Project: 6012 (Run 0, Clone 32, Gen 168)

bollix47 · Post by **bollix47** » Mon Jan 25, 2010 12:37 pm

FYI

Running on E6600 @ stock of 2.4Ghz using notfred's distro.

Was getting frame times of 1hr12min for ~94 PPD and definitely not going to make preferred deadlines(i.e. no bonus and WU would be re-issued).

Logged into console and ran top to find that only one instance of the core was running. Killed the fah process and restarted from command line. Times dropped to 15min for around 436PPD. With bonus it will still be well below the PPD of the a2 core WUs that run on this computer which is around 2000PPD.

When notfred's starts it has the arguments:

Code: Select all

Arguments: -local -forceasm -smp 4

When I restarted I used:

Code: Select all

Arguments: -verbosity 9 -smp

Not sure if the arguments had anything to do with the problem but the resulting decrease in frame times is obvious.

Code: Select all

[10:11:33] Folding@Home Gromacs SMP Core
[10:11:33] Version 2.13 (Tue Aug 18 18:28:33 CEST 2009)
[10:11:33] 
[10:11:33] Preparing to commence simulation
[10:11:33] - Ensuring status. Please wait.
[10:11:43] - Assembly optimizations manually forced on.
[10:11:43] - Not checking prior termination.
[10:11:43] - Expanded 1796466 -> 2078149 (decompressed 115.6 percent)
[10:11:43] Called DecompressByteArray: compressed_data_size=1796466 data_size=2078149, decompressed_data_size=2078149 diff=0
[10:11:43] - Digital signature verified
[10:11:43] 
[10:11:43] Project: 6012 (Run 0, Clone 32, Gen 168)
[10:11:43] 
[10:11:43] Assembly optimizations on if available.
[10:11:43] Entering M.D.
[10:11:49] Using Gromacs checkpoints
[10:12:26] Resuming from checkpoint
[10:12:26] Verified work/wudata_07.log
[10:12:26] Verified work/wudata_07.trr
[10:12:26] Verified work/wudata_07.edr
[10:12:27] Completed 8222 out of 500000 steps  (1%)
[10:39:33] Completed 10000 out of 500000 steps  (2%)


*******  Restarted using command line. *******


--- Opening Log file [January 25 11:00:19 UTC] 


# Linux SMP Console Edition ###################################################
###############################################################################

                       Folding@Home Client Version 6.29

                          http://folding.stanford.edu

###############################################################################
###############################################################################

Launch directory: /etc/folding/1
Executable: ./fah6
Arguments: -verbosity 9 -smp 

[11:00:19] - Ask before connecting: No
[11:00:19] - User name: bollix47 (Team 39340)
[11:00:19] - User ID: 1988EAD83989163D
[11:00:19] - Machine ID: 1
[11:00:19] 
[11:00:20] Loaded queue successfully.
[11:00:20] 
[11:00:20] - Autosending finished units... [January 25 11:00:20 UTC]
[11:00:20] + Processing work unit
[11:00:20] Trying to send all finished work units
[11:00:20] Core required: FahCore_a3.exe
[11:00:20] + No unsent completed units remaining.
[11:00:20] Core found.
[11:00:20] - Autosend completed
[11:00:20] Working on queue slot 07 [January 25 11:00:20 UTC]
[11:00:20] + Working ...
[11:00:20] - Calling './FahCore_a3.exe -dir work/ -nice 19 -suffix 07 -np 2 -checkpoint 30 -verbose -lifeline 1553 -version 629'

[11:00:20] 
[11:00:20] *------------------------------*
[11:00:20] Folding@Home Gromacs SMP Core
[11:00:20] Version 2.13 (Tue Aug 18 18:28:33 CEST 2009)
[11:00:20] 
[11:00:20] Preparing to commence simulation
[11:00:20] - Looking at optimizations...
[11:00:20] - Files status OK
[11:00:20] - Expanded 1796466 -> 2078149 (decompressed 115.6 percent)
[11:00:20] Called DecompressByteArray: compressed_data_size=1796466 data_size=2078149, decompressed_data_size=2078149 diff=0
[11:00:20] - Digital signature verified
[11:00:20] 
[11:00:20] Project: 6012 (Run 0, Clone 32, Gen 168)
[11:00:20] 
[11:00:20] Assembly optimizations on if available.
[11:00:20] Entering M.D.
[11:00:26] Using Gromacs checkpoints
[11:00:27] Resuming from checkpoint
[11:00:27] Verified work/wudata_07.log
[11:00:27] Verified work/wudata_07.trr
[11:00:27] Verified work/wudata_07.edr
[11:00:27] Completed 10252 out of 500000 steps  (2%)
[11:15:10] Completed 15000 out of 500000 steps  (3%)
[11:30:40] Completed 20000 out of 500000 steps  (4%)
[11:46:10] Completed 25000 out of 500000 steps  (5%)

Unfortunately notfred doesn't keep a previous log so the above does not show a complete step for the 1hr12min frame times but can easily be prorated.

