour prime95. Could this be the effect? I booted it back up around 5 minutes later and the installation worked fine. :confused:
 
Do you overclock with manual voltages, or do you use offsets? If latter, you may be encountering crash when voltages drop on idle. Needs more voltage tweaking.


Also in Prime, do you have SUM(INPUTS) error checking and round-off checking on? If not, you may have had small errors but they were ignored. Such errors may result in crashes in other programs.
 
Could be a RAM issue. If you are running 4 DIMMs, you may have to nudge the voltage up a tad from 1.5 (somewhere between 1.5-1.55) Even if you are not, could be your voltage is low for some reason on the RAM. If you use the small fft run in prime, it will not test ram that much.
 
Is PLL overvoltage enabled though? Sometimes it acts as a switch and let's you increase the multi more.
 
Enable internal PLL voltage, CPU PLL voltage should not exceed 1.9 so 1.8V is fine for me.

I have my load line calibration @ 130% and CPU Current Capability @ 130%. VCore is 1.4V.

5.0GHz Prime Stable (for about 2 hours at least). When I had my Current Capability @ 140%, I'd get random BSODs.
 
tacohunter52 said:
Turn on PLL overvoltage.
Click to expand...

either the internal pll voltage is broken on the gene z III or my processor reached a limit. It freezes at the windows logo at 4.9. Im at 120% and i did use 130% no go.You guys have chips made in Maylay?
 
UnrealCpu said:
either the internal pll voltage is broken on the gene z III or my processor reached a limit. It freezes at the windows logo at 4.9. Im at 120% and i did use 130% no go.You guys have chips made in Maylay?
Click to expand...

I'm sure it's your board or some board setting that is out of whack. I have a 2500K bought in May and a 2500K bought in September. Both are Malay. In fact, all 2500Ks should be Malay IIRC.

S2008 3.30GHZ
Malay
L108A709

You manually set the vCore in BIOS? Are you on Offset mode?
 
Spare-Flair said:
I'm sure it's your board or some board setting that is out of whack. I have a 2500K bought in May and a 2500K bought in September. Both are Malay. In fact, all 2500Ks should be Malay IIRC.

S2008 3.30GHZ
Malay
L108A709

You manually set the vCore in BIOS? Are you on Offset mode?
Click to expand...

mines costa rica
plus ive heard rumors that there is a batch the performs better around your time you got yours then todays models that a month old
 
UnrealCpu said:
either the internal pll voltage is broken on the gene z III or my processor reached a limit. It freezes at the windows logo at 4.9. Im at 120% and i did use 130% no go.You guys have chips made in Maylay?
Click to expand...

You may have just hit the wall for your chip. Mine struggles to hit 5ghz on almost the same board as you (The LX) and almost the same cooler.
 
djsi said:
A lot of people have reported topping out at 4.9.
Click to expand...

I actually got into windows but got a BSOD at 4.9. The only way was to reduce my stock timmings on my ram which i really dont want to do because its 1600mhz ram. Anyway its funny because the chip is super stable at 4.8 in prime95. I consider it stable if i i can do a torture test for at least 20-30min. 24/7 i feel is really no need to consider it stable. All my years in overclocking i never seen anything like this. You would think i could increase ram speed or base clock 10 mhz and it would be stable but its not. I reduce my cpu volts from a stable 1.45 to 1.435 and boom not stable lol. Funny chip.

I can get into the bios at 5.0 or even 5.1 but exhibits the same issues at 4.9. Makes me wonder if it is a bios issue with the GeneZ/III . It is the first bios but i am sure its practically the same thing as the original GeneZ
 
UnrealCpu said:
I actually got into windows but got a BSOD at 4.9. The only way was to reduce my stock timmings on my ram which i really dont want to do because its 1600mhz ram. Anyway its funny because the chip is super stable at 4.8 in prime95. I consider it stable if i i can do a torture test for at least 20-30min. 24/7 i feel is really no need to consider it stable. All my years in overclocking i never seen anything like this. You would think i could increase ram speed or base clock 10 mhz and it would be stable but its not. I reduce my cpu volts from a stable 1.45 to 1.435 and boom not stable lol. Funny chip.

I can get into the bios at 5.0 or even 5.1 but exhibits the same issues at 4.9. Makes me wonder if it is a bios issue with the GeneZ/III . It is the first bios but i am sure its practically the same thing as the original GeneZ
Click to expand...

Again, I'm sorry to say but I think it is your board or some strange board behavior because a 48x hard limit where you cannot even raise the base clock 10MHz seems too artifical to be real.

