Sign in with
Sign up | Sign in
Your question

I5 overclocking issue?

Last response: in CPUs
Share
February 2, 2013 4:41:09 AM

I have been lurking around various forums to start getting into over clocking my i5. Now I have a few questions, regarding voltages temperatures and speeds.

Here's a quick summary of my system, I basically just am looking to mess around with over clocking everything I can to see what all the fuss is about. I'm not really looking to go berserk, but I do want to see how far it'll go. after reading around a bit I figured 4.4ghz would be a pretty reasonable expectation out of the CPU. As far as RAM and GPU are concerned I really have no expectations.

Case: BitFenix Prodigy - White
MoBo: Asus P8Z77-I Deluxe Mini ITX
CPU: Intel i5 3570K
RAM: Corsair Dominator GT 8GB 1866 9-10-9-27
GPU: XFX HD7870 Black Edition
PSU: PC Power & Cooling Silencer MKIII
Cooler: Corsair H100i
Rad Fans: Cougar CF-V12HPB Vortex
Fan 1: BitFenix Spectre 230mm
Fan 2: Corsair AF140 Quiet Edition
SSD: Mushkin Catalyst 50GB
HDD: Western Digital WD Black 1TB
OS: Win7 Pro

The 230mm fan is at the front pulling air into the case, the 140mm fan is at the rear and exhausting air. the radiator is mounted at the top of the case with both fans pushing air above it pushing air into the case. The rad fans are controlled by the Corsair Link software and throttle up depending on water temp.

Now my question comes to the over clocking. Right now I'm working on the CPU and as far as I have come to understand IB does run relatively hot. However, after seeing some of the results I am getting I don't know if I simply got a terrible chip, there is a problem with my cooling or everyone just runs their computers in a freezer. I'm using AIDA 64 currently for stress testing. I use the CPU, FPU, cache, and Memory to test under load and then I use just FPU to get maximum temperatures. Just to give a point of reference I put the exact setup given in the Ivy Bridge/Asus over clocking guide on overclock.net:

http://www.overclock.net/t/1291703/ivy-bridge-overclock...

Using what is outlined here I got some temps after running for 5 mins on both tests outlined above. The maximum multiplier I was able to achieve was 42x and under load the CPUID says it's at 1.192v. Under CPU/FPU/cache/Memory test temperatures maxed at 67/70/70/63 for each core. When I put just the FPU test the temps hit 75/81/79/72. After letting it sit for 5 mins idle comes to about 33/32/38/26. Ambient sits around 22-23 during these tests, however those are winter temps and I do want to be ok up to about 35 in the summer just to have some headroom.

On the over clock I have been working right now I'm at 4.34GHz with a 101.0 BCLK and 43x Mult. Fully Loaded it hits max temps of 76/82/80/72@1.288v and FPU hits 85/94/89/82 @ 1.272v. Not sure if it's stable yet, only run about 15 mins test on it. This is using an offset voltage and it is at +0.050

Again, this is the first system I'm over clocking and would just like some input on the speeds and temperatures I'm seeing as compared to what most others are getting the speed seems really low and it seems to be running relatively hot, I'm pretty sure I've seen people that said they've gotten significantly better results on the stock coolers.

TL;DR My temps seem too high for an h100 and my clocks seem too low for a 3570k.

More about : overclocking issue

February 3, 2013 4:20:15 AM

Hello :hello: 

First off your temps seem HOT! Really hot for the minimal voltages that are being applied. I have a simple air cooled Intel Core i5 2500K cooled with a Coolermaster V8 and my temps don't near that. Have you checked the "seat" of your processor to see it is tight and has enough paste in the right places? I have a core i5 2500k OC'd to 4.5 Ghz and at full load under Prime95 the cores have a Max at 65 degrees C(one in particular). A few of my settings are.....

45 multiplyer
100 Frequency
CPU voltage 1.35
Dram 1.65
Intel advanced speed stepping tec enabled
turbo mode enabled

I don't offer any suggestions although other than this sight.
m
0
l
February 3, 2013 3:45:59 PM

Thanks hillmanant, thats the reason I actually ended up posting on here, the temps just seam rather unreasonable for the cooling solution... as far as the seat of the cooler, there are a few things. First of all, the back plate provided by corsair came with a thicker clear plastic sheet in the shape of the plate that was attached under the hold downs. I wasn't sure if it is supposed to stay on there or not, but none of the instructions indicated it should be removed and while not exactly difficult, it seems that removing it would be a rather intricate process for a simple shipping cover, so I left it on there. Secondly as far as the cooler actually sitting on the cpu, I first installed it using the pre-applied thermal paste, after seeing the temps I lifted it off of the cpu and re seated it. Temps dropped maybe a couple of degrees, but nothing that would indicate a major problem. I then re seated it a second time and temps stayed the same. After making the post I decided to see if maybe there was an issue with the thermal compound so I pulled the block off of the cpu, cleaned both with 90% alcohol and coffee filters, and applied some AS5 I had sitting around from my last build about 2 years ago or so. I used the method outlined by AS which include "tinting" both surfaces and then I used the line method for actually applying the paste (their suggested method for IB). And tbh, temps got worse. I can hit 98 degrees on a 4.25ghz OC running the FPU test which I believe to be about equivalent to IBT, essentially getting the CPU as hot as possible. It'll hit about 85 degrees on the blended test. This is at about 22 degrees ambient and 1.256V on the FPU and it sits at about 1.264V for the blended, though it does occasionally bounce to 1.256-1.272.

At this point I don't really know what the issue could be. Either I have a CPU that just thinks it's in the sahara, I'm incapable of installing this cooler properly, the cooler is bad, or maybe it's a limitation of the SFF size of the bitfenix? though I really would have thought with all the cooling I jammed in there it shouldn't be a significant issue... Maybe my chip is just awful? I don't know how microcenter would feel about me trying to get a different one in that instance. I don't think the cooler itself is the issue, as the water temp doesn't rise above 34 degrees... and installing it really isn't exactly technically demanding, you just put the back plate on and screw down all the hold downs... somehow I'm leaning towards the cpu being funky as it seems it should be able to hit higher clocks with less voltage than its getting usually.

I will try this setup from the OC forum you posted though and report back.
m
0
l
Related resources
February 3, 2013 4:00:37 PM

My god you didn't take off the plastic cover? There's your problem that has to come off. the CPU connection to the cooler is metal on metal, the plastic is insulating the CPU from the cooler. This guy shows how to install correctly.

http://www.youtube.com/watch?v=qkeuZPPdScw
m
0
l
February 5, 2013 1:35:14 AM

To specify, the plastic cover I am referring to is not on the CPU block but the back plate for the assembly. I did some googling on it yesterday and found that it is there to insulate the metal back plate against any possible shorts with the motherboard. However, while googling this topic I actually came across another topic explaining a loose back plate on an H100i with an ASUS motherboard. I checked this out and it turns out that with the H100i and certain motherboards the back plate does not secure properly when the stand offs are screwed in (motherboard is too thin?). I didn't think much of it because I figured the thumb screws would pull it back down, but others were saying it was the cause of their excessive temps. I went out to lowes, got some rubber washers like others with this issue suggested and installed them in between the back plate and the motherboard to take up the excess space. To be fair, I also went a bit overboard thinking there might be some issue with the heat spreader on the CPU I got that exchanged (before I found out about the issue with the back plate)... However, full load with the FPU test only now peaks at 59/60/60/58 with stock clocks @1.032v and a 36x multiplier. Gonna start pushing it and see what happens with the temps and such now.

If anyone is interested the washers I used are Hillman brand 3/16 x 1/2 x 1/16 size rubber washers. They sit slightly loose around the post of the back plate but are the closest size I could find.
m
0
l
February 5, 2013 1:39:31 AM

As A quick note, the plastic I'm talking about can be seen at 3:40 in the video, the shiny portion on top of the plate is actually a clear plastic cover.
m
0
l
a c 135 à CPUs
February 5, 2013 1:40:01 AM

thats quite high for full load temps dude.

m
0
l
February 5, 2013 1:43:05 AM

Fe59 said:
To specify, the plastic cover I am referring to is not on the CPU block but the back plate for the assembly. I did some googling on it yesterday and found that it is there to insulate the metal back plate against any possible shorts with the motherboard. However, while googling this topic I actually came across another topic explaining a loose back plate on an H100i with an ASUS motherboard. I checked this out and it turns out that with the H100i and certain motherboards the back plate does not secure properly when the stand offs are screwed in (motherboard is too thin?). I didn't think much of it because I figured the thumb screws would pull it back down, but others were saying it was the cause of their excessive temps. I went out to lowes, got some rubber washers like others with this issue suggested and installed them in between the back plate and the motherboard to take up the excess space. To be fair, I also went a bit overboard thinking there might be some issue with the heat spreader on the CPU I got that exchanged (before I found out about the issue with the back plate)... However, full load with the FPU test only now peaks at 59/60/60/58 with stock clocks @1.032v and a 36x multiplier. Gonna start pushing it and see what happens with the temps and such now.

If anyone is interested the washers I used are Hillman brand 3/16 x 1/2 x 1/16 size rubber washers. They sit slightly loose around the post of the back plate but are the closest size I could find.



Lol Hillman that's my name, glad to see you got it figured out.
m
0
l
February 5, 2013 1:45:24 AM

iceclock, do you mean the 59/60/60/58? or the numbers I posted previously?
m
0
l
a c 135 à CPUs
February 5, 2013 1:49:14 AM

the previous numbers

m
0
l
!