Hyperlife · Post by **Hyperlife** » Mon Jan 25, 2010 4:11 pm

The top command will only show one instance of Core A3 if you don't view threads. If the %CPU used was hovering around 200%, then you're OK. You can see the separate core threads in top if you type "H" while it's running.

The frame times look OK after you restarted. Not sure what caused the difference.

weedacres · Post by **weedacres** » Mon Jan 25, 2010 5:05 pm

I have 3 P6102's running. One on a Q6600 and the other two on dual core machines. All on VMWare & linuxfah's slackware client.
The Q6600 is running fine at 5300ppd, about what I'd expect from most A2's.

The 2 core machines were down around 120ppd when they normally run 1600-1900 ppd on A2's. I changed from -smp 4 to -smp and that seems to have helped. I'm now up to around 1100-1200 ppd. Not up to A2 standards, closer to A1 for these machines.

_r2w_ben · Post by **_r2w_ben** » Mon Jan 25, 2010 5:43 pm

Here's my guess:

-smp 4 overrode the automatic configuration of the number of threads to run in the new A3 core.
4 threads on a dual core results in poor performance because of the context switching involved.
Two threads shared each CPU core and execution went something like this: Load one thread, perform a few computations, unload, load the other thread, perform a few computations, unload...

Removing the '4' override made the client launch the appropriate 2 threads (one per core) and increased performance.

The initial launch string is missing from bollix47's log so I can't confirm my theory. However, this line in the log shows -np (number of processors) as 2 after the change:

Code: Select all

Calling './FahCore_a3.exe -dir work/ -nice 19 -suffix 07 -np 2 -checkpoint 30 -verbose -lifeline 1553 -version 629'

weedacres · Post by **weedacres** » Mon Jan 25, 2010 6:05 pm

_r2w_ben wrote:Here's my guess:

-smp 4 overrode the automatic configuration of the number of threads to run in the new A3 core.
4 threads on a dual core results in poor performance because of the context switching involved.
Two threads shared each CPU core and execution went something like this: Load one thread, perform a few computations, unload, load the other thread, perform a few computations, unload...

Removing the '4' override made the client launch the appropriate 2 threads (one per core) and increased performance.

The initial launch string is missing from bollix47's log so I can't confirm my theory. However, this line in the log shows -np (number of processors) as 2 after the change:
Code: Select all
Calling './FahCore_a3.exe -dir work/ -nice 19 -suffix 07 -np 2 -checkpoint 30 -verbose -lifeline 1553 -version 629'

Sounds reasonable.

Here's a log on a 2 core machine, prior to removing the -4. I shut it down after 50 min with no update.

Code: Select all

--- Opening Log file [January 25 14:15:52 UTC] 


# Linux SMP Console Edition ###################################################
###############################################################################

                       Folding@Home Client Version 6.29

                          http://folding.stanford.edu

###############################################################################
###############################################################################

Launch directory: /usr/local/fah
Executable: ./fah6
Arguments: -smp 4 -verbosity 9 

[14:15:52] - Ask before connecting: No
[14:15:52] - User name: Porky (Team 52523)
[14:15:52] - User ID: 651D58F87A25D954
[14:15:52] - Machine ID: 1
[14:15:52] 
[14:15:52] Loaded queue successfully.
[14:15:52] 
[14:15:52] - Autosending finished units... [January 25 14:15:52 UTC]
[14:15:52] Trying to send all finished work units
[14:15:52] + Processing work unit
[14:15:52] + No unsent completed units remaining.
[14:15:52] - Autosend completed
[14:15:52] Core required: FahCore_a3.exe
[14:15:52] Core found.
[14:15:53] Working on queue slot 03 [January 25 14:15:53 UTC]
[14:15:53] + Working ...
[14:15:53] - Calling './FahCore_a3.exe -dir work/ -nice 19 -suffix 03 -np 4 -checkpoint 15 -verbose -lifeline 836 -version 629'

[14:15:53] 
[14:15:53] *------------------------------*
[14:15:53] Folding@Home Gromacs SMP Core
[14:15:53] Version 2.13 (Tue Aug 18 18:28:33 CEST 2009)
[14:15:53] 
[14:15:53] Preparing to commence simulation
[14:15:53] - Ensuring status. Please wait.
[14:16:02] - Looking at optimizations...
[14:16:02] - Working with standard loops on this execution.
[14:16:03] - Previous termination of core was improper.
[14:16:03] - Files status OK
[14:16:03] - Expanded 1797587 -> 2078149 (decompressed 115.6 percent)
[14:16:03] Called DecompressByteArray: compressed_data_size=1797587 data_size=2078149, decompressed_data_size=2078149 diff=0
[14:16:03] - Digital signature verified
[14:16:03] 
[14:16:03] Project: 6012 (Run 0, Clone 121, Gen 50)
[14:16:03] 
[14:16:03] Entering M.D.
[14:16:09] Using Gromacs checkpoints
[14:16:29] Resuming from checkpoint
[14:16:29] Verified work/wudata_03.log
[14:16:29] Verified work/wudata_03.trr