Why does Afterburner Jump around like a rabbit 8 degrees variation in 1/2 a sec

[FullmetalJacket]

Why does Afterburner Jump around like a rabbit 8 degrees Cpu temp in 1/2 sec . I have watercooling it not possible to have such variation in 1/2 a sec ,Is there any really good monitoring software to accurately monitor temps

 

[JackNaylorPE]

Afterburner jumps around because it's doing what it is supposed to do. Your card's load varies, and therefore the power it uses varies and if that varies so will card voltage and temperature. As for water cooling, if we are talking CLC type water coooling, this is very different from custom water cooling.

a) This is primarily true because CLCs have very, very low flow. The **recommended** flow rate for custom loops is 1.00 to 1.25 gpm ... below 1 gpm, we see a drop off in cooling performance which gets more and more significant the lower it gets. The H100i for example delivers just 0.11 gpm. So when you get a heat spoike, there will be a lag before, the system can draw that heat away.

b) Custom loop system are designed based around a target of a Delta T (difference between ambient air and coolant) of 10C. CLCs operate at a much higher delta T. The greater the delta T, the harder it is for the system to respond to change and the lower the overall heat transfer.

c) CLCs have small aluminum radiators and no reservoir. As such, the "thermal mass" of the coolant is much smaller than in a custom loop. The larger the thermal mass, the lower the temperature variation within the system.

All that being said 8C is a lot. I suspect there may be a thermal interface issue. You might try resetting the block using new paste:

http://pcpartpicker.com/product/mDPfrH/masscool-thermal-paste-g751

Same thermal transfer capablity nof AS% but w/o AS5s vcuring and capacitance problems (see AS5 web site home page)

http://archive.benchmarkreviews.com/index.php?option=com_content&task=view&id=150&Itemid=62&limit=1&limitstart=12

 

[The_Staplergun]

The temperature of the device itself could easily spike if put under sudden load. It's not like the water in your system that has to gradually change temps.

What I'd be concerned about is if it's getting super high temperatures.

If it spikes and can return to a lower temperature very quickly (when under a quick temporary load, or at the end of the load) or spikes and stays without climbing (if under load), that should be about right.

Reseating the heatsink obviously would be the first answer.

 

[FullmetalJacket]

Jake its CPU temps and,And stapler yeah thanks i have done that ,I reseated the block twice due to motherboard changes and used very expensive thermpaste ,but 8 degree spike and that is idling , is the difference between a cooked turkey and one walking around when it gets real hot . Do you know any good free temp monitoring software because even the cores vary in temp especially in games the first four, really cop it ,Its just not accurate it physically impossible for temps to vary that significantly in such short periods ,Anyway thanks for the input

 

[JackNaylorPE]

The laws of thermodynamics apply equally to GPU and CPU. Please note that just because you aren't doing anything, your system may not be idling. proceesd can and will activate in the background and your CPU will not downclock immediately when you arrive in Windows ... typically, the CPu will remain at full speed for 2 - 5 minutes before dropping down to 800 MHz or so.

It most certainly is possible for temps to vary that fast ... run a stress test and then stop, you will see 20C or more drop in < 1 second. Yes, the temperature of the coolant will vary slowly ... but the surface temp of the CPU/GPU will drop extremely fast as soon as heat stops being applied because the thermal mass of the CPU / GPU surface is miniscule compared to that of the block that's cooling it.

As for monitoring tools, for CPU, MoBo and everything else, HWiNFO is the best utility for temp / voltage monitoring I have ever used ... and since installing I haven't used anything else. Even provides GFX VRM info. I still use AB for GFX, ... I dunno, force of habit maybe.

 

[The_Staplergun]

You gotta be joking. The temps CAN vary that much because the item is so small and so much power is thrown through it so fast.

It's not like a large hotplate that requires time to warm up because of its sheer mass in comparison.

You're putting up to 200+ watts of power through a space the size of your thumb nail. IfAn older incandscent lightbulb runs by resistance and heat generating light and almost instantly turns on. A processor can very easily jump around. It's how well your system or the IHS can dissipate this hear that matters. Unless you're direct die mounted you still have an IHS for the heat to transfer through.

It happens on my CLC cooled CPU.

 

[FullmetalJacket]

You gotta be joking. The temps CAN vary that much because the item is so small and so much power is thrown through it so fast.

It's not like a large hotplate that requires time to warm up because of its sheer mass in comparison.

You're putting up to 200+ watts of power through a space the size of your thumb nail. IfAn older incandscent lightbulb runs by resistance and heat generating light and almost instantly turns on. A processor can very easily jump around. It's how well your system or the IHS can dissipate this hear that matters. Unless you're direct die mounted you still have an IHS for the heat to transfer through.

It happens on my CLC cooled CPU.

What your saying seems extremely possible but 8 degress and then down 8 degress in 1/2 sec and less if the heat has trasfered how does it disapear in the same time and then happen repeatedly ,now admittedluy its not a hotplate but there is some thermal mass in conjunction with the water ,I am not doughting that you are right because I haven't done the maths ,It just doesn't seem probable

 

[FullmetalJacket]

The laws of thermodynamics apply equally to GPU and CPU. Please note that just because you aren't doing anything, your system may not be idling. proceesd can and will activate in the background and your CPU will not downclock immediately when you arrive in Windows ... typically, the CPu will remain at full speed for 2 - 5 minutes before dropping down to 800 MHz or so.

It most certainly is possible for temps to vary that fast ... run a stress test and then stop, you will see 20C or more drop in < 1 second. Yes, the temperature of the coolant will vary slowly ... but the surface temp of the CPU/GPU will drop extremely fast as soon as heat stops being applied because the thermal mass of the CPU / GPU surface is miniscule compared to that of the block that's cooling it.

As for monitoring tools, for CPU, MoBo and everything else, HWiNFO is the best utility for temp / voltage monitoring I have ever used ... and since installing I haven't used anything else. Even provides GFX VRM info. I still use AB for GFX, ... I dunno, force of habit maybe.

Thanks I will try that program HWiNFO and if it keeps happening ,I know the cpu is doing some weird throttling stuff ,Its quite possible that afterburner being made for GPUs is just not set up perfectly for CPU monitoring ,because it soesn't claim to be ,its for overclocking and gpu monitoring first and foremost

 

[The_Staplergun]

It's because it's such a small amount of material. It can easily heat up and based on TDP can dissipate it just as fast.

 

[FullmetalJacket]

I have to put a caveate on this posts something weird happened. Temps suddenly became amazingly low and much more consistant 40 to 60 playing battlefront all afternoon .Solid as a rock ,I don't even wanna investigate ,its like the computer processor became self learning and changed WENT AI,Joke incase you were going to roast me about it ,Anyway don't wanna JINX IT,it s working beautifully and performing like I dreamt it would be.

 

[The_Staplergun]

That works. Don't mess with it!

 

[FullmetalJacket]

That works. Don't mess with it!

Thanks dude I wont