W701ds CPU and Motherboard temperature problem

Hello Lenovo Community,
I have a w701ds 4323 laptop with nvidia fx2800, an SSD drive and a second 500gb hdd.
I use this laptop for web development and image processing and i kept it on a cooling pad, The CPU and MB temperature has always been like between 70-80 °C, but now i've removed the cooling pad and it goes between 80-90.
This is not normal, so can you guys share your experiences with this.
This is supposed to be a powerful machine with i7 and 16gb RAM, but it doesn't seem so, so far i had the nvidia problem but i think i finally fixed it with disabling powermizer and i'm thinking about disabling the overheat slowdown because it just freezes from time to time.
I'm running windows 7 64 and a linux mint which i barely use. The power settings are for maximum performans, the fans are set to active with no power saving and the bios options for power saving are all disabled.
Please advise.
Thank you
Solved!
Go to Solution.

I have W701 built 10/2010 with i7 920XM and 16GB RAM. SSD and 500GB hard drive like yours. I'm also running 128GB SSD in the Ultrabay in place of DVD drive. All USB ports in use, including the eSATA port. When set to highest performance, I often pegged the meter on this machine during Photoshop usage. I was finding temps reaching 85F to 90F frequently (no cooling pad). I chose to open the laptop up and inspect the cooling system. After 3 years, alot of dust and junk had accumulated on both fans (slowing them down) and other various parts of the cooling ducts, etc. Cleaned all that crap out, made it look like new inside.
Running same routines as before, sustained temps above 80F are very rare. In Thinkpad default mode plugged in, average temp is mid-50's. Battery save mode, the sucker idles in the high-40's now. The Nvidia 2800 card also runs about 5F cooler than before.

Similar Messages

Arch "eats" cpu and raise temperature

Good morning guys,
I'm on a Arch x86 with gnome.
I'm not a complete linux-noob since I've been using gentoo since 2006.
My problem is that on arch the cpu and gpu temperature are always from 5 to 10 °C higher than Gentoo.
Thanks to arch wiki, I configured cpufreq.
Trying to reproduce Gentoo settings, I installed cpufreqd (from aur, with gentoo's cpufreqd.conf) and configured the nvidia card to downclock when I log in gnome.
What looks weird to me is that if I run:
watch grep \"cpu MHz\" /proc/cpuinfo
There is ALWAYS a core at full throttle. Maybe for just 2 seconds, then it goes back to slowest speed and immediately the other one raise its speed to maximum.
I set the ondemand governor, and cpufreq-info return it set correctly.
When I'm in gnome, there is always the process "gnome-shell" that takes 2-6 % of cpu.
But when I shutdown X, and I am all alone in the shell, the top command shows no processes using too much cpu, even if the strange behaviour of cpu-speed still happens (according to watch).
In gentoo the cores stay to slowest speed for much a longer time.
My conclusion is, the temperature are so high cause the pc never really "relax" due to core strange policy with speed.
In my rc.conf, that's what about modules:
MOD_AUTOLOAD="yes"
MODULES=(fuse acpi-cpufreq cpufreq_ondemand cpufreq_performance cpufreq_powersave)
and about daemons:
DAEMONS=(@hwclock @syslog-ng @crond @dbus @gpm @net-profiles @bluetooth @cpufreq @sensors @acpid @rcpbind @nfs-common @alsa @avahi-daemon @cups @cpufreqd @gdm)
I hope somebody can help me, or maybe just show me the way to fully debug the system.
Thank you very much,
MJ
ps- I hope not to posted in the wrong forum, but the other sections seems not for this kind of problem.

