Ubuntu :: CPUs At 100% - Therefore Excess Fan Noise
Feb 20, 2011
My setup:
HP XW8200
2x 3600MHz Xeon CPUs
4GB RAM
Quadro FX3800 PCIe graphics
Dual boot Ubuntu 10.10 32bit (primary use) and XP (gaming use only)
The issue I have is that the CPUs seem to run at 100% constantly in Ubuntu meaning that the CPU fans are constantly at high. This makes the machine sound like a wind tunnel!
It is nice and quiet in XP.
I installed GKrellM monitor which is what is telling me about the 100% CPU useage.
Would using the 64bit Ubuntu make any difference? (dont really want to reinstall tbh)
I have just purchased an AMD Phenom II 1055T 6 Core CPU to replace my aging CPU. The problem is only 1 core is visible to Ubuntu, Does it actually support 6 cores?
I'm using Ubuntu 10.04 x64, and I have an AMD Phenom 2 x4 940 CPU. First I have to underline that four cores worked perfectly in previous versions of Ubuntu as well as Windows.
When looking at system monitor it tells me that I have two CPUs/Cores, it should be four.
This is the output of CPU info:
Code: processor: 0 vendor_id: AuthenticAMD cpu family: 16 model: 4
I'm buying a new fanless desktop unit with 2 x 1.66GHz Intel Atom D510 CPUs. The vendor calls these a '64-bit processor'. But the unit has just 4Gb of memory and I don't want the fuss of 64-bit for the basic stuff I do. So I asked them to load Lucid-32.
Will the processors work with 32-bit?
The vendor tells me 32-bit won't make full use of the 4GB, only three and a half, but all else will be fine. So they'll load 32 and I'll try a live DVD of 64 in my own time. But I thought 32 bit was fine up to and including 4GB of memory?
I have an i7 980x (hexacore) machine, installed with ubuntu 10.04. Has the stock 2.6.28-15-server #52-Ubuntu SMP kernel. When I check my /proc/cpuinfo, I can see only 8 CPUs. Was expecting to see 12. Is there any way to get all my 12 CPUs.
I've got some weird stuff going on with my F14 setup, I seem to be getting loads of CPU activity with nothing happening! Both CPUs are running about 60% as soon as I log in to my normal account. If I log into another account it drops to about 15%. Sometimes when I log out I get a message saying that 'an unknown process' is still running. Looking at processes it is not apparent what is causing this as nothing is using that much CPU. It does show some processes with no name, which seem to come and go and change ID. The only things I've installed recently are dropbox and myth tv. I've removed dropbox but no change. how I can find out where my CPU time is going!
I have a high resolution file (1080p) encoded with H.264 and when I play it back, it lags at certain parts (and enough parts to destroy the job of watching the video).
I don't think that my computer is that crappy and it has been able to play many H.264 encoded files with high resolution - so there is always a possibility that this file is just encoded poorly (lacking a better computer, I cannot test).
For information, I'm running Ubuntu 9.10 64 bit. I have a Intel Core2 Quad CPU Q9550 (@2.83 GHz). I have 8 GB of RAM (forgot the speed off the top of my head). And a Nvidia GeForce GTX 280 (1 GB of memory).
During the playback, the RAM is far from being completely used; however, it seems that one (and only one) of the cores of my CPU becomes saturated at 100% precisely at the moments in which the file lags. When the CPU utilization is below 100%, there is smooth playback.
My question is - is it possible to use all four cores during playback, or is such a process not capable of easy parallel processing.
Another side question is, how can I be sure that my GPU is being used in the decoding process - or is that also being under-utilized?
Not so long time ago I have installed Centos 5.3 on my AMD Phenom II X4 server (I am hosting game servers on it). Only today I noticed, that cpuinfo shows me only two CPUs. I watched for the info in the dmesg and found there that the system boots up only 2CPUs.
