• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

multiplier dropping down in Prime95

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

Gunny51

Registered
Joined
Mar 16, 2009
Location
West Texas
So, I am running Prime95 while monitoring RealTemp and CPUz and noticed that the clock multiplier is dropping down from 20x to as low as 10x. I am not seeing in my my MSI Eclipse bios where to stop that from happining EIST is disabled. I think there is a place in Vista 64 to stop that as well, but can not remember where to find that setting. Little Help?:confused:
 
clock multiplier cutting back

So, I am running Prime95 while monitoring RealTemp and CPUz and noticed that the clock multiplier is dropping down from 20x to as low as 10x. I am not seeing in my my MSI Eclipse bios where to stop that from happining EIST is disabled. I think there is a place in Vista 64 to stop that as well, but can not remember where to find that setting. :confused:Little Help?
 
I know that MSI's GreenPower software, when set to Max Saving, clocks down the multiplier to 10.5 (10.5 x 133MHz). And I think the Optimise setting drops the multiplier to 12 (12 x 133MHz).
 
Hi guys, thanks for your input. redduc900 I have GreenPower off; Evilizer, I have power plan on High Performance in control panel. Perhaps this is merely a perceived problem that I haddn't noticed before. When running the sensor test in RealTemp which invokes Prime95, I had been watching temps only, not CPUz. In the 10 minute test of everything the multiplier stays up (as indicated in CPUz) as proc it brought to 98% load in heat up test#1. then as it goes through the seven one minute "cool-down" tests to proc idle load I am seeing the multiplier drop to 18x then 16x then 13x and so on, down to 3X and proc idle. So my questions now are: is that a normal function of said test, wherein prime 95 overides the suposidly fixed proc clck multiplier during the "cool-down steps #'s 2 through 8? And under all other normal operation such as gaming is the proc clck and multiplier constant as max set point of 20x?
 
Last edited:
prime95 doesnt override anything, what you are saying. is as if eist is enabled and the cpu is idle, speed drops to lower power usage. it is not a problem but a normal function intel put into the cpu. if all those are disabled the last one to look for is C1E in the bios. after that you need to update the bios as it sounds like the settings dont work in bios.
 
prime95 doesnt override anything, what you are saying. is as if eist is enabled and the cpu is idle, speed drops to lower power usage. it is not a problem but a normal function intel put into the cpu. if all those are disabled the last one to look for is C1E in the bios. after that you need to update the bios as it sounds like the settings dont work in bios.

OK Evil... will look in bios. Can you define C1E for me I have seen that don't recal if it is currently in auto or disabled. EIST is disabled. I flashed to the latest (30 days ago) version 1.4 bios version. Perhaps it got a little messed up in my applying the "short CMOS" buttion a few time in my last round of OC attempts or I have missed something. C1E Auto or disabled?

BTW my current speed listed in my sig. is thus far, as fast as I have been able to attain and stable. Still working on that. Your suggestion to disable htt significantly lowered the temps: 39 idle and just 12 degree increase to 51 at full load.
 
c1e should be disabled, it functions just like EIST. it can somehow also lower the multi but its main function is to lower cpu votlage at idle.
 
c1e should be disabled, it functions just like EIST. it can somehow also lower the multi but its main function is to lower cpu votlage at idle.

K, Evil... C1E disabled, EIST double checked: disabled. ran RealTemp 10 minute sensor test again while monitoring it with CPUZ as well. Same thing (looks almost like it might be a normal occurrance): Voltage remains constant, Multiplier after heat-up step#1 and begining with cool step #2 all the way down to 3 x 172.6 for 553mhz in step#8 and #9 Step #10-Idle multiplier snaps back up to 20x.
 
unless there is something wrong with the bios, i have both disabled and on XP. my multi doesnt jump at all.. though it doesnt matter if it changes,if you leave your computer idle there is no point in letting idle at oc'd speeds. then agian if your computer is idle then your not on T32@FAH or OCF@rosetta team!
 
I am only seeing the multiplier drop as far as I know, during the cool down tests in RealTemp ''sensor test"/Prime95. Anyother time that I start CPUz it shows the set multiplier and resultant 3441.5mhz cpu speed. I would like to upload some bios screen shots as others have done to see if you or others can see anything that I might me missing. Do they snap a shot with a digital camera then transfer to PC and upload in those cases?
 
I am only seeing the multiplier drop as far as I know, during the cool down tests in RealTemp ''sensor test"/Prime95. Anyother time that I start CPUz it shows the set multiplier and resultant 3441.5mhz cpu speed. I would like to upload some bios screen shots as others have done to see if you or others can see anything that I might me missing. Do they snap a shot with a digital camera then transfer to PC and upload in those cases?

no they dont use digi cams, they hit the "print screen" button on the keyboard. then open up paint in windows and past it in, then save the image as a .jpg file. then when you post you go advanced and scroll down, then click on "manage attachments" then have it upload the pic.
 
hmmmm... I run that sensor test on mine (have the gigabyte board tho) and my multi never moved... but I'm not OC'ing it yet.. burning the AS ceramic in I think.. my temps shot up to 68 according to realtemp with a 99.7% load... so... but idle, after the test, was at 37-38 ... is there something wrong?

Also cpu-z reads my 920 at 2790 ish for speed... Is that because of the turbo thingy on my board?
 
Last edited:
hmmmm... I run that sensor test on mine (have the gigabyte board tho) and my multi never moved... but I'm not OC'ing it yet.. burning the AS ceramic in I think.. my temps shot up to 68 according to realtemp with a 99.7% load... so... but idle, after the test, was at 37-38 ... is there something wrong?

Also cpu-z reads my 920 at 2790 ish for speed... Is that because of the turbo thingy on my board?

im confused, why do you think something is wrong?

yes at 2790mhz means the multi is at 21x, provided you havent oced the BCLK. stock BLCK of 133x21=2790mhz, so your cpu is running in turbo mode.
 
just my idle looks high... Could this be good or bad?
But is it because Gunny is running a MSI board and i'm running a giga that his board is possibly dropping his multi???
And nope.. haven't OC'd it yet... might be flashing the F5 bios soon....
 
so 68 w/o an OC is acceptable? Hopefully when I get my 3.4 it will only be around 75... Isn't realtemp like 5 degrees cooler or something like that ???
 
so 68 w/o an OC is acceptable? Hopefully when I get my 3.4 it will only be around 75... Isn't realtemp like 5 degrees cooler or something like that ???

well i dont know what cooler your using but 68c at only 2.7ghz is hot. i peak in the low 70's at 3.4ghz on a ninja heatsink. from what i have seen of i7 80c is what you want to stay below. i like to keep things cooler then that, it is just me thing.

no realtemp isnt 5c cooler but if it was cooler then what? real temp and core temp would both reporting the same temps underload. now keep in mind coretemp polls the temp sensor more (well in a lower ms poll, i think its 100ms by default). leaving temps to be close but could be off by 1C since real temp will not be hitting the temp sensor at the same time as core temp.
 
I'm using the Xig Dark Knight cooler...with the AS ceramic paste. I was reading in the applications section that one of the temp reading software was off from the other, just can't remember which one is off.
 
Back