various systems and credit

Message boards : Number crunching : various systems and credit

To post messages, you must log in.

AuthorMessage
AMusingFool

Send message
Joined: 16 May 06
Posts: 5
Credit: 17,168,068
RAC: 0
Message 52311 - Posted: 7 Apr 2008, 16:45:27 UTC

So, I've only been running things (I think, despite what it says about my start time) since the new credit system went into effect, so I think that's not part of the issue.

However, the first machine I put this on was a windows XP laptop. The units I was running were generally around 10k seconds time, claiming 40-45 credits per unit, and being granted 10-15% more credits than that per unit.

The next machine was a MacBook Pro, with a CPU only a hair slower than the XP box (2.16GHz vs 2.33GHz of the same make and model CPU). It gets around 15-20 credits claimed per unit (of ~10k second units), and is granted 10-15% less credits per unit.

I was wondering if there was a heavy windows bias, just based on that, but now I know there definitely is. The reason is that I installed ubuntu on the XP box, and it's now getting 25-30 credits per unit (but still being granted less than that).

So it would appear that the windows client is faster than the linux client, which is faster than the mac client.

So I was just wondering if anyone in the know would like to comment on that. (All of this is done with the generic client on each machine, if anyone's wondering.)

Thanks.
ID: 52311 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile dcdc

Send message
Joined: 3 Nov 05
Posts: 1829
Credit: 116,848,329
RAC: 78,241
Message 52319 - Posted: 7 Apr 2008, 19:33:47 UTC
Last modified: 7 Apr 2008, 19:35:31 UTC

the T2600 is a Core Duo (Yonah) CPU whereas the T7600 is a Core2 Duo (Conroe). Conroe is quite a bit more efficient (although they're both good CPUs):

http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=2808&p=6

There are differences between the different OS's but I don't think they're particularly profound. Maybe it was under more load or getting more cache-swapping when running the graphics-intensive OSX? XP is possibly better optimised as the compilers have been around longer...

HTH
Danny
ID: 52319 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Path7

Send message
Joined: 25 Aug 07
Posts: 128
Credit: 61,751
RAC: 0
Message 52320 - Posted: 7 Apr 2008, 20:20:46 UTC
Last modified: 7 Apr 2008, 20:21:36 UTC

Hello AmusingFool,
A few weeks ago I installed Ubuntu as a dual-boot, next to my Windows XP. Boinc on Linux was (at first) crunching much less credits than XP. After checking out some other threads (eg https://boinc.bakerlab.org/rosetta/forum_thread.php?id=3593) I realized my AMD 3000+ didn't run its normal 1.8 G Hz but slowed down to 1 G Hz because the low priority R@h appication didn't urge Linux to go faster.
Now My single core machine crunches more credit running Ubuntu than XP (in the same period of time).

Well, this is of course one point of view, this might be different on a multi-core machine.

Have a nice day,
Path7.
ID: 52320 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
j2satx

Send message
Joined: 17 Sep 05
Posts: 97
Credit: 3,670,592
RAC: 0
Message 52334 - Posted: 8 Apr 2008, 22:07:12 UTC - in response to Message 52320.  

Hello AmusingFool,
A few weeks ago I installed Ubuntu as a dual-boot, next to my Windows XP. Boinc on Linux was (at first) crunching much less credits than XP. After checking out some other threads (eg https://boinc.bakerlab.org/rosetta/forum_thread.php?id=3593) I realized my AMD 3000+ didn't run its normal 1.8 G Hz but slowed down to 1 G Hz because the low priority R@h appication didn't urge Linux to go faster.
Now My single core machine crunches more credit running Ubuntu than XP (in the same period of time).

Well, this is of course one point of view, this might be different on a multi-core machine.

Have a nice day,
Path7.


