compilation shorten Computer life?

[minsoehan]

Eg. if I install Firefox, compilation time takes a bit long. It's ok. No question.
But compilation takes up 52% of my CPU, Intel Core i3-4005U (1.7 GHz, 3MB L3 cache) and 25% of my 8GB Ram.
It is absolutely without any other task.

So the question is that can using Gentoo shorten my computer's life because of compilation than using any other Distros ?

Don't flame me. it is just curious.

[v_andal]

Potentially - yes, practically - impossible to say. I have 15-year old laptop with Gentoo and it still works. My colleague had laptop with Windows and it died after 4 years.

In real life, quite often people replace their PCs because they want newer specs. So, the decisive point is not compilation but quality of device and speed of hardware development.

[Perfect Gentleman]

playing games kills cpu/computer much faster, turning on/off every day too.

[The Doctor]

I'd say the quality of the computer plays a big role there. If you are running the cheapest generic you could find it probably will. If you are running a designer end machine with top notch cooling then it probably won't.

Also, laptop vs desktop. No laptop ever invented has great cooling characteristics. It just doesn't have the space. Desktops can spread out and cool. That makes a huge difference in CPU temps.

Practically it probably matters more how well you take care of your computer, ie dusting it and not dropping it. I've seen a 6 inch drop kill a netbook.
_________________
First things first, but not necessarily in that order.

Apologies if I take a while to respond. I'm currently working on the dematerialization circuit for my blue box.

[asturm]

Yes, don't buy cheap and pay attention to temperatures once in a while. My Thinkpad is still going strong after more than 6 years of Gentoo.

[minsoehan]

[schorsch_76]

@minsoehan: Actually i do it the same like you. Suspend/hibernate created much more trouble to me than the saved time is worth it. With the SSD and gentoo (openrc) it is on X and workable in under 10 sec (including boot password).

About the lifetime of computers (especially CPUs), there i read an intersting "hack" how to get CPUs slower over time.
Sorry, link is german
http://www.heise.de/newsticker/meldung/Boesartige-Software-kann-Prozessoren-schneller-altern-lassen-2853329.html
_________________
// valid again: I forgot about the git access. Now 1.2GB big. Start: 2015-06-25
git daily portage tree
Web: https://github.com/schorsch1976/portage
git clone https://github.com/schorsch1976/portage

[Perfect Gentleman]

when you turn on computer there are transient events which are very harmful for CPU and other equipment

[khayyam]

[Naib]

*sigh*...

right the real damage to CPU's and components are either

1) overtemperature. sort out your cooling ...
2) Thermal stress. Rapid high->low->high->low ... Could turning off the computer, turning it back on & compiling... turning it off... etc be enough to accelerate its aging? (bondwire thermal stress)... depending on how fast it cools...
3) Poor handling at build time or clean (DO NOT USE A VACUUM!!!). ESD is a real issue not some fairy tail spun by the electronics industry (here is my EE.SO link: https://electronics.stackexchange.com/questions/93293/wood-workbench-as-esd-protection/159107#159107 )
4) really cheap parts...
_________________

[khayyam]

[Naib]

[albright]

[NeddySeagoon]

minsoehan,

It a well established fact in electronics reliability that a 10 deg C rise is the operating temperature of electronic equipment doubles the random failure rate.
Two questions arise from that.
a) Will the equipment be obsolete before it fails (so you don't care)
b) Will the equipment fail beforehand from non random causes.

Non random failures are difficult to predict. There are well known non random effects that equipment designers try to minimise. e.g. gold embrittlement of solder joints. The pads on printed circuit boards are often covered with gold, as this prevents oxides forming which make the pads difficult to solder.
The down side is that in the finished solder joint, the gold contend must be very low, so that the gold embrittlement problem is avoided.

Another is temperature cycling. This causes repeated mechanical stress on parts due to differential expansion/contraction as parts heat and cool.
The classic example is that light bulbs and vacuum tube heaters normally fail at switch on. This effect is seen in other electronic equipment too.

All the above are statistical truths and statistics only work with reasonable sample sizes.
They cannot be applied to a sample size of one.

Summing up, for your particular system, the answer is a definite maybe.
_________________
Regards,

NeddySeagoon

Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail.

[Naib]

I would say the bigger problem with commercial electronics is tin wiskers (goo fscking EU and their short sightedness... ) & their associated failures rather than all these other stuff...
_________________

[desultory]

[yagami]

Well, one trick I do is to configure portage make.conf to only use half my cpu cores.

If for overnight compilation, I even "downgrade" further.

For example, I have an i7 with 8 cores. During day, I compile with -j4, and at night I use -j2.

This has two upsides: Cooler temp's and makes my system completly responsive while compiling !

EDIT: s/"not only use halt my cpu"/"not only use half my cpu"

[Cyker]

I used to have PORTAGE_NICENESS=20 set in my make.conf which meant it ran at the idle CPU frequency. Good for temps and power usage but took so long to compile I took it out

(NB: this does require e.g. /sys/devices/system/cpu/cpufreq/ondemand/ignore_nice_load to be set to "1"!)

[JDisaster]

If you have a SSD you probably want to setup a tmpfs for portage this will lower the number of Read/Writes on your solid state drive.

Here is a guide

https://wiki.gentoo.org/wiki/Portage_TMPDIR_on_tmpfs

[Ant P.]

...or just leave things as they are and enable laptop_mode for mostly the same effect, except that any working set that doesn't fit in the VFS cache won't repeatedly hammer the tiny fixed area of the drive assigned to swap.

[1clue]

Really, as long as the system gets rid of heat adequately I don't think there's much difference with respect to CPU or other solid state components.

A storage device, whether it spins or is an SSD, has a finite number of writes that can be done to any sector before damage occurs. It benefits both speed and longevity of any drive to set up tmpfs if you have the RAM for it. Time powered on for a storage device is probably a bad thing as a general rule.

[Buffoon]

Once upon time I purchased a new used refurbished T23, 1.2 GHz, 2 GB of RAM, a real monster ... err, it was A. D. 2004. Alright, I had to run some errands, took it into my car and started emerging KDE. Arrived somewhere, locked up the car in a hot summer day in Rhode Island. Guess what, when I got back a few hours later my brand new used refurbished T23 was dead. Dead!

Draw your own conclusions.

[ct85711]

Even then, it wasn't necessarily because of you compiling KDE. You could have been running windows, and just as well fried your laptop. You could easily just take a brand new laptop, lock it in an enclosed room that is 100+ with the laptop running, and see how long it lives too. Sure my last laptop, which I just happen to also fry, went out because it wasn't able to handle to the heat load for extended periods of time. The only reason why it lived as long for me, is that I underclocked that laptop to the bare minimal amount to turn the CPU on just to get a little bit more out of that laptop.

[Cyker]

[1clue]

@Buffoon,

Do you somehow suspect that, had your laptop been in an air conditioned office, it would have still died an early death? Or that if you had not been compiling but left it running in the car, it wouldn't have?

Clearly it's not the compiling that killed your laptop, it's the heat.