I'm running 2.6.39-ARCH.
Ok for the power regression, but it is correct that my core 2 duo 9400 never stay at 800MHz? Is this kernel suppose to switch back to full power one core at time for just 1-2 seconds? And I am not pretty sure this could be related to high temperature ("higher temperature" is better).
I just did a deeper test: turning off one daemon at time, this strange behavior still continue to happen. Even when "rc.d list" show no active daemon. Looking to the output of "top" and "powertop" give me no info, but maybe just cause my ignorance.
For now, thanks to everyone is trying to help.
MJ
edit:
@ chemicalfan and lukaszan
I am not complaining about performance. Instead, it seems to me that this Arch is quite faster to boot and to use. I think - but I'm not very sure of it - that this fact of compiling software under Gentoo is a bit overestimated. I can agree about the possibilities of certain personalization or just the kernel lightness, but I think that if the arch kernel (or some process I am unable to identify) can't keep my proc at the lowest frequency some problem exists.
Last edited by MajinJoko (2011-07-06 16:10:46)

CPU and Motherboard upgrade for Premiere Elements 11

Hi all. Been perusing the Premiere Elements and Pro forums but still have a few questions.
I’m upgrading Premiere Elements from PrE9 to PrE11 and am upgrading some of my desktop’s hardware at the same time. I could use some advice about choosing a CPU, graphics card, and motherboard. The goal of this upgrade is to considerably speedup transcoding without breaking the bank. The budget’s a bit flexible but I’d like to keep it at or under $500 for the motherboard/CPU combo.
The computer will be used for editing and transcoding AVCHD videos. The videos are shot in either 1080i or 1080p using a Panasonic HDC-TM700.
I’m processing about 2-3 hours of video per week and my current system takes a long time to transcode. For example, it takes well over 3 hours to convert 1 hour of AVCHD video to wmv. An Intel Core 2 Quad Q8300 with 3 GB of RAM was fine for the occasional video, but it’s too slow for what’s currently needing to be processed.
New CPU:
At this point I’m looking at the following processors:
i5 3570K
Ivy Bridge
No hyper threading
i7 2600K
Sandy Bridge
hyper threading
i7 3770K
Ivy Bridge
hyper threading
Questions:
) Does PrE11 make use of hyper threading?
) If PrE11 doesn’t use hyper threading is there much of anything to be gained by getting the i7? The i5 is about $90- $100 cheaper than the i7 CPUs and seems pretty quick.
New Motherboard:
I’ve had good luck overclocking Asus motherboards, so I’m leaning towards getting another one, but am open to alternatives.
The choice of board will depend mainly on Intel’s northbridge chip:
Ivy Bridge – Z77 chip
Sandy Bridge – Z68 chip
Question:
) Are there any other northbridge chip/CPU combos I should be looking at?
Current Graphics Card:
nVidia Geforce 8400 GS 512 MB DDR2
Drivers are up to date and downloaded from nVidia.
Question:
) Since PrE11 doesn’t use the graphics card’s GPU, will it be okay to continue using this card for the new build? I’m not a gamer and this card works just fine for the current build.
Additional hardware/software details:
PSU: Seasonic X660
HDD: Four 1TB 7200 rpm drives
Memory: 16 GB DDR3 RAM
OS: Win7 64 bit or Win8 64 bit (haven’t decided yet whether to go with Win 7 or 8)
Thanks in advance for your comments and help.

I have been going through the same process. Yes PE does support multiple threads and definatley go with the ivybridge i7 cpu. Asus is also the board I'm going with. They have made OCing much easier but be careful of high temps that leads to early failure of the cpu . Most of them let you have a maximum of 32 gig of ram and with memory so cheap check out upgrading to a higher grade memory. This can be overclocked as well. One of the ivy bridge cpus has 6 cores but cost $500 plus and requires a more expensive MB as well. too expensive for me though. I believe the built in graphics processor in the i7 k series the HD4000 is enough for PE. For about $650 I got an i7 at Microcenter for $229 and I think they just lowered the price to $200 an Asus MB at Newegg for $140 and 16gig ram for $70 also check out SSD drives in the 128 gig range are about $100 which you could use for you C: drive maybe take it up to 256 gig for more. They are getting pretty cheap
Good luck with your build.

