Overclocking the i2500K... Let's talk about it

blackhawk777

Limp Gawd
Joined
Jun 10, 2010
Messages
386
There isn't a thread in here on the i2500K so here it is. Mine is currently at 4.2 with a simple multiplier bump and slight voltage bump but I know some of you are getting more out of your chips. What clock are you stable at, and how did you get there? I would love to sit stable at 4.5 or better but I haven't had enough time to work my way there. Please [H]elp!
 
Do a search for 2600K overclocking. There have been a lot of discussion on overclocking the 2600K. The settings are almost nearly identical. I would start there for now.

Also lots of review sites have overclocking subsection in their reviews of the 25k and 26k processors. They post charts and what they did to get stable clock speeds.
 
There isn't a thread in here on the i2500K so here it is. Mine is currently at 4.2 with a simple multiplier bump and slight voltage bump but I know some of you are getting more out of your chips. What clock are you stable at, and how did you get there? I would love to sit stable at 4.5 or better but I haven't had enough time to work my way there. Please [H]elp!

Since you as using a MSI board as well, good reading material here -

http://www.bit-tech.net/hardware/cpus/2011/01/07/how-to-overclock-the-intel-core-i5-2500k/5
 
There isn't a thread in here on the i2500K so here it is. Mine is currently at 4.2 with a simple multiplier bump and slight voltage bump but I know some of you are getting more out of your chips. What clock are you stable at, and how did you get there? I would love to sit stable at 4.5 or better but I haven't had enough time to work my way there. Please [H]elp!

I gather you just used the ratio that's all?
Now put the ratio on 45. All you do now is push the vcore up by 0.025 till it pass in prime95 or if you reach 80 degrees.
80->stop

But you'll get there way before 80 if your not using a stock cooler that is.

When your done with that you go to the ram
Put it on the rams specs in the bios. Then memtest86. If it doesn't pass. Push the VccSa up by 0.025 till you get a pass in memtest86 or if you get to 1.2v
1.2v stop
If you don't get a past when you hit the 1.2v mark then drop the ram speed a bit then memtest86 again
 
1.4v on the core should get you close to 4.7ghz (according to the review here) without toching anything else.
 
Purely depends on luck of the draw.
It's possible for a CPU to multiplier wall at X44, which limit you to a maximum of x49 with PLL override enabled (PLL should allow 5 higher multipliers).

1.4v is quite low for 5 ghz. Only golden or very good chips can do 5 ghz at 1.35-1.4v stable. My first 2600k needed 1.48v load to be stable enough to game at 5 ghz. Second 1.404v (to be fair, this is with HT on; off=lower voltages). If you're unlucky, you'll need 1.5v+ and there are a lot of unluckies out there, too.

Of course, if I disable hyperthreading, I could probably be doing 5 ghz at 1.375v, but I'm not going to bother stress testing to figure that out.

I know some people who need more than 1.5v.
And from what I have seen, the 2500k's tend to have SLIGHTLY more difficulty doing 5 ghz than the 2600k's, despite lack of HT.
 
Sitting at 4.6 with coretemp reporting 1.37 VID..... played a couple hours of Crysis Warhead last night and it was running like a champ. I keep seeing different forums claiming that the max safe voltage for 24/7 operation is 1.35-1.38, and one site claimed anything higher than 1.4 would risk killing the CPU or significantly decreasing it's lifespan. What are you guys running for a 24/7 voltage? Are there people really keeping it at 1.47 or more all the time?
 
Sitting at 4.6 with coretemp reporting 1.37 VID..... played a couple hours of Crysis Warhead last night and it was running like a champ. I keep seeing different forums claiming that the max safe voltage for 24/7 operation is 1.35-1.38, and one site claimed anything higher than 1.4 would risk killing the CPU or significantly decreasing it's lifespan. What are you guys running for a 24/7 voltage? Are there people really keeping it at 1.47 or more all the time?

its safe to go over 1.4v if your cooling is up to it. Yes for benching its fine but for 24/7? No 4.7 or 4.6
screenshot007qg.png
 
Purely depends on luck of the draw.
It's possible for a CPU to multiplier wall at X44, which limit you to a maximum of x49 with PLL override enabled (PLL should allow 5 higher multipliers).

