Does Linux generate more heat?

Moderators: Site Moderators, PandeGroup

Re: Does Linux generate more heat?

Postby Stonecold » Fri Feb 17, 2012 1:40 am

bruce wrote:Does the FahCore report the same version number?

What do you mean?
Stonecold
 
Posts: 392
Joined: Sun Dec 25, 2011 9:20 pm

Re: Does Linux generate more heat?

Postby k1wi » Fri Feb 17, 2012 1:41 am

Stonecold wrote:
bruce wrote:Does the FahCore report the same version number?

What do you mean?
I'm not sure who he was asking.
k1wi
 
Posts: 1152
Joined: Tue Sep 22, 2009 10:48 pm

Re: Does Linux generate more heat?

Postby Stonecold » Fri Feb 17, 2012 1:42 am

k1wi wrote:
Stonecold wrote:
bruce wrote:Does the FahCore report the same version number?

What do you mean?
I'm not sure who he was asking.

I wasn't sure either...
Stonecold
 
Posts: 392
Joined: Sun Dec 25, 2011 9:20 pm

Re: Does Linux generate more heat?

Postby bruce » Fri Feb 17, 2012 1:46 am

Stonecold wrote:
bruce wrote:Does the FahCore report the same version number?

What do you mean?


Whenever a WU starts, you'll see something like this in the log (though it's slightly different for each FahCore.
Code: Select all
19:05:23:WU01:FS00:FahCore 0xa4 started
19:05:24:WU01:FS00:0xa4:
19:05:24:WU01:FS00:0xa4:*------------------------------*
19:05:24:WU01:FS00:0xa4:Folding@Home Gromacs GB Core
19:05:24:WU01:FS00:0xa4:Version 2.27 (Dec. 15, 2010)


In this case, I'm running Version 2.27 of FahCore_a4. Those numbers may not be identical for Linux and for Windows. If you have different versions, you've found the difference. If they're the same version, we have to keep looking.

FAH manages which version you have and it will upgrade automatically if it needs to.
bruce
 
Posts: 20833
Joined: Thu Nov 29, 2007 10:13 pm
Location: So. Cal.

Re: Does Linux generate more heat?

Postby Stonecold » Fri Feb 17, 2012 1:49 am

bruce wrote:In this case, I'm running Version 2.27 of FahCore_a4. Those numbers may not be identical for Linux and for Windows. If you have different versions, you've found the difference. If they're the same version, we have to keep looking.

FAH manages which version you have and it will upgrade automatically if it needs to.

Both Windows and Linux are using v2.27 as well.
Stonecold
 
Posts: 392
Joined: Sun Dec 25, 2011 9:20 pm

Re: Does Linux generate more heat?

Postby k1wi » Mon Feb 20, 2012 4:25 am

I have switched over to Linux (in a VM) so will start logging data on temperatures and performance.

Very initial data is hinting at a decrease in TPF for the initial project (6095), from 9:30 on the windows client to 9:00 on the linux client, which by my maths is approximately a 5% increase in performance. At this initial stage, temperatures while running the linux client appear to be in line with those on windows.

[UPDATE]: It does look like the core temperatures are a bit warmer on the Linux VM, the 5% increase is responsible for an increase of about 2 or 3 degrees. Not much but certainly noticeable and this is a highly water cooled computer. I would imagine that if native linux is 10-15% more efficient (and I can see it being so), this computer running native linux could easily be 6-9 degrees warmer than when running the windows client.

I only imagine that a laptop, with it's less effective cooling, would be even more noticeable...
k1wi
 
Posts: 1152
Joined: Tue Sep 22, 2009 10:48 pm

Re: Does Linux generate more heat?

Postby k1wi » Mon Feb 20, 2012 11:33 pm

Below are the latest graphs of my computer's temperatures, outlining the shift from Windows FAH to FAH in a Linux VM on Windows.

The dip at x = ~50 in Graph One reflects the idle time as I switched the client over from Windows to the Linux VM. Immediately you can see the temperatures climb by several degrees, before tailing off gently as the ambient temperature decreases. At X = ~125 the client was shut down to undertake non-FAH related work, before being restarted again at x = ~180 (represented by the sharp climb in temperatures).

The increase in temperatures is more noticable in Graph Two. In Graph Two, x = ~290 represents the switch over to the client in the linux VM, while the first 'dip' at x = ~310 represents the client being shut down. The increase in temperature is fairly noticeable at this point, despite ambient temperatures being similar to those 3 days prior (x = ~150). Switching to running the linux client in a VM is the first time my computer has hit 80 Degrees C in quite a while.

So my interpretation is that there is a temperature increase of around two or three degrees, which, at around a 4% increase in temperature, is in line with the 5% increase in performance.

Image
Graph One
Image
Graph Two
k1wi
 
Posts: 1152
Joined: Tue Sep 22, 2009 10:48 pm

Previous

Return to Linux CPU V6 Client

Who is online

Users browsing this forum: No registered users and 1 guest

cron