Mapping CPU and HD Temperature

I am aware of the Activity Monitor application that comes with the OS but I was wondering if anyone had any suggestions for a widget/app that can not only show me a real time temperature readout of the CPU and HD (I don't care too much about ram) but also allow me to go back and view those readouts over a period of usage (an hour, a day etc.).
Thanks for your time =)

I have been using this widget for a while now.
I only have a question:
which of the following 5 options is most useful in understanding whether the temperature of the mini is within healthy parameters?
Right now the widget compares the CPU A temperature Diode with the Ambient Temp.
Regards
Ambient temperature
CPU A temperature diode
CPU A Heatsink
Northbridge Position 1
Northbridge Position 2

CPU and motherboard upgrades for DV7-2278ca

Hi
I have a DV-2278ca laptop with Intel Core™2 Quad Q9100 2.26-GHz with 12-MB L2 cache and 1066-MHz FSB and ATI Mobility Radeon HD 4650 discrete graphics subsystem and 1-GB of dedicated memory.
I want to upgrade them to something that uses second/third generation INTEL (I5/I7)
Is it possible? if yes what are my options?
Thanks!

I am going to be honest here. The chances of upgrading those components is very low and not very likely to work with that notebook case.
A second generation Core i5/i7 process is two generations away from the current processor which will require a motherboard upgrade. I'm certain you are aware of this since you mention upgrading the CPU and the motherboard. This is an issue since the case for the notebook was specifically matched to the motherboard it came with. It is possible that finding one that supports what you want and fits that case will be challenging. It can also add up in cost and risk. Risk being the components just don't work or a different feature of the notebook doesn't work any more.
I did do some research to find other posts, in this forum and outside this forum, where others have successfully changed those components. I found plenty on replacing it with the same motherboard, but none that were specifically upgrading it. Perhaps I did not use the right search terms. Notebooks are generally intended to be as they are configured with limited upgrades available for the hard drive and RAM.
↙-----------How do I give Kudos?| How do I mark a post as Solved? ----------------↓

Cpu and motherboard upgrade

I want to upgrade my cpu to a good one so I need a mother board and the cpu my budget is 150 or should I build my own pc I have a gpu in my pic at the moment any suggestions

hi,
firstly you need to be clear as to whether you want any future upgrades or not(namely a graphics card,in the case of which a power supply upgrade would also be needed)
assuming you only need a cpu and mobo upgrade i must say that 150$ is a little low for upgrading both...a decent cpu comes at that price...
for the cpu id recommend,firstly ,something from intel as they last longer than their amd counterparts...you could look at the following-
http://www.amazon.com/AMD-FD6300WMHKBOX-FX-6300-6-Core-Processor/dp/B009O7YORK/ref=sr_1_1?s=electron... you REALLY want something from amd this is it)
intel core i3-3210,4150(i'd go for this)
as for the mobo...if you want the intel cpu it has to be an lga 1150 socket type...at a lower price range,the
Intel DH87RL 4th Generation
mobo is quite good,the MSI B85M-G43 Motherboard is also worth a look(this one is my recommendation)
for an amd it has to be of the am3 socket type-the
Gigabyte GA-78LMT-USB3 Motherboard is what id get
youll have to stretch to around 200 at the least
this is just an outline...anything else you need,just ask

Upgraded CPU and motherboard

How do I make arch get back in touch with the install? Do I need to delete a couple files? Rerun mkinitcpio? I can chroot back into it, just need to know what to do. Thanks.

I'm sorry, I meant to get back to you guys on this, but couldn't. Arch is up, baybe.
There wasn't really a problem, actually. You think kernel panic is a problem, and it is, but in this case it was something I caused before I even switched the motherboard and CPU.
I can't remember what it was because I had not been using arch for some time. I can say I do know I left it with a kernel panic, before, and just never bothered fixing it..and then...
...on top of that the image configuration file had specified modules for the old motherboard. So, I just updated with pacman, and set the configuration to default, re-ran mkinitcpio and voila. (Sorry, I nearly forgot to mention it was necessary to regenerate the xorg.conf) So you see, had I not been monkeying around with it so much before, it would have just worked. LOL.
But now that I have your attention, I'm wondering why the system seems noticeably less responsive than a much slower VIA motherboard that I had been using. The new one is a K8T800 with an Opteron 144 in dual channel with 1gb of RAM and a 6600GT 128-bit 128mb AGP card. It replaced an Nforce2. The more responsive system is an AMD 1700+ with 512mb of RAM in single-channel with a Dell OEM MX4000 64-bit video card using 64mb of RAM. It seems like the hesitation is DMA related, as the Hdd always reads while the application waits. Any thoughts at all, including why rain has to be so cold, are entirely welcome.
Additionally, glxgears gives FPS about half what it should be.

