| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: Smash Childhood Cancer
Thread: SCC: Windows or Linux |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 17
|
|
| Author |
|
|
JEklund2
Advanced Cruncher Finland Joined: Aug 10, 2006 Post Count: 119 Status: Offline Project Badges: 
|
.. which one is more "productive" ( faster and stable ) ?
---------------------------------------- |
||
|
|
Crystal Pellet
Veteran Cruncher Joined: May 21, 2008 Post Count: 1320 Status: Offline Project Badges: 
|
.. which one is more "productive" ( faster and stable ) ? Tasks run times are varying a lot, but I'll give you the figures from the same hardware over the last 30 valid tasks each: Windows: Average 4.963 hours - Minimum 3.39 hours - Maximum 6.11 hours Linux: Average 2.027 hours - Minimum 1.52 hours - Maximum 2.54 hours |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Had asked about this before, the science being VINA. My Q6600 at 1.6ghz on Linux with 4.9.7 kernel per yesterdays data did about 2.3 hours average. The W8.1 on 4770K at 3Ghz did 2.2 hours. Sample size small, but still strongly suggests same applies as all the other VINA sciences.
|
||
|
|
KLiK
Master Cruncher Croatia Joined: Nov 13, 2006 Post Count: 3108 Status: Offline Project Badges:
|
I've not been running SCC exclusively like some...but have a similar MBO with X3360 & Q9550S, so those can easily compare Win10 vs. Ubuntu 14.04! Will post the results in a few days...
----------------------------------------Or check device names DP35DP & DG33FB. Beware that DP35DPM is a Ubuntu 16.04 machine. |
||
|
|
Sid2
Senior Cruncher USA Joined: Jun 12, 2007 Post Count: 259 Status: Offline Project Badges: 
|
No question, SCC runs better on Linux [Mint]
---------------------------------------- |
||
|
|
Falconet
Master Cruncher Portugal Joined: Mar 9, 2009 Post Count: 3295 Status: Offline Project Badges: 
|
My Linux Mint 64-Bit AMD A8-6500 3.5GHZ average from the last 30 tasks: 1.82 CPU hours.
----------------------------------------They got a bit longer in the last 2 days or so. AMD Ryzen 5 1600AF 6C/12T 3.2 GHz - 85W AMD Ryzen 5 2500U 4C/8T 2.0 GHz - 28W AMD Ryzen 7 7730U 8C/16T 3.0 GHz |
||
|
|
hiimebm
Senior Cruncher United States Joined: Oct 19, 2014 Post Count: 305 Status: Offline Project Badges: 
|
Core 2 E7200 @2.53Ghz running Server 2012
----------------------------------------2.59-4.69 Hours.  |
||
|
|
KerSamson
Master Cruncher Switzerland Joined: Jan 29, 2007 Post Count: 1672 Status: Offline Project Badges: 
|
This question has been asked regularly and the answer (fortunately
---------------------------------------- ) remains the same for all WCG projects:
In the past, the efficiency differences between Linux and Windows could have varied a little bit, even if always in Linux favour. Today, the difference is really significant, for example for SCC1 (daily boinc points average):
As long as Windows will not handle memory more efficiently, Linux will be in all cases more efficient. Cheers, Yves |
||
|
|
ca05065
Senior Cruncher Joined: Dec 4, 2007 Post Count: 325 Status: Offline Project Badges:
|
The following comparison work units were run on a dual boot 8Gb I7-2600k at stock speed. It was a small sample of work units (8 linux and 7 Win 10) which were downloaded the same afternoon.
----------------------------------------Average Time (hours) 1.79 (Linux) 2.59 (Win 10) Average Points(Result Status Page) 57.7 (Linux) 68.4 (Win 10) [Edit 1 times, last edit by ca05065 at Feb 5, 2017 12:29:11 PM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Classic, Linux does 2.59 / 1.79 = 1.62 times more work per unit of time, but gets (57.7 / 1.79) / (68.4 / 2.59) = 1.22 times points, but hey we're here for the work, not for the points. BOINC servers are clearly suffering from the same AI as they do at Penn.Ave 1600, Alternate Intelligence.
 |
||
|
|
|