10-24-2014
Newer kernels include more support for CPU throttling, C-states. Probably stepping down the CPU when it can. What kind of load is the system usually under? Honestly even if your CPU is pegged 100% 24/7 you shouldnt be in 75-80C. Thats way too hot. 60C is where you want to be for a non-overclocked CPU even at 100% duty cycle. 30-45 for idle or normal use.
There is an entry in the kernel cpu config to enable or disable turbo speeds. ( not overclocking specifically) Maybe the old config happened to have that enabled, and the new one is disabled by default?
Not that you can compare to values pre-upgrade, but two things to look at that might be of use is to check the physical power usage of the box with a power-meter inline from the socket to the PSU. They can be had for $20-30. Again, not all that usefull without having "before" values to compare too. But more info on your system is always good to have. Second thing to look at is install "turbostat". It will show you the clock-rates and precentage of time each core is spending in various C states ( full power through idle/power-save modes). There are a handful of other tools that show the same info. Thats just the one I happened to use recently.
This User Gave Thanks to agentrnge For This Post:
6 More Discussions You Might Find Interesting
1. Solaris
Hello,
Can give me some tips to trace the GPU usage on a solaris8 update7 box?
I try to understand why a Cadcam app is so slow on my Solaris box compared to a win$ box. I guess it comes from the poor graphic card I have but i would like to emphasize it.
My bench is 150sec long on a Solaris... (0 Replies)
Discussion started by: solea
0 Replies
2. UNIX for Advanced & Expert Users
Hi ,
i want begin programming using CUDA
which enviroment can i get .I don't have desktop to buy GPU graphics card.
what should to do to get CUDA enviroment.
i'm thinking to buy desktop has this card or laptop (1 Reply)
Discussion started by: Scotch
1 Replies
3. Hardware
Dear all,
I set up a external Gforce GPU using the PE4H (Pcie passive adapter) from HWTOOLS.NET.
I'm able to add and remove the device doing
and
The point is the kernel does not initialized the device correctly.
Here is what dmesg gives after the scan:
lspci -t gives
and lspci... (2 Replies)
Discussion started by: aihake
2 Replies
4. Hardware
I am trying to start troubleshooting an error on a virtual server that uses the ubuntu 14.04 OS. Basically what happens (seeming random) is that the GPU stops processing and terminates. What Imean by seeming random is that for 3 runs there is no error then on run 4 the error appears. It has... (2 Replies)
Discussion started by: cmccabe
2 Replies
5. Hardware
Hi All
I'm find out a way to disable the DGD: AMD Radeon HD 7470M on my Ubutu 16.04.2 LTS because radeon open source module is not capable to support this GPU and consequence the boot is very slow.
I have tried serveral way with pci-stub.ids in the grub menu but not work.
In general how... (11 Replies)
Discussion started by: _Fabio_79
11 Replies
6. Linux
The situation
videocard n°1 Nvidia Ge-force(used on host linux)
videocard n°2 Ati radeon(used on guest windows 7)
host is Slackware 14.2,kernel 4.18.15
I had set vfio to pass a ati card to windows7 guest
Configure /etc/modprobe.d/vfio.conf
options vfio-pci... (1 Reply)
Discussion started by: Linusolaradm1
1 Replies
LEARN ABOUT CENTOS
cpupower-set
CPUPOWER-SET(1) cpupower Manual CPUPOWER-SET(1)
NAME
cpupower-set - Set processor power related kernel or hardware configurations
SYNOPSIS
cpupower set [ -b VAL ] [ -s VAL ] [ -m VAL ]
DESCRIPTION
cpupower set sets kernel configurations or directly accesses hardware registers affecting processor power saving policies.
Some options are platform wide, some affect single cores. By default values are applied on all cores. How to modify single core configura-
tions is described in the cpupower(1) manpage in the --cpu option section. Whether an option affects the whole system or can be applied to
individual cores is described in the Options sections.
Use cpupower info to read out current settings and whether they are supported on the system at all.
Options
--perf-bias, -b
Sets a register on supported Intel processore which allows software to convey its policy for the relative importance of performance
versus energy savings to the processor.
The range of valid numbers is 0-15, where 0 is maximum performance and 15 is maximum energy efficiency.
The processor uses this information in model-specific ways when it must select trade-offs between performance and energy efficiency.
This policy hint does not supersede Processor Performance states (P-states) or CPU Idle power states (C-states), but allows software to
have influence where it would otherwise be unable to express a preference.
For example, this setting may tell the hardware how aggressively or conservatively to control frequency in the "turbo range" above the
explicitly OS-controlled P-state frequency range. It may also tell the hardware how aggressively it should enter the OS requested C-
states.
This option can be applied to individual cores only via the --cpu option, cpupower(1).
Setting the performance bias value on one CPU can modify the setting on related CPUs as well (for example all CPUs on one socket),
because of hardware restrictions. Use cpupower -c all info -b to verify.
This options needs the msr kernel driver (CONFIG_X86_MSR) loaded.
--sched-mc, -m [ VAL ]
--sched-smt, -s [ VAL ]
--sched-mc utilizes cores in one processor package/socket first before processes are scheduled to other processor packages/sockets.
--sched-smt utilizes thread siblings of one processor core first before processes are scheduled to other cores.
The impact on power consumption and performance (positiv or negativ) heavily depends on processor support for deep sleep states, fre-
quency scaling and frequency boost modes and their dependencies between other thread siblings and processor cores.
Taken over from kernel documentation:
Adjust the kernel's multi-core scheduler support.
Possible values are:
0 - No power saving load balance (default value)
1 - Fill one thread/core/package first for long running threads
2 - Also bias task wakeups to semi-idle cpu package for power savings
SEE ALSO
cpupower-info(1), cpupower-monitor(1), powertop(1)
AUTHORS
--perf-bias parts written by Len Brown <len.brown@intel.com>
Thomas Renninger <trenn@suse.de>
22/02/2011 CPUPOWER-SET(1)