New cpu and motherboard+something is wrong

Hi!
I am thinking of bying a amd 64 3400+ or 64 3500+ with a nforce3 motherboard,
But what is the difference between 3500+ and 3400+? I read that the 3500+ is 2.2ghz and the 3400+ is 2.4ghz. But is it worth it? I mean will it do a lot difference in games.
btw. I think my comp is quite sloow, I cant run farcry att the highest details with out freezez and it wont run smooth. I only have agp 4x, is that the problem? And when I play farcry its some strange colors everywhere I have flashed my X800 PRO into a X800XT PE but it does not go over 65 celsius, So I dont think its because its to hot right?

Quote
Power:............antec truepower 380w(ordered q-tek 650w)
I hope that is a joke, you should keep the Antec TP380 over anything Q-tech have (regardless of how many imaginary watts they print on the sticker).
The 3500+ is 2.2GHz in socket 939 (meaning dual channel DDR400) while the 3400+ is 2.4GHz in socket 754 (only one DDR400 channel).
They trade off clock speed for memroy bandwidth to reach the same preformance on average.
Themost cost effective CPU right now is probably the A64 3000+ or the 3200+ (2.0 and 2.2GHz respectively).
The 3400+ and 3500+ are a huge price hike for very little added preformance.
To play farcry your main problem right now is only 512MB of RAM, when using high settings far cry uses 1GB of RAM and you want as little of that as possible in your swap file.
The 2.53GHz P4 is good enough and the X800 card is excelent.
It will of course be a big improvement if you go for a new CPU which uses DDR400 (and replace the PC2100 with 1GB of DDR400) rather than just adding 512MB to the current system, but it's up to budget.

Plugin-container in Firefox 4; High CPU and Graphics Switching Problems and Workaround

I recently upgraded to Firefox 4, and unfortunately, I've had many problems running Firefox under OSX on my Macbook Pro.
The first problem I came across was that Firefox is now hardware accelerated. This should be a good thing, but due to the way Apple has implemented graphics switching, Firefox caused the discreet graphics card to be used (as seen with GFXcardstatus)
To prevent this, I went into preferences, and disabled "Use hardware acceleration where available" under Advanced>General.
But my problems didn't stop there. Now, it seems Firefox has decided to run plugins such as flash under a separate process called plugin-container in order to protect the browser from plugin crashes. This is a good idea (I believe Chrome and Safari do the same), but even with hardware acceleration disabled in Firefox and Flash, plugin-container still triggered the discreet graphics.
This would actually be acceptable if plugin-container was terminated after I navigated away from the page with the flash content, but this is not what happens. Instead, plugin-container keeps using the high-powered GPU until Firefox is closed.
At this point, I thought that I'd try to force the computer to only use integrated graphics (using gfxcardstatus), but then I discovered another problem:
The plugin-container process was using large amounts of CPU time when flash content was visible in the Firefox window. Typically, viewing just a static flash ad would cause plugin-container to use about 60% (in addition to firefox using about 20%, that means 80% total). Obviously, this was a battery life problem.
After all of this I think I found a temporary workaround, but there are clearly problems here that need to be resolved by the Firefox developers. There are many non tech-savvy users who are no doubt experiencing poor battery life from these problems.
'''My Temporary Solution'''
1. Disable hardware acceleration in Firefox in Advanced > General
2. Disable hardware acceleration in Flash by right clicking a flash element, clicking settings, and then unchecking "enable hardware acceleration"
2. In about:config, search for dom.ipc and set all boolean values that appear to false. This will disable Firefox's plugin-container process for all plugins.
(At this point, if you restart Firefox, Flash will not work at all)
3. Set Firefox to 32-bit mode by getting info on Firefox.app and then checking the "Open in 32-bit mode checkbox"
4. Restart Firefox. Flash should be working with reasonable CPU usage (I had something like 20% with static flash ads as opposed to 80% before), plugin-container will be no more, and your Macbook will no longer switch to discreet graphics for Firefox or Flash.
'''UPDATE:''' It seems that firefox still sometimes causes a switch to discreet graphics for certain flash elements (video players mostly). Also, I've still seen some cases where flash elements are using large amounts of the CPU, even when they are not in the current tab.
I hope this was helpful to somebody, and if anyone has a suggestion for how to make this workaround obsolete or better, please respond.

