• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Intel 560 Throttled?

Associate
Joined
23 Nov 2005
Posts
32
I am not sure if this is the correct forum, but I need some advice about my CPU.

Its a Intel PIV 560 (3.6Ghz), its about 13 months old now and has constantly run at a high temp since I first got it. I used to have the Asus Star Ice cooler but it was too noisy (and so heavy it warped the mobo) so I removed this and replaced it with the Retail cooler (and Artic silver 5). Under stress my temp now reaches 75c which I know is not good, however I have a more worrying problem.

My CPU never goes above 70% - at the moment I am running the torture test on Prime95 (Blend) and my cpu is averaging between 60 and 70%, the CPU temp is peaking around 75c.

it looks like my cpu is throttling itself because of the high temps, is that correct?

I have ordered the freezer 7 pro so hopefully that will help with the temps.

Any response welcomed - except those buy an AMD posts hahaha I will be buying an AMD as soon as they support DDR2 lol spent £140 on my RAM want to make the most of it dammit :)

**Edit for spelling DOH!**
 
i haven't used it myself but you might want to try throttlewatch.... :)

http://www.panopsys.com/Downloads.html

ThrottleWatch™ is a Windows utility for detecting and reporting CPU throttling activity in Pentium 4, Xeon, Pentium M, and Athlon 64 processors. New version 2.01 also graphs (and optionally logs) fluctuations in the CPU's core frequency and core voltage made by Intel SpeedStep™ Technology, Intel Enhanced SpeedStep™ Technology, and AMD Cool‘n’Quiet™ Technology. Throttlewatch™ and InSPECS™ share the same hardware detection engine.
 
Make sure you have adequate case temps. Intel recommend 38C or less.

#1 reason that P4s get hot (apart from voltage leaks) is that people do not always have enough case fans to provide a good airflow though the case & enough to extract the hot air out the back. Even a couple of low RPM fans will provide good air flow provided there are no case obstructions like cables in the way.

Your 560 is the 2nd worse P4 of all for temps (570 is the worst).

With adequate case cooling & a decent HSF you should get no more than 60-70C max.

Best HSF to get for a P4 is either Zalman 9500, Thermalright XP90C/120.
 
My wife has one of these in an SB81P, the white shuttle job. Its a BTX style case with very average airflow, the chip can reach almost 80C but won't throttle. Using throttlewatch the CPU sits at 2.8Ghz until it starts doing something, then switches and stays at 3.6Ghz.
 
I had a 560 that did temps excactly like yours on the stock cooler, a XP-90c and AS5 dropped them from 50idle to 30idle + 75load to 45load, that artic cooler should sort them out.

Edit: Also if you are using the latest version of CpuZ then that will show you if it is actually throttling or just perfoming worse due to the heat.
 
Thank you for all the comments I will do some testing tonight and see what I can do, the case is a Lian Li V2000 so the air flow should be pretty good, but I did notice it was pretty warm around the northbridge and GPU cooler.

Thanx again for all the feedback.
 
You have a hyperthreading CPU, you wont get windows task manager up much past 60-70% unless you run TWO cpu heavy programs at the same type. Try running two separate copies of prime, you'll get 100% right off the bat no problem.

Task manager is not a measurement of clockspeed, it simply measures how much time the processor is in idle state. If the CPU's throttling its more likely that there will be no idle time left, which would push the cpu to 100% use.

But as I said, primarily its the hyperthreading, normally P4's sit around 50-60% when infact they are really running at 99% load. Because the virtual CPU is very lightly loaded, and the main number you see is the 'average' load on all CPU's (virtual or real) in the machine.

Hope that makes some sence.
 
Ahhhh thank you, that makes perfect sense.

I tried all the suggestions above (Thanx guys) and could still not get it passed 60%.

I then tried running two instances of Prime95 as you suggest but Prime95 would not load twice, I could see a new instance open but it would close immediately if I had another copy already running, but I ran CPUBurnIn while I was running Prime and this pushed it to 100% so you were right. Thank you at least I know the CPU is actually doing what its mean to then lol.

I was going to say that whenever I do video encoding I could see that it went to around 90% and I know Hyperthreading is great for video encoding so I guess it makes sense that it would go over the 60% all other tasks are averaging at.

Thank you again for your suggestions - I can put my mind to rest and work on bringing those temps down.
 
The % for CPU in task manager is per hyperthread so 2 x50=100% overall CPU utilisation!!

If you do not have 2 x CPU usage history graphs in task manager then your Windows copy has not properly installed the correct HAL and needs to be re-installed again.
 
I definately have both CPU graphs in Task Manager, I have taken some screen shots, as you can see both CPU's in task manager are definately being utilised but they dont seem to go past 60%.

cpuz.gif

taskman.gif

perfmon.gif

The spike you see above is when I was opening Photoshop to paste the screens.
 
Off topic a bit but what motherboard are you using, I notice you dont have cpu voltage on the CpuZ shot, is it an Abit board by any chance as mine seems to do the same, it has to be something to do with the u-guru sensor.
 
Hi Agr3sive,

Yeah its an Abit NI8 SLI mobo, I guess it must be something to do with the uGuru sensor, I have tried various sensor monitoring tools and have yet to find one that reads the uGuru sensors correctly. I am stuck with the uGuru application at the moment.
 
UPDATE:

I installed the freezer pro 7 last night - initially everything looked good, temps were low (well low for a 560 lol). However even with low temps it would just hang under a stress test.

I fiddled with a few things and then decided to check the chipset since I already knew that it ran pretty hot. I have a Abit NI8 SLI mobo which has a heatsink with a heatpipe that runs to a set of fins in the backplane of the mobo (Where Audio inputs usually are). When I removed the heatsink the thermal pad just fell off.

It looked pretty cooked and the chip had some residue on which appeared to be some melted thermal paste. Anyhow I cleaned this up and applied some AS5 and reset the heatsink.

Turned the PC on and managed to get a good 15 minutes of stress testing out of it (it was now 3am in the morning so I really wanted to get to sleep haha).

CPU temp after 15 minutes of stress testing was 62 degrees, case temp was 36 degrees.

When idle the CPU temp is 46 degrees and case temp 32 degrees.

With the boxed cooler I could only do about 2 minutes of stress testing before the mobo would start beeping since the CPU temp rose above 75 degrees, so the cooler has reduced the temps by about 15 degrees.

I think I will be investing in a chipset cooler to help reduce the heat a little more.

The CPU cooler is also much quieter than the boxed cooler.

Definately worth the £15 it cost, wish I had done it months ago.

Thanx for all your suggestions :)
 
Glad everything is working alright now, those 560's really do take some taming, When I had mine at 4.3ghz with 1.6v, temps were going crazy with a XP-90c and low speed fan, I had to stick a 100cfm Tornado on there to get acceptable temps.
 
I have the exact same processor and the highest temp its been at is 64 degrees when im playing CM 03/04.

Just checked out your task manager and you have 84 processes running at once! I currently have 33.
 
Back
Top Bottom