I just purchased a brand new Costa Rica 2500K and while it runs a little hotter, it still works at 5GHz at the same settings as my previous Malay CPU from months ago. Primed for 1 hour now @ 1.416V

Try this. In the BIOS set your DDR3 to 1333MHz and 1.5V (for safety for your CPU's memory controller, trust me). Then set manually to 100 BCLK and 50x multiplier. Set the offset voltage to + sign to about 1.10V to 1.20V. Set load line calibration and CPU current capability to ultra-high or extreme and 130 or 140%. Set internal PLL voltage to Auto and DIGI+VRM to Extreme.
 
Last edited:
Hmmm... I thought this was a thread about helping out r2yneth?

r2yneth, can you please provide more details about your PC? So far all I've seen is your CPU and MB models, and the voltage you have the CPU set to. I think telling us which GPU you use, PSU wattage, amount of RAM [value of each stick and stick amount please].

Also, like a lot of people, I do not believe running Prime95 for 12hrs automatically notarizes your PC as stable; you should really try LinX and run it for at least 20-25 loops. The thing with Prime95 is that it is not stressing the CPU to its maximum capacity 100% of the time during those 12hrs you ran; meanwhile LinX will stress the hell out of that CPU each and every run, to the point that you will know nothing else will stress it to that level hence telling you that it is rock solid.
 
Telo said:
Hmmm... I thought this was a thread about helping out r2yneth?

r2yneth, can you please provide more details about your PC? So far all I've seen is your CPU and MB models, and the voltage you have the CPU set to. I think telling us which GPU you use, PSU wattage, amount of RAM [value of each stick and stick amount please].

Also, like a lot of people, I do not believe running Prime95 for 12hrs automatically notarizes your PC as stable; you should really try LinX and run it for at least 20-25 loops. The thing with Prime95 is that it is not stressing the CPU to its maximum capacity 100% of the time during those 12hrs you ran; meanwhile LinX will stress the hell out of that CPU each and every run, to the point that you will know nothing else will stress it to that level hence telling you that it is rock solid.
Click to expand...

Tyvm, I will try out LinX! Are there any safety precautions I should note before running this program? (eg monitoring the temps and manually turning off the program in case temps reach 80+)

At the moment I'm running

i5 2500k @ 4.5 GHz (1.31 Vcore)
P8Z68-V Pro
Corsair H60
4x4GB of G.Skill Ripjaw X Series DDR3-1600
OCZ ZX-Series 850W
Gigabyte GTX 560 Ti factory OC
Case: Antec 900
- 2 120mm front intake fans
- 1 120mm side intake fan
- 2 120mm push-pull configuration exhaust fans
- 1 200mm top exhaust fan

I've added one more fan to the H60 (2x scythe typhoons) in order to get the push-pull configuration, however, the temperatures are exactly the same with the stock corsair fan - any clues on this issue?

For prime95, I am getting temperatures of 62C @ 1.31 Vcore and idle temps at 32-40C. I've read that with the H60 push-pull configuration, I should be getting anything from under 58C under load and 24-30C idle, honestly have no clue what's up with that.

Many thanks for the help provided thus far! :)
 
Spare-Flair said:
Again, I'm sorry to say but I think it is your board or some strange board behavior because a 48x hard limit where you cannot even raise the base clock 10MHz seems too artifical to be real.

I just purchased a brand new Costa Rica 2500K and while it runs a little hotter, it still works at 5GHz at the same settings as my previous Malay CPU from months ago. Primed for 1 hour now @ 1.416V

Try this. In the BIOS set your DDR3 to 1333MHz and 1.5V (for safety for your CPU's memory controller, trust me). Then set manually to 100 BCLK and 50x multiplier. Set the offset voltage to + sign to about 1.10V to 1.20V. Set load line calibration and CPU current capability to ultra-high or extreme and 130 or 140%. Set internal PLL voltage to Auto and DIGI+VRM to Extreme.
Click to expand...

Did not work after several reboots.. Fully stable at 4800 still.
must be a crappy chip in a few months i will be sellign it and getting the ivy anyway.. I am pretty satisfied at this speed.. The only thing i could think of is the computer does not like booting with 16gb ram or my PSU.. Its a Antec Tru power 650w.
I am running 5 HDD 1 SSD a GTX 570 , DVD, etc.. The PSU is about 3 years old so it could be that idk. I almost got into windows at 4900 but gave a bSOD. This is as stable as it gets guess. Also before i got my H100 my chip was running a constant 60-85c for the 1st week the chip was installed. Could of damaged something idk..
 
r2yneth said:
Tyvm, I will try out LinX! Are there any safety precautions I should note before running this program? (eg monitoring the temps and manually turning off the program in case temps reach 80+)