Ive downgraded from FF4 to 3.6 at this point. The CPU usage is out of control on OS X.

Q9550 cpu and p965 platinum problem

I have P965 platinum m/b with 1.8 bios. The old CPU is E6400. I just tried to instal Q9550. PC starts but stack at MSI logo. Any ideas?
EDIT: 1+3 led are red

Hello,
which Revision of the Q9550 are you using?
The CPU Support List only includes the Q9550/C1 [sSpec number = SLAWQ]. If you have the Q9550/E0 you have a CPU Support problem. That processor is not included in the CPU Support list. It requires a MicroCode-Update and currently the latest BIOS for your Board is v1.80 which does not include such an update.

What are normal Core and Motherboard temperatures

I recently purchased a HP Probook Laptop 4530s. I installed CPUID and it says that the core temp is 56c and the motherboard is 56c is this a normal operating temp?
Nothing left in my right brain, nothing right in my left brain.

""""There's no need to use any fan control software. Besides the software only increases the fan idle speed and has no other impact. The use of it could also void your warranty.""""
first of all.....smc fancontrol does what its supposed to do...cool off your mac....when i boot in to vista and play cod4 on my mac i use fan control...it keeps the temp down....there is an impact to how fast your fans spin.....and it will not void your warranty......it is an extremely usefull tool...it will keep your video card cool when playing games....as i type this iam ripping cd's and noticed the temp at about 56celsius.....i increased my fan speed and the mac is at 37celcius...obviously this works and has a purpose....
Message was edited by: ImarkG5-osx

Can i take out the motherboard and CPU and replace it with a newer one?

I was wondering if i could take out the CPU and Motherboard of an early 2011 13" macbook pro with a 2.3GHz dual core i5, and replace it with a 2012 2.9GHz 13" dual core i7? I don't care about the costs but i just want to know if it's possible? I've upgraded my computer to the max so far, but i need some more processing power and i think the new ivy bridge processors will do me some good. Thanks for any replies

NOPE.
I'm about 99.9% sure it won't fit.
Plus the rest of the stuff won't plug or won't be compatible.

P67A-C45 (B3) CPU and Memmory Compatibility Question

(sorry to my english)
Hi all i hawe 2 queestions i have CPU G630 and memorry (RAM) work to 1333MHZ if i activated in bios XMP to work on 1600 i have some problems stop working on windows, My friends tell g630 dont suport 1600ram
i dont have more money but i wont to give my cpu to friend and buy G860 becose my friend have it and i see he work with 1600MHZ but he have Z77 MB not as mine (,
1 will work my RAM on 1600MHz if i put G860 CPU to my MB?
2 i see to Global site have update bios 22nm to this mb is really posible this ? if i wont a 22nm cpu need to upgrade or bios in not for my MB ? becose on start this mb is for 32 nm?
MSI P67A-C45(B3)
Intel G630 2x2,7
Kingston Blue 2x4gb ram 1600MHZ
.....

if my P67A-C45(B3) work with Intel Core i3 2130 tray or CPU Intel Pentium G860 on Ram X.M.P. 1600MHZ i will buy some this but not wont to risk
if upgrade i sell mb+cpu and buy next
if Z77A-G43 or G45 i buy Intel® Core™ i3 3240 i dont like SLI or owerclocking and mabby is G43 low cost
if Z87-G43 i buy Intel Core i5-4570 Haswell but is to hight costs mb and cpu and have some problems with money.
i ask you all becose to my country Moldova we dont have more money on job to buy extrime somfing and i wont last tehnologis to me ))) i like Z87-G43 is good but not have some low cost cpu (
I'm afraid that MSI will go strictly by 1150 socket, and will only continue to develop them, and forget about 1155

