Do modern laptops really run this hot?

Talk about just about anything else that is non-gaming here, but keep it clean
User avatar
neilencio
64-bit
Posts: 386
Joined: Thu Aug 06, 2009 12:13 am
Location: Philippines
Contact:

Re: Do modern laptops really run this hot?

Post by neilencio »

Thanks for the in depth reply, pakopako.
pakopako wrote:
neilencio wrote:Any info on the pros and cons of undervolting?
Under-clocking is never bad. Especially software under-clocking.

I won't make claims it will lengthen the life of your computer -- I really don't know -- I do know that because you're restricting your computer from pushing itself harder, you extend your battery life and reduce heat output. (See my previous post about "throttling down".)
I'm not sure I can express it well, but it's not that I'm trying to make it live way beyond the expected shelf life, I just want to reduce the stress it gets from heat (which reduces the life of the computer below normal, if the heat is consistently above average).
You will, of course, take a hit in productivity -- flash games or movies that would normally require your machine to run at-or-near full-speed will stutter if you shunt your system down to half-speed. (Of course programs that only require half of your processing power won't notice a thing.)
In my case, I didn't actually half the clock, a couple of hundred megahertz didn't really slow my computer down that much, especially since I don't really use it for resource intensive tasks. It can still play hd movies and flash smoothly (especially since the IGP is much, much better than Intel's crappy GMA)
As I said, I under-volted a few of my systems. The one I wanted to get the most out of was the one that didn't take to it very well, but the others had minor improvements to battery life. Maybe an extra 10-15%, or roughly 20 minutes, of battery life and maybe 3~5°C overall.
This is the result I was going for
RyaNtheSlayA wrote:Unless you're pushing the upper limits of your CPU temp wise, I don't think it will make it last longer. Besides, I doubt you'll have that same laptop 10 years down the road.
I was talking with wootcube about that a few months ago. He has a good system of not being tied down to a single machine for too long, which is great.

I, however, drag the hell out of whatever I get my hands on. Maybe it's the "car" mentality (that each time you turn the ignition, the price keeps cutting itself in half), but I've been able to tweak and refine and tune-up decade-plus machines. I wouldn't recommend this if you can help yourself, but it is possible.
I'm on the same boat, at least, in the context of laptops. Laptops, IMO, should be taken care of and held on to for a lot longer than desktops (and at least in my country, there are people still using old Pentium III laptops, and stores selling refurbs of said models), primarily because they cost a hell of a lot more than a similarly specced Desktop, and you don't have the option of changing and upgrading the parts willy nilly. I go upgrade crazy on my desktop, but I treat my laptop as a long-term purchase. I wouldn't go so far as 10 years (but I don't mind it living that long), but I want it to last as much as possible and the need for replacement and repair as far away as possible.
neilencio wrote:I use VM for work, and it always maxes out one core. my average temps during use is on the high 60s-low 70s. I think that's a bit stressful for the cpu, considering I work 8-12 hrs a day nonstop.
Missed this part.
Because you're running the VM 8-12 hours, you're not really going to "extend the life" of your machine -- as a work-machine it is going to RROD faster than most people's personal machines, you just want it to not die as quickly.

As for consuming the entire core, have you tried under-clocking/speed throttling (down) the system? It will still using 100% of the core, but if you don't take a hit in productivity, then that's just inefficient coding on the part of the VM (to max out the processor without delivering additional performance).
I actually got this fixed. Turns out i've been using an SVN of virtualbox that had resource management issues. I used a different revision that fixed the bug. Also, I found a setting that allows me to limit the amount of cpu cycles the virtual machine is allowed to use. It runs like peaches now without turning into a cpu stress test (temps are still a bit higher than if I were running programs natively, but I understand that it's the price of virtualization. I can't have my cake and eat it too)

OT, but one neat thing about running VMs is that I usually need to run several automated scripts and programs at the same time. If I do it natively, my whole system crawls and I can't use it for other things. But If I use a VM, the scripts and programs slow down the VM, while the main system still runs well and I can use it normally.
Visit my wife's blog: Pinay Geek
pakopako
Next-Gen
Posts: 1654
Joined: Wed Jul 07, 2010 2:29 pm

Re: Do modern laptops really run this hot?

Post by pakopako »

neilencio wrote:I'm not sure I can express it well, but it's not that I'm trying to make it live way beyond the expected shelf life, I just want to reduce the stress it gets from heat (which reduces the life of the computer below normal, if the heat is consistently above average).
Ah. I undervolt/underclock/throttle-down because I like holding my laptop. Holding something with a surface temperature close to the boiling point of water is not fun. (...I have stories about cooking bacon with one my old machines...)
neilencio wrote:(... there are people still using old Pentium III laptops, and stores selling refurbs of said models), primarily because they cost a hell of a lot more than a similarly specced Desktop, and you don't have the option of changing and upgrading the parts willy nilly.
I met a GameFAQs poster last week who bought a P3 laptop for $20 as a dedicated DOS machine. (Because their netbook couldn't handle tasking Skype and DOSbox and Productivity at the same time.) Worked out pretty well.
neilencio wrote:Turns out i've been using an SVN of virtualbox that had resource management issues. I used a different revision that fixed the bug. Also, I found a setting that allows me to limit the amount of cpu cycles the virtual machine is allowed to use. It runs like peaches now without turning into a cpu stress test
8)

I really should learn how to use VirtualBox... but I actually have such a dearth of (outdated) hardware, I can run programs in their native environment at the cost of a HDD install and boot-up time.
My scheduling skills have died of dysentery; I hope to visit at least on a monthly basis.
Still, don't forget to tip your waitress.
Post Reply