Here is some info from the dmesg:
Quote:
Linux version 2.6.18-164.11.1.el5 (mockbuild@builder10.centos.org) (gcc version 4.1.2 20080704 (Red Hat 4.1.2-46)) #1 SMP Wed Jan 20 07:32:21 EST 2010 Command line: ro root=/dev/VolGroup00/LogVol00 ACPI: Local APIC address 0xfee00000
[Code]....
Am I correctly understand that something wrong in BIOS? Or it is something else?
Not so long time ago I have installed Centos 5.3 on my AMD Phenom II X4 server (I am hosting game servers on it). Only today I noticed, that cpuinfo shows me only two CPUs. I watched for the info in the dmesg and found there that the system boots up only 2CPUs.Here is some info from the dmesg: Linux version 2.6.18-164.11.1.el5 (gcc version 4.1.2 20080704 (Red Hat 4.1.2-46)) #1 SMP Wed Jan 20 07:32:21 EST 2010
ACPI: Local APIC address 0xfee00000 ACPI: LAPIC (acpi_id[0x00] lapic_id[0x00] enabled) Processor #0 0:4 APIC version 16
I have fresh install of Ubuntu 10.04 LTS Added Wine from Software Center but have no any Win apps loaded or running yet. I have also enabled visual Effects and the system installed and enabled NVidia's driver for the video-card after which the picture quality improved. Here is the screen of my CPU monitor (sorry labels arre in Russian but the charts are quite intuitive) As you can see this 100% load "jumps" from one core to another periodically. On the Processes tab there is nothing fancy:
Less than 10% processes load. I have tried to play with CPU Clock Gnome widget, it doesn't really change the picture whether I use MAX CPU clock of 2Ghz or reduced power-saving 1,2 GHz. I have found couple similar cases on forum but see no solution so far.BTW, I have already launched the update manager and currently Ubuntu does not offer any new updates.
[code]....
PS. In the attachement - Report of some built in System Diag Tool
I'm using bioinformatics programs i run from console on my system or on the server and some of them don't have a option for use multiple cores/cpus.There's a way to force it? some programs have to run for days and use a single core...
I've got a Toshiba Satellite L655D-S5109 and the gnome cpu temp applet only shows 3 cpu's. Its a dual core AMD Turion II - I would have expected 4 cpu in the temp applet.
When running cat /proc/cpuinfo under Linux, a variety information is kicked-back. For example:
> cat /proc/cpuinfo processor : 0 vendor_id : GenuineIntel cpu family : 6 model : 15 model name : Intel(R) Xeon(R) CPU 5130 @ 2.00GHz
[Code]...
First, what does all of that actually mean? I see I have a processor 0 and processor 1. Does that mean Linux is reporting both cores of the CPU, or, since it is a VM, the two that I happen to have right now (even if they're on physically different CPUs)?
Second, how can I get a similar information dump form the command line in Windows? Third, is there a way using either platform to determine the number of physical CPUs versus total CPU cores?
While searching for the maximum numbers of CPUs supported by CentOS 5 x86_64 I found the following page: [URL]. The question I have regarding the specified information ("64/255" logical CPUs) is, which number means what. Does this mean 64 real CPUs with a maximum of 255 cores/hyperthreaded CPUs or something similar or totally different?
While searching for the maximum numbers of CPUs supported by CentOS 5 x86_64 I found the following page: [URL] The question I have regarding the specified information ("64/255" logical CPUs) is, which number means what. Does this mean 64 real CPUs with a maximum of 255 cores/hyperthreaded CPUs or something similar or totally different?
I am somewhat familiar with SGE (Sun Grid Engine, now Oracle Grid Engine) commands but am having a problem when running parallel jobs.
Present Machine configuration: machine I - 12 cpus machine II - 12 cpus machine III - 12 cpus .... so on ... One_machine - 16 cpus
(I have all machine of 12 cpus and 1 machine having 16 cpus)
I want to schedule jobs on these machine such that if I ask for 12 cpus - my jobs should execute on any of the machines which has all 12 free cpus (eg machine I or machine II) in this case.
