Sign in with
Sign up | Sign in
Your question

Topping out at 4ghz 1100T

Last response: in Overclocking
Share
February 14, 2011 7:50:27 PM

I have a 1100T BE, Crosshair IV Formula,4 patriot sector 5 1600mhz 2 GB ram cards, and Sapphire HD 6870.

I can get the system to boot at 4ghz on all 6 cores.
the voltage ranges from 1.38 to 1.40 set at default.
I have the multiplier set to 20.0

Temps when running Prime are idling around 48 up to around 51 max.

now if i push it any harder say 20.5 or up it wont even boot.

So what I've been reading is that if it doesn't come up you might need to raise the voltage.

So I didn't really want to push it to much without consulting someone but I did try booting at 1.425 and still it would not start.

Are my temps to high?
or do I need to change the voltage?

Idk any help would be great.
thanks!

EDIT: for the MB the north bridge is also at about 50 I had to put a fan on it to cool it down though. Stock it was running 55 idle!!!

More about : topping 4ghz 1100t

February 15, 2011 1:47:41 AM

update...
set multiplier to 20.5 and turned up the vcore to 1.4125.
running at 4.111ghz.

idle...
cpu temp is 23C
cpu volt 1.4090-1.4220

prime maxes at 47C

any suggestions?
m
0
l
a b K Overclocking
February 15, 2011 2:14:48 AM

If all you are changing is the multiplier, then increasing the voltage is really your only option. From what I've seen on forums, you're already getting near that CPU's max overclock.

Download Core Temp. That will give you accurate temps for each core.

Run Prime95 Torture Test using the "In-place large FFTs" setting for an hour, and report back the peak temps for each core.
m
0
l
Related resources
February 15, 2011 11:11:06 PM

yea i figured.

I set it to 4.1ghz and upped the voltage to 1.4125 and was stable but voltage was going way to high. it was 1.4660 at one point. I have a H50 with two 110 CFM fans on it

Would changing those to deltas 230 CFM make any difference when i overvolt?
m
0
l
a b K Overclocking
February 16, 2011 8:20:44 AM

Quote:
@ leaps
Large FFTs are for RAM testing are they not...?

Nope. That's the setting that creates the most stress on the CPU and the most heat. It only uses 8MB of RAM on that setting, just enough to partially test the on-chip memory controller.

I actually use IntelBurnTest to test the CPU's memory controller and RAM, as on the Maximum setting it will use as much RAM as your system has free. I have 12GB in my system, and it tests about 11GB of it on Maximum.

My overclock stability test routine:
1. 24 hours of Prime95 In-place large FFTs
2. 50 runs of IntelBurnTest on Maximum
If it doesn't pass both, it's not stable.
m
0
l
a b K Overclocking
February 16, 2011 2:52:52 PM

You're just not going to be happy till you push it as hard as you can now are you ^_^. Well just don't kill it.
m
0
l
a b K Overclocking
February 27, 2011 11:51:30 PM

well to be honest thats a pretty decent overclock.. i have a 1090t and i have gotten it to boot up at 4.2 ghz and 4.4 ghz you can get it to boot up at 4.4 but is it stable .. i took me setting my 1090t to 1.525 before i could get it stable at 4.4 Ghz running on a ud5 board .. using the multiplier apporach .. was able to get my ram timeings to 8-8-8- 30 at 1333mhz but it took 1.55 volts do it .. on the ram i also was running my front side buss at 2600 mhz with a modest voltage increase of .125 on the northbound really the purpose was to get it stable enough to run 3dmark 11 and get a score .. i had my 6870 overclocked to 1000mhz and ram 1100 mhz any higher on the ram and it would just crash during the test run.. it was a wild and hairy ride but fun .. i actually got a really really good score 4669 but yea.. there was no way i was going to leave it on those settings though .. more than likely would of killed my processor funny thing was at that speed my h50 was able to keep it around 32 to 40c while running the test i wouldnt dare stress test at that speed i dont trust the h50 that well..
m
0
l
February 28, 2011 12:03:33 PM