Weird CPU temperature problem

Hi everyone.
The first time I run my system in the morning, I'm having high CPU temp readings by Core Center: 71 C, at around 40% CPU usage, running Norton Antivirus '05, and 67 at iddle. In the same conditions, Motherboard Monitor 5 reads: 67 C for N. Antivirus and 63 C at iddle.
So, I restart computer and check BIOS's PCHealth Status. It reads 46 C for CPU temp.
Then I get back to Windows XP (SP2), Core Center reads CPU temp while running N. Antivirus as 53 C, and MBM reads 50 C at 40%-80% CPU usage. At iddle, Core Center reads 49 C, and MBM5 reads 45 C.
Environment temp is about 30 C.
What do you guys think is happening? Do I need to worry about it or not?
Thanks in advance for your input.
My System:
Windows XP PRO (SP2)
MSI K8N Neo 2 Platinum BIOS 1.30
Athlon64 3800+ no OC
Leadtek A400 6800GT 256 MB
Crucial Ballistix PC4000 2x 512 MB slots 3 and 4
Zalman 7000B AlCu
Vantec Stealth 470A
Raptor 74GB SATA
Barracuda 7200.7 160 GB SATA
SoundBlaster Audigy2 ZS Platinum
NEC 3500 DVD-RW
Samsung SB616 DVD-ROM
SpeedTouch 330 ADSL modem.

I was having the same problem as described by the forum user. Despite my laziness, I even unhooked my CPU fan clean the fan and the CPU and put it back together with Arctic silver. This helped the idle temperature at START up from 50 degree C down to 43 degree C reading from Core Center. However, eventually the CPU temp will creep back up to 47 to 50 degrees C.
My breakthrough came when I read one of the thread which said to uninstalled Core Center, once I done that, my CPU temperature at Cold Start and Warm Start seemed to be consistent at around 36 to 40 degree C (I used SpeedFan 4.18 to verify my temp readings I even crossed check the temp reading with SisSandra 2005 and the two readings are consistent).
I then installed CoreCEnter 1.7.1.0 back and during cold start the temperature will creep back up and keep on increasing. Once I do a warm start everything is back to normal again. So after cold start I open both SpeedFan and CoreCenter side by side and I notice the following:
During cold start, CoreCenter never starts the CPU fan at max RPM, it starts around 2200 RPM or so. Then it would start loading the CPU varying from 1 to 4 to 100% (this happens intermittently so you won't see it if you are not staring at the screen), which would increase the CPU temperature (this is observed in the SpeedFan monitoring screen). Then it would lower the CPU fan back to 1850 RPM or so and does not increase the fan speed even when temperature goes back up. The problem does not seem to exist when one does a warm start then CoreCenter lets the fan stays at maximum speed and the temperature is stable.
I noticed when I downloaded the new CoreCenter version 1.7.1.0 software from the MSI web site, it said "fixing fan speed never reach maximum problem", so I believe this is a CoreCenter software bug. BTW, I just went back to the MSI tech support web site; it no longer showed this message. I have a feeling that they removed the message from potential liability with high temperature issue that might have led to many MB returns. (I know this is sounding awfully like a conspiracy theory)
This is my conclusion:
(1) CPU temp reading from core center is accurate minus about 3 to 4 degree C cross checked again other utilities. (If you system reaches 60 to 70, I should still be alarmed, do not just let the system idle, because it could be an actual reading minus 3 to 4 degrees).
(2) CoreCenter never let the fan reaches maximum speed during cold start and will load the system gradually therefore increase CPU temperature even more.
(3) After warm start, Core Center goes to normal and let the CPU fan goes back to maximum speed.
I know this is not a very scientific observation, nevertheless it is reproducible on my system. Please note that I have a Neo Pla set up not Neo Pla 2 therefore results might be different if you have a Neo Pla2 set up. I really doubt that would be the case since CoreCenter only has one version for both motherboards.
Please give this a try and see what you find, I am curiously to see what other gets. If we can all produce the same issue, perhaps MSI will come out with an updated version of CoreCenter again.

K8N Neo2-FX new CPU temperature problem

Well, I'm not sure if this problem is unique, but I didn't find any help by using search or google.
Problem is this: CPU temperature keeps going up slowly but very persistent and then the computer shuts down about somewhere between 80-82 degrees Celsius (says BIOS). Heatsink is warm when is touch it, i would say the CPU is max 40-50 degrees "hot". CPU is 3500+ Venice. These 2 are brand new parts. BIOS version is 1.B.
Computer was running fine one evening, night and morning (about 10 hours) with Windows XP, and then all started: shutdown. I waited couple minutes, re-started, went to BIOS H/W monitor. Temperature reading was about 50, then it was slowly rising to 80 degrees and shutdown. I instantly restarted and went to BIOS. Temp was about 70 and rising. I shutdown it manually, touched the heatsink and it felt perfectly normal (warm). I'm using the boxed cooler that came with the CPU. Fan speed is about 3150 rpm.
Right now only attached parts are PSU, MB, CPU, AGP card and 512MB DDR so I just can go to BIOS and check the temp. I have tried different PSU, AGP card and memory, the problem stays. It always starts from a reasonable value: 25 degrees after being down for a while and higher if i have tried it on. It takes about 2 minutes to go from 30 to 80 degrees.
Is this a sensor problem, CPU problem or what? One more notice: Motherbard monitor gave strange temps in Windows. Most of the time it was OK (50-55), but then both values were 0, and then back to normal. I noticed the zero values before the 1st crash too so this can be a sensor problem.
Installation failure is not possible, i have worked with computer HW over 10 years. I removed the heatsink, put it back and it fits perfectly. There is no way the CPU is 80 degrees hot. I've touched thousands warm heatsinks.
Should I send the MB/CPU package back or do you have any tricks I can try?
Thanks for reading and sharing my pain

Quote from: inf on 26-October-06, 14:57:17
I didn't change the thermal compound when I removed the heatsink. I checked that it is OK, some sort of gray grease, but not anything compared to arctic silver, but not a rubbery "pad" either.
Like I said that original compound is a one shot deal. It needs to be replaced... with the same material according to AMD for warranty purposes.
Quote
- I decided to start the computer to Windows, everything seemed to be ok, MBM showed 42 degrees about 1 minute. Then it got to 45-46 and stayed there.
- I downloaded the SpeedFan and installed it. Core temp 45, Vcore 1,40 - 1,41V. Stared the screen couple minutes, stable.
- I decided to try with a load. I Started 3DMark2001 for 5s and shut it. Core temp was 65! Very fast peak Then it went down to 53 and the weird dance began... 53-63-54-64-54-64-53-63... with the same rate than SpeedFan refresh rate. This state was stable now, I stared the screen 5 minutes and the graph showed 10 degree peaks all the time, fully idle CPU.
- Another start with 3DMark, now 30s and i halted it. 80 degrees! Heatsink was as warm as normal heatsink can be in idle computers. And then it went back to 53-63-53-63 dance.
- 15 minutes passed and I decided to boot to see BIOS values.
Weirdest part: BIOS is showing 67 degrees, rising 1 degree/ 2 secs to 80 and shutdown Vcore 1,37 - 1,39.
Conclusion: Windows is more stable than BIOS In BIOS it raises the temp to 80 and crashes, but in Windows I can do surfing and normal work and the temps stays reasonable as long as I put minimal load to CPU.
Like Syar said, CPU "halts" in Windows will result in lower temp.
Quote
Now I really think there is some kind of sensor failure. 10 degree dance in idle state is not normal. Fast peaks are normal if the heatsink is not touching the whole CPU (core) surface, but in this case it's fine. Crashing in BIOS is 1st time for me if the computer is idle stable in Windows. I'm sending both back, CPU and MB.
When you get the 10 degree jump, what does SpeedFan show for CPU usage? Also Look in Task manager/Performance to see what the CPU graph looks like and if you run Process Explorer from www.sysinternals.com you will be able to see which process is using CPU. It sounds to me like there is a process which is running at some interval and that just a small amount of CPU usage is boosting temperature excessively. Before giving up on it, I'd recheck heatsink fitting with fresh thermal compound -- a good heatsink paste, like Artic Silver will be sufficient -- to be sure, before giving up on it.

W701ds CPU and Motherboard temperature problem

Similar Messages

Maybe you are looking for