Eg. suppose I ask for 24 cpus
Option I : 12@machine1 12@machine2 ----- I need this
Options II : 10@machin1 10@machine2 4@machine3 ---- I don't need such a distribution
Hence, Option I is ideal here. Also, when running 12 jobs on machine1 (say) - even if all 12 cpus are not being used at some instance of time, none of the 12 cpus should be freed. In short, until my run finishes, all blocked cpus should remain blocked.
If you may understand, the purpose here is to run some performance tests.
I have an install of Mint 9, and the fan runs a lot less then it did in Vista, but it still is way to loud because it's not very warm at all. I was wondering if there is a tool for Ubuntu/Mint that you could adjust this issue. It would be nice if it was a gui tool, but anything will do, because it's really bugging me.
Since about a few weeks, I have a problem with the sound on my Ubuntu Jaunty 64-bit. From any application it would simply provide noise instead of the real sound. The motherboard is working fine and providing sound under Vista. Initially, I then managed to change the settings in System-> Sound from Auto to one of the options, which provided for a decent test result. Now this is not working anymore either ... so I only get noise out of my speakers.
Is it possible to keep the hard drive off most of the time, so that the laptop runs silently? (The fan has been taken care of already. Are there any other noise sources?)
I have no practical knowledge about this, but in theory it should be possible to cache frequently used files in RAM and to store disk writes on a flash drive...? UnionFS?
I'm running the latest version of Ubuntu and have been for a while now without issue. However today (actually now yesterday) i made the giant mistake of upgrading my on board HD3150 graphics to a PCIe HD5570 since then my whole life has fallen to bits.
First i noticed some visual noise on the screen. Concluding this was due to incorrect or out of date drivers i attempted to update them from the Hardware Drivers section. This gave the error:
SystemError: installArchives() failed
Activating from the command line using jockey gave the same error.
Deciding a reboot might be needed i did so. On power up all ubuntu and recovery modes produced an unskinned GUI with no desktop icons that responded to no mouse or keyboard input.
I was forced to reinstall the operating system.
On reinstalling and losing everything i had installed previously (it's going to take many days to get that all back and working again!), i still have the screen noise, have still been getting the same errors on attempting to activate the drivers, except now i also get this:
the installation of this driver failed.
Please have a look at the log file for details.: /var/log/jockey.log
Code: 2010-09-22 01:40:58,589 DEBUG: querying driver db <jockey.detection.LocalKernelModulesDriverDB instance at 0xa332aec> about HardwareID('modalias', 'usb:v045Ep00E1d0007dc00dsc00dp00ic03isc01ip02') 2010-09-22 01:40:58,589 DEBUG: querying driver db <jockey.detection.LocalKernelModulesDriverDB instance at 0xa332aec> about HardwareID('modalias', 'usb:v046Dp0804d0009dcEFdsc02dp01ic0Eisc01ip00')
I am dual booting Windows 7 and Ubuntu 10.10. When I am on Windows 7, I have no noise from the hard drive. When I use Ubuntu, every now and then I hear a short scratch noise from the hard drive. It didn't start doing it until installing Ubuntu when I was setting up the partitions.
I'm running Ubuntu 10.10 in a Samsung R510. I've tried to record sound with an external microphone and there have always been a permanent noise accompanying it.I thought it could be the internal microphone, that's causing the noise, but even when I disabled it the noise remains. Or maybe I didn't disable it the right way !
I'm getting this weird noise and (sometimes) delayed audio on HDMI out on VLC. I've tried banshee, miro, ..... on FF and all of them work just right. I would settle changing VLC's sound output to ALSA, but it's also buggy as I get no audio after pausing. Restarting or reloading Pulseaudio and ALSA didn't work either.
Does anybody have a recommendation? Obviously, I'd like to stick with VLC.