Thanks for all the replies guys you've all been a very big help. Unfortunately yesterday my H70 sprung a leak during prime95 and fried my whole computer. Gonna be a while before I can replace stuff and try some of this out.
m
0
l
March 1, 2011 3:40:24 AM

excuse me, but shouldn't he be doing "small FFT's" for 24 hours to test cpu? I though large was for testing memory stability.
m
0
l
a b À AMD
a c 242 K Overclocking
March 1, 2011 4:13:41 AM

sparkle_ftw said:
excuse me, but shouldn't he be doing "small FFT's" for 24 hours to test cpu? I though large was for testing memory stability.


I think it needs to test all of whose names overclock though the focus remains on ram overclock everything must be tested for stability in use do not have a problem,
overclock is stable or not depends on how long we have been testing it with prime95
m
0
l
March 1, 2011 4:16:06 AM

oh... I got 24 hours stable in small fft's and can not pass large fft's yet never bsod before. I am disappointed now.
m
0
l
a b À AMD
a c 242 K Overclocking
March 1, 2011 4:26:29 AM

sparkle_ftw said:
oh... I got 24 hours stable in small fft's and can not pass large fft's yet never bsod before. I am disappointed now.


if you want stable re check again whether the RAM settings are too far from JEDEC and how to Vdim. Never use Large FFT as a measure of stability. It is prone to crashes for no reason. !

to test the RAM I just use a test set intelburn maximum 20 loops and software OCCT 2hours.!
m
0
l
March 1, 2011 6:20:11 AM

So was I stable before or not? My ram is at the default speed and voltage. I have an x4 955 for overclocking, no ram change needed.
m
0
l
a b À AMD
a c 242 K Overclocking
March 1, 2011 6:25:42 AM

if you have tested it with small FFT at least 8 hours I think it is and no errors, then your system is stable
m
0
l
March 1, 2011 6:38:29 AM

Ok cool, can somebody else confirm that for me? Why did I crash is Large FFT's then?



EDIT: Holy sh!t phsyco!? Nothing but the latest and greatest in your rig, huh?
m
0
l
a b À AMD
a c 242 K Overclocking
March 1, 2011 6:52:17 AM

When Prime95 fails in any way, the very first thing you should do is run a memory test to rule that possibility out. MemTest86 is a very thorough memory testing program.
If your memory passes the tests then you're pretty much down to the CPU and the CPU/RAM interface as the most likely causes although there are some other possibilities.
m
0
l
a b K Overclocking
March 1, 2011 3:12:39 PM

I do not agree with henydiah that the Large FFTs setting should not be used. It doesn't error out for no reason, it errors out because the CPU's not stable.

The "In-place large FFTs" setting is the maximum-stress setting for the entire CPU and creates the most heat. If you cannot pass that, you are not stable.

The "Small FFTs" setting is the maximum stress setting for the FPU (math part of CPU) only. It doesn't test the whole thing.

With a stable system, you should be able to pass all three Prime95 Torture Test settings for 24 hours each.

My overclock stability test routine:
1. 24 hours of Prime95 In-place large FFTs (this stresses mainly the CPU)
2. 50 runs of IntelBurnTest on Maximum (this puts lots of stress on the CPU's memory controller and the RAM as well as the CPU itself. It tests most of your RAM -- I have 12GB of RAM, and it tests about 11GB of it)
If it doesn't pass both, it's not stable.
m
0
l
a b À AMD
a c 242 K Overclocking
March 1, 2011 3:40:51 PM

Leaps-from-Shadows said:
I do not agree with henydiah that the Large FFTs setting should not be used. It doesn't error out for no reason, it errors out because the CPU's not stable.

The "In-place large FFTs" setting is the maximum-stress setting for the entire CPU and creates the most heat. If you cannot pass that, you are not stable.

The "Small FFTs" setting is the maximum stress setting for the FPU (math part of CPU) only. It doesn't test the whole thing.

With a stable system, you should be able to pass all three Prime95 Torture Test settings for 24 hours each.