At the moment I'm running

i5 2500k @ 4.5 GHz (1.31 Vcore)
P8Z68-V Pro
Corsair H60
4x4GB of G.Skill Ripjaw X Series DDR3-1600
OCZ ZX-Series 850W
Gigabyte GTX 560 Ti factory OC
Case: Antec 900
- 2 120mm front intake fans
- 1 120mm side intake fan
- 2 120mm push-pull configuration exhaust fans
- 1 200mm top exhaust fan

I've added one more fan to the H60 (2x scythe typhoons) in order to get the push-pull configuration, however, the temperatures are exactly the same with the stock corsair fan - any clues on this issue?

For prime95, I am getting temperatures of 62C @ 1.31 Vcore and idle temps at 32-40C. I've read that with the H60 push-pull configuration, I should be getting anything from under 58C under load and 24-30C idle, honestly have no clue what's up with that.

Many thanks for the help provided thus far! :)
Click to expand...

You are right the H60 should be able to handle that CPU, but it is also dependent of your ambient temps. Have you tried re-seating the H60 block/pump? I dropped a good 5-8 degrees from re-seating my H70 the last time.

Getting back to the BSOD, it doesn't look like your PSU is the culprit, so that trumps 1 of my theories.

Do you have a spare video card laying around?
Have you tried the install with default settings? Give that a try next when you get a chance.
 
Telo said:
You are right the H60 should be able to handle that CPU, but it is also dependent of your ambient temps. Have you tried re-seating the H60 block/pump? I dropped a good 5-8 degrees from re-seating my H70 the last time.

Getting back to the BSOD, it doesn't look like your PSU is the culprit, so that trumps 1 of my theories.

Do you have a spare video card laying around?
Have you tried the install with default settings? Give that a try next when you get a chance.
Click to expand...

bump the vcore to 1.45 and do 4800 mhz

you probably have the same type of chip i do.

1.31 is too low you need at least 1.35+ for 4500. I think these new batches of i2500k suc my idle at a cold start is 25-26 c then stable at 34-35 c. Full load is 72c
stable at 1.45 volts 4800mhz
 
Telo said:
You are right the H60 should be able to handle that CPU, but it is also dependent of your ambient temps. Have you tried re-seating the H60 block/pump? I dropped a good 5-8 degrees from re-seating my H70 the last time.

Getting back to the BSOD, it doesn't look like your PSU is the culprit, so that trumps 1 of my theories.

Do you have a spare video card laying around?
Have you tried the install with default settings? Give that a try next when you get a chance.
Click to expand...

Do you mean taking off the H60 and applying new thermal paste and repositioning? Do I have to scrap off all the previous thermal paste?

I've installed everything back to stock since I BSOD'd twice, but the issues have disappeared also when I upped the Vcore to 1.31. So I'm not sure if they both combined for the greater good, but I'll be installing the latest drivers soon to test it out. Unfortunately, I do not have a spare video card lying around, I really wish I did - I can't tell if its my gpu causing the problem or not :(

UnrealCpu said:
bump the vcore to 1.45 and do 4800 mhz

you probably have the same type of chip i do.

1.31 is too low you need at least 1.35+ for 4500. I think these new batches of i2500k suc my idle at a cold start is 25-26 c then stable at 34-35 c. Full load is 72c
stable at 1.45 volts 4800mhz
Click to expand...

The thing is, I want to stabilize at 4.5 before I up it any more. I'm not sure what's causing the issues atm but if I can figure it out, then I can OC to 4.8 while catering to the problems at the same time.
 
r2yneth said:
Do you mean taking off the H60 and applying new thermal paste and repositioning? Do I have to scrap off all the previous thermal paste?

I've installed everything back to stock since I BSOD'd twice, but the issues have disappeared also when I upped the Vcore to 1.31. So I'm not sure if they both combined for the greater good, but I'll be installing the latest drivers soon to test it out. Unfortunately, I do not have a spare video card lying around, I really wish I did - I can't tell if its my gpu causing the problem or not :(



The thing is, I want to stabilize at 4.5 before I up it any more. I'm not sure what's causing the issues atm but if I can figure it out, then I can OC to 4.8 while catering to the problems at the same time.
Click to expand...

What exactly are your issues? We can't help you figure them out if you don't talk about them. Just BSOD? My most common cause of BSODs on the same board and CPU as you have is that when EIST/Speedstep kicks in that the voltage scales too far and the system crashes. It's Prime stable for 24 hours @ 4.5 - 5GHz at a pretty low voltage but it's too low for EIST/Speedstep. Once I stopped Prime 95 and the CPU drops itself to 1600MHz and under 1.00V, doing simple things like browsing the internet will crash it!
 
Spare-Flair said:
What exactly are your issues? We can't help you figure them out if you don't talk about them. Just BSOD? My most common cause of BSODs on the same board and CPU as you have is that when EIST/Speedstep kicks in that the voltage scales too far and the system crashes. It's Prime stable for 24 hours @ 4.5 - 5GHz at a pretty low voltage but it's too low for EIST/Speedstep. Once I stopped Prime 95 and the CPU drops itself to 1600MHz and under 1.00V, doing simple things like browsing the internet will crash it!
Click to expand...

I have SpeedStep disabled, the problem is I have no idea why my system BSOD'd in the first place and was wondering if anyone had similar issues. I was at 1.3 Vcore, prime95'd for 12 hours and a bit more and everything was running perfectly fine, then I suddenly BSOD almost instantly on SWTOR installation. Now that I have upped it to 1.31 Vcore and downgraded my gpu's latest drivers, nothing seems to be a problem. But if it was really as simple as downgrading drivers then I can overclock even more with no worries now.

My second problem is the temperature problem - with the H60 in a push-pull configuration. With the stock fan I was getting idle temps of 30-40 and full load temps of under 65. Now with double scythe typhoon fans, I'm getting nearly the same exact temps. There has to be something wrong with this, but I've done everything correctly and set my fans up in the correct position.

Case: Antec 900
- 2 120mm front intake fans
- 1 120mm side intake fan
- 2 120mm exhaust fans (configured in a push-pull, so they are both exhaust)
- 1 200mm top exhaust fan

I haven't tried to re-position the H60, I'm not sure about what to do with the thermal paste from the first attempt. However, I turned the screws a bit more tightly until I can't turn them anymore with force - haha, I know, it's not going to help :( - and the temperatures are exactly the same, if not, worse.
 
r2yneth said:
I have SpeedStep disabled, the problem is I have no idea why my system BSOD'd in the first place and was wondering if anyone had similar issues. I was at 1.3 Vcore, prime95'd for 12 hours and a bit more and everything was running perfectly fine, then I suddenly BSOD almost instantly on SWTOR installation. Now that I have upped it to 1.31 Vcore and downgraded my gpu's latest drivers, nothing seems to be a problem. But if it was really as simple as downgrading drivers then I can overclock even more with no worries now.

My second problem is the temperature problem - with the H60 in a push-pull configuration. With the stock fan I was getting idle temps of 30-40 and full load temps of under 65. Now with double scythe typhoon fans, I'm getting nearly the same exact temps. There has to be something wrong with this, but I've done everything correctly and set my fans up in the correct position.

Case: Antec 900
- 2 120mm front intake fans
- 1 120mm side intake fan
- 2 120mm exhaust fans (configured in a push-pull, so they are both exhaust)
- 1 200mm top exhaust fan

I haven't tried to re-position the H60, I'm not sure about what to do with the thermal paste from the first attempt. However, I turned the screws a bit more tightly until I can't turn them anymore with force - haha, I know, it's not going to help :( - and the temperatures are exactly the same, if not, worse.
Click to expand...

IIRC, Corsair's instructions for all their H series coolers (my H80 was the same) was to run them as intakes, not exhausts. Have you tried that? I find that works perfectly with the top exhaust as the hot air from the radiator naturally rises and will want to exit out the top as well. Make the fronts exhaust as well.

With the stock fan, what voltage and clockspeed were you at? The same or different?
 
Spare-Flair said:
IIRC, Corsair's instructions for all their H series coolers (my H80 was the same) was to run them as intakes, not exhausts. Have you tried that? I find that works perfectly with the top exhaust as the hot air from the radiator naturally rises and will want to exit out the top as well. Make the fronts exhaust as well.

With the stock fan, what voltage and clockspeed were you at? The same or different?
Click to expand...

Ah, well noted, I will change them to intakes. The front fans however, I tried for a very long time to take them out but to no avail. I guess I'll be heavily relying on the top 200mm exhaust fan? With the stock fan, I was at the same Vcore of 1.31 with clockspeeds of nearly 1900-2000 RPM. The new 2x scythe silent typhoons I have are at 1850 RPM, but their overall performance should easily outmatch corsairs single fan.

Would it be able to have everything intake but the top fan? Many thanks for your help :)
 
r2yneth said:
Ah, well noted, I will change them to intakes. The front fans however, I tried for a very long time to take them out but to no avail. I guess I'll be heavily relying on the top 200mm exhaust fan? With the stock fan, I was at the same Vcore of 1.31 with clockspeeds of nearly 1900-2000 RPM. The new 2x scythe silent typhoons I have are at 1850 RPM, but their overall performance should easily outmatch corsairs single fan.

Would it be able to have everything intake but the top fan? Many thanks for your help :)
Click to expand...

Hmm...

http://www.corsair.com/cpu-cooling-...g-cpu-cooler/hydro-series-h60-cpu-cooler.html

37005551.jpg


====

Hydro H60 Fan Static Pressure: 3.2mm/H20 @ 1700 RPM

Gentle Typhoon @ 1850 RPM = 0.081inH20 (~2.06mm) and 0.113inH20 @ 2150 RPM (~2.87mm)

===
In other words, the static pressure on Gentle Typhoons is not great and that's what you need to push air through a radiator. That's why Typhoons are quiet. They push high CFMs but have poor static pressure because of longer blades. I would use Typhoons for system exhaust but never for pushing air through an object like a heat sink or radiator. But in a push-pull pair, it would definitely outpower the stock Corsair fan.
 
Last edited:
That's something I've never seen :(

http://www.overclock.net/t/612436/official-corsair-h50-h70-club/19890

The scythe gentle typhoon has been recommended countlessly as "THE" pair of fans to use with this radiator. This is something new to me...

EDIT: I posted my question on there as well, hopefully some experts who have been working closely with the H-series might know whatsup.

Can't thank everyone here enough though, the continued help has been much appreciated!
 
Last edited:
r2yneth said:
That's something I've never seen :(

http://www.overclock.net/t/612436/official-corsair-h50-h70-club/19890

The scythe gentle typhoon has been recommended countlessly as "THE" pair of fans to use with this radiator. This is something new to me...

EDIT: I posted my question on there as well, hopefully some experts who have been working closely with the H-series might know whatsup.

Can't thank everyone here enough though, the continued help has been much appreciated!
Click to expand...

Stupid metric to standard conversions. Why would Scythe, a Japanese company have their measurements in inches? I am assuming that's what "inH20" means instead of "mm/H20". I had to convert to mm and the Gentle Typhoons seem to be not as bad anymore as I thought and would outpower the Corsair fan when you are using two of them.
 
Spare-Flair said:
Stupid metric to standard conversions. Why would Scythe, a Japanese company have their measurements in inches? I am assuming that's what "inH20" means instead of "mm/H20". I had to convert to mm and the Gentle Typhoons seem to be not as bad anymore as I thought and would outpower the Corsair fan when you are using two of them.
Click to expand...

Whew! That's one big question wiped out :D

BTW, is there a way to give rep on these forums? Can't seem to find the option anywhere...
 
There are a lot of SlipStream fans that are quiet; I own about 12 of them and I have [right now] 8 of them in use [in 3 dif setups] and they are hardly noticeable.

You can run the H60 in push/pull config taking the inside case air out, that is how I have my H70; you just need to align an intake fan as close to parallel as possible so it picks up that nice cool air as fast as possible.

As for the re-seating comment from before, yes it means to take off the block/pump from the H60, then even out the thermal paste on your CPU and block/pump. If you used to much paste the contact is not that great; I usually use a bead the size of the head of case screw [the part with the phillips indentation] right at the middle of the cpu, and four very small ones [the size of the applicator's tip] between each corner and the middle bead. I don't bother spreading it out, pressure from holding down the block/pump will do the job.
 
Telo said:
There are a lot of SlipStream fans that are quiet; I own about 12 of them and I have [right now] 8 of them in use [in 3 dif setups] and they are hardly noticeable.

You can run the H60 in push/pull config taking the inside case air out, that is how I have my H70; you just need to align an intake fan as close to parallel as possible so it picks up that nice cool air as fast as possible.

As for the re-seating comment from before, yes it means to take off the block/pump from the H60, then even out the thermal paste on your CPU and block/pump. If you used to much paste the contact is not that great; I usually use a bead the size of the head of case screw [the part with the phillips indentation] right at the middle of the cpu, and four very small ones [the size of the applicator's tip] between each corner and the middle bead. I don't bother spreading it out, pressure from holding down the block/pump will do the job.
Click to expand...

I think he mentioned he used the stock compound that was pre-applied. That stuff is usually pretty good compared to people who haven't done much thermal compound application before.

Personally I do the line method as recommended by AS for multicore CPUs...BUT it's either me or the new chip but when I had the stock compound from Antec it was 10° cooler at maximum load than it is now with my paper thin application of AS on my new chip.
 
You must log in or register to reply here.
Back
Top