1.4v is quite low for 5 ghz. Only golden or very good chips can do 5 ghz at 1.35-1.4v stable. My first 2600k needed 1.48v load to be stable enough to game at 5 ghz. Second 1.404v (to be fair, this is with HT on; off=lower voltages). If you're unlucky, you'll need 1.5v and there are a lot of unluckies out there, too.

Of course, if I disable hyperthreading, I could probably be doing 5 ghz at 1.375v, but I'm not going to bother stress testing to figure that out.

I know some people who need more than 1.5v.
And from what I have seen, the 2500k's tend to have SLIGHTLY more difficulty doing 5 ghz than the 2600k's, despite lack of HT.

with sandy it looks the boards play a big role. Dead chips doing 5.9Ghz on gigabyte latest revisions. They haven't worked out the bios yet the manufacturers but they're getting there
 
OK; forced PLL enabled and got it to 4.7 at 1.375 for 30 minutes of Prime95 and a 3dMark11 run.... 69C max under load. I think I'll stay here; this is definitely my favorite CPU that I've owned.
 
OK; forced PLL enabled and got it to 4.7 at 1.375 for 30 minutes of Prime95 and a 3dMark11 run.... 69C max under load. I think I'll stay here; this is definitely my favorite CPU that I've owned.

did you set your ram yet?
 
I set it to it's rated timings of 1600 at 9-9-9-24 1.5V; any suggestion on tweaking it or should I just leave it as is?
 
I'm stable at 5ghz

Vcore of 1.45 temps are hitting 74ish though which is too high for me....
 
When prime 'fails', are you guys talking BSOD, or the OS just locks up?

It seems when IBT fails: BSOD..
When prime fails: OS hang up..

is that pretty much how it goes for everyone?

I can boot and bench with a 50x multi @ 1.35, but I see temps get to 80 and have to stop. Isn't that a low voltage to be seeing 80c with? Just installed a new cooler (sig), and saw about a 15 degree drop in temps at idle and with my moderate OC (4.6), so I think I have good pressure and a proper install.

But I'm wondering if theres somethign else I'm missing that could be driving up my temps.
 
When prime 'fails', are you guys talking BSOD, or the OS just locks up?

It seems when IBT fails: BSOD..
When prime fails: OS hang up..

is that pretty much how it goes for everyone?

I can boot and bench with a 50x multi @ 1.35, but I see temps get to 80 and have to stop. Isn't that a low voltage to be seeing 80c with? Just installed a new cooler (sig), and saw about a 15 degree drop in temps at idle and with my moderate OC (4.6), so I think I have good pressure and a proper install.

But I'm wondering if theres somethign else I'm missing that could be driving up my temps.

1.35 is a good vcore for a 50x. I need V1.35 for the 45x and tons more crossing the 45x multi..

I usually get BSOD running prime. Sometimes the last core fails though.
 
When prime 'fails', are you guys talking BSOD, or the OS just locks up?

It seems when IBT fails: BSOD..
When prime fails: OS hang up..

is that pretty much how it goes for everyone?

I can boot and bench with a 50x multi @ 1.35, but I see temps get to 80 and have to stop. Isn't that a low voltage to be seeing 80c with? Just installed a new cooler (sig), and saw about a 15 degree drop in temps at idle and with my moderate OC (4.6), so I think I have good pressure and a proper install.

But I'm wondering if theres somethign else I'm missing that could be driving up my temps.

What is your idle temp at 4.6? Mine is under 29C at 4.7.... just curious because that definitely seems like low voltage to hit 80C at..... maybe your case isn't exhausting the hot air fast enough?
 
What is your idle temp at 4.6? Mine is under 29C at 4.7.... just curious because that definitely seems like low voltage to hit 80C at..... maybe your case isn't exhausting the hot air fast enough?

Same here - 29C with 1.28v and a 46multi.

My case (Antec P180) has a fan exhausting in two directions outward from the CPU
antec-p180-side.jpg


and I have the fan on my Mugen mounted on the right side of the heatsink, blowing into (through) the heatsink towards the case fan at the left.

Like so:
real-subtle-2.jpg


And all case fans are manually set to the highest setting.

I haven't been able to manually adjust the CPU fan speed. I have it set to go full-on when my cpu reaches 60C. This RPM seems to be about 1420rpm. Should I be able to set a constant CPU fan speed in the bios/etc?? (Asus P67 TUF). I know I still have a lot to read/look into.

Bad idea (the fan placement)? Suggestions welcome!
 
I'm not sure about CPU fan speed setting in your bios, but I would try to add another fan on the left side for a push-pull setup as my first step. Also, I would try to add at least one intake fan to the front of the case. It looks like that case only has exhaust fans and no side intake fan either.... is that right? If so, there might not be enough airflow to displace the heat once things get cooking. Don't forget, your GPU is most likely exhausting air out of the case also. Ideally, you want to have an intake airflow equal to or higher than your exhaust airflow. Equal to means that there is a constant steady flow of new, room temp air flowing over your equipment. Higher means there is a "positive pressure" inside the case which some case designers feel is an advantage. It's like turning a whole house fan on without opening the windows. Sure, it will suck out some of the hot air, but if you open the windows it happens MUCH faster, and also replaces it with cooler air immediately.
 
Last edited:
You know I think you're right about the lack of intake flow - and my buddy mentioned that too as he was watching me put this all together.

This case came with a large intake mount for a fan to pull air in:
p180-needthisback.jpg


but there's no way it could fit in the case with all the cables. Not to mention, wouldn't it pull hot air from the PSU/CPU at the back of the case anyway?

I'll have to buy a fan and mount it at the front. Not sure I have a place for that but I'll check it out. thx!
 
I'm stable at 5ghz

Vcore of 1.45 temps are hitting 74ish though which is too high for me....

jesus you got a beauty there. 80 is the stop point. Your cpu will never run that stressed in 24/7 only in prime situations. 4.6~4.7Ghz your set. What's your batch number btw?
 
You know I think you're right about the lack of intake flow - and my buddy mentioned that too as he was watching me put this all together....I'll have to buy a fan and mount it at the front. Not sure I have a place for that but I'll check it out. thx!

There are a number of companies that sell inserts with fans that fit into 5.25" drive bays if your really tight on options; even replacing the empty bay covers with mesh inserts would be better than nothing.
 
There are a number of companies that sell inserts with fans that fit into 5.25" drive bays if your really tight on options; even replacing the empty bay covers with mesh inserts would be better than nothing.

I've seen mods where they watercooled a psu lol
 
There are a number of companies that sell inserts with fans that fit into 5.25" drive bays if your really tight on options; even replacing the empty bay covers with mesh inserts would be better than nothing.

Im searching around for people who needed an intake fan with my case - and the consensus is the provided bracket is useless (bad spot), and no one else uses it either, lol.

So, I have two air filters at the front of the case I could pull out to mount a fan (but then would also have to remove the front panel to my case to make sure it never shuts on the fan.

Or, I could reverse that top-fan at the top of the case, couldn't I? That would put the IN and OUT rather close to eachother - but would it do the trick, at least?

edit: also seems I could move both of my HD's to the bottom right corner section (where I have two extinct raptors (36 & 74gb) just hanging out not even plugged in anyway :p), then remove the uppder HD mount entirely, and there's room there for a big 120mm fan (there's even mounting holes). The door removal would again be necessary.. but I think this is my best bet. Sounds like (for most people) it made a HUGE difference.

Anyway - enough derailing... sorry OP
 
Last edited:
...also seems I could move both of my HD's to the bottom right corner section (where I have two extinct raptors (36 & 74gb) just hanging out not even plugged in anyway :p), then remove the uppder HD mount entirely, and there's room there for a big 120mm fan (there's even mounting holes). The door removal would again be necessary.. but I think this is my best bet. Sounds like (for most people) it made a HUGE difference...

Sounds like the way to go; turning around the top fan would most likely only create a current from that top fan directly to the back fan... not very effective, plus heat rises so it goes against the design. A 120 on the front taking in air would definitely make a much bigger impact. Good luck either way.
 
Can't you put the hdd up on a higher bay so that its heat can flow in the circulation quicker?
 
Back
Top