My overclock stability test routine:
1. 24 hours of Prime95 In-place large FFTs (this stresses mainly the CPU)
2. 50 runs of IntelBurnTest on Maximum (this puts lots of stress on the CPU's memory controller and the RAM as well as the CPU itself. It tests most of your RAM -- I have 12GB of RAM, and it tests about 11GB of it)
If it doesn't pass both, it's not stable.


@shadow you seem don't finish read all post massage @henydiah . He don't recommend use large fft , never
m
0
l
a b K Overclocking
March 1, 2011 7:08:37 PM

I understand.

You do not recommend using large FFTs setting. You said this: "Never use Large FFT as a measure of stability. It is prone to crashes for no reason."

I DO recommend using the large FFTs setting. It doesn't crash for no reason, it crashes because the processor isn't stable.
m
0
l
March 2, 2011 2:19:45 AM

Is it possible that my proc was stable for anything real world though? I never noticed a single issue with it at 3.8GHz. 3.7GHz wasn't stable either in Large FFt. If Large is so much more stressful, why does tf2 make my proc hotter than large FFt does? May not get too hot, but I still crash. Bleh, this is so lame! Is there a way to make it pass large but not small? I thought I had a good chip getting 3.8 at stock voltage, I guess not.
m
0
l
a b K Overclocking
March 2, 2011 3:18:20 AM

If your processor gets hotter playing a game than when doing Prime95, that means it's working harder during the game. That's bad, as it crashes during Prime95 which indicates that it's giving out bad information.

The errors may not affect normal programs though. Prime95 needs exacting, precise information but games generally don't.

If you feel it's stable and safe to use, go ahead. People do all the time.

Would I trust it? Nope. But I'm obsessive about such things. I will settle for a completely stable lower overclock rather than keep a higher overclock that gives Prime95 errors, even though I know that most apps/games don't stress the processor like Prime95 does and therefore wouldn't experience errors.
m
0
l
a b K Overclocking
March 2, 2011 6:19:41 AM

purespeed86 said:
Thanks for all the replies guys you've all been a very big help. Unfortunately yesterday my H70 sprung a leak during prime95 and fried my whole computer. Gonna be a while before I can replace stuff and try some of this out.


That horrible,what did you loose?
m
0
l
March 3, 2011 3:41:46 AM

I take that back. Large FFt is about 1-3C hotter than tf2. Although some games might make my cpu hotter than Large FFt then, I have no idea.

Small FFt is upwards of 12C hotter than Large FFt though. I don't really understand prime95.

Will a system that passes 24 hours of Large FFt for sure be stable in Small FFt? Or do I have to check both?
m
0
l
a b K Overclocking
March 3, 2011 4:03:30 AM

Once it passes one of those settings for 24 hours, I'd test it with 50 (yes fifty!) runs of IntelBurnTest on Maximum setting. That will test nearly all of your RAM, the CPU's memory controller, plus the CPU itself. That won't take nearly as long -- about 7.5 hours on my system when testing about 11GB of RAM (I have 12GB).
m
0
l
March 17, 2011 10:15:31 AM

got my 179 dollar 1055t to 4ghz easy. It's running at 1.45v (14x286) nice and cool under a Noctua D-14 HSF :)  idle 31c load 46c

m
0
l
a b K Overclocking
March 17, 2011 10:47:44 AM

@masterofevil22
Thats a lot of tweaking to do.286 from 200.What is your NB freq at?
m
0
l
February 19, 2012 4:02:51 AM

HT is at 2287.7Mhz and I apologize, but it's actually at 287*14 = 4.02Ghz.

I've gotten it as high as 4.4Ghz windows boot (1.65v) and I've run it at 4.2 before through Prime and Cinebench. I'd still say that 4.1Ghz at 1.475, 1.48 or so is about as high as it "likes" to go for regular use.

I think I got a pretty good chip, but you gotta remember; this is essentially the Phenom II X4 with two more cores and as a result, some more cache. AMD's been manufacturing essentially the same cpu's with the same 45nm fab for years. They're pretty consistent at this point.
m
0
l
!