Make sure "Cool n Quiet" is disabled in BIOS and try the Linux again.
ID: 52334 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
AMusingFool

Send message
Joined: 16 May 06
Posts: 5
Credit: 17,168,068
RAC: 0
Message 52365 - Posted: 10 Apr 2008, 15:55:19 UTC

Thanks for the pointers. I found out that the CPU was running at far less than top speed (less than half, actually)... After some poking around, I discovered how to fix that. I also poked around with renice-ing boinc. It didn't do much until I renice'd the actual rosetta applications themselves (since they default to a nice level of 19, does anyone who knows python know how to change that?). I still need to see if renice'ing will do anything on the macs.
ID: 52365 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Greg_BE
Avatar

Send message
Joined: 30 May 06
Posts: 5664
Credit: 5,711,666
RAC: 764
Message 52368 - Posted: 10 Apr 2008, 19:48:06 UTC

How much does that cool'n'quiet take out of your cpu speed?
I have a 64 2800+ runing win xp sp 2.
ID: 52368 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile dcdc

Send message
Joined: 3 Nov 05
Posts: 1829
Credit: 116,848,329
RAC: 78,241
Message 52370 - Posted: 10 Apr 2008, 20:30:12 UTC - in response to Message 52368.  

How much does that cool'n'quiet take out of your cpu speed?
I have a 64 2800+ runing win xp sp 2.

it reduces the clock speed dependant upon the CPU load, but depends upon the OS as to whether low priority tasks are considered as load...
ID: 52370 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
j2satx

Send message
Joined: 17 Sep 05
Posts: 97
Credit: 3,670,592
RAC: 0
Message 52374 - Posted: 10 Apr 2008, 21:42:24 UTC - in response to Message 52368.  

How much does that cool'n'quiet take out of your cpu speed?
I have a 64 2800+ runing win xp sp 2.


It typically drops the CPU to 1GHz on my puters.
ID: 52374 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
j2satx

Send message
Joined: 17 Sep 05
Posts: 97
Credit: 3,670,592
RAC: 0
Message 52375 - Posted: 10 Apr 2008, 21:45:24 UTC - in response to Message 52368.  

How much does that cool'n'quiet take out of your cpu speed?
I have a 64 2800+ runing win xp sp 2.


It typically drops the CPU to 1GHz on my puters.
ID: 52375 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
AMusingFool

Send message
Joined: 16 May 06
Posts: 5
Credit: 17,168,068
RAC: 0
Message 52526 - Posted: 16 Apr 2008, 17:00:29 UTC

Just in case anyone cares, renice'ing the rosetta programs in linux seems to've made a very large difference in credit. However, it seems not to've made any difference at all under OSX (intel or PPC). Kind of weird.
ID: 52526 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Gen_X_Accord
Avatar

Send message
Joined: 5 Jun 06
Posts: 154
Credit: 279,018
RAC: 0
Message 61423 - Posted: 28 May 2009, 7:03:10 UTC

Here's an interesting thing. While I disabled the Cool n Quiet feature after putting in my new motherboard, processor and new memory, I decided to activate it and see what it did when I ran Rosetta. Now I have XP sp3, a new AMD X2 7850, and DDR2 memory that I only run at 800mhz because dual channel isn't supported at 1066mhz. After setting it all up, the Cool N Quiet will let the processor run at full speed when Rosetta is running, which is 2800 mhz, and only idles it back when I put Rossetta on snooze, to about 1400mhz. I tested it by changing the power scheme to home/office, which seems to disable Cool N Quiet without having to go all the way back into the Bios, and when I put Rosetta on snooze, the processor speed stays the same, even if the processor isn't under load anymore. I think I will leave it on, for the strange occasions that Rosetta stops and I'm not around, or my kids messes with something, at least it will save me a little electricity.
I will also monitor my credit to see if it takes a hit, but it seems that CnQ will let Boinc run Rosetta full blast for my computer.
ID: 61423 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote

Message boards : Number crunching : various systems and credit



©2024 University of Washington
https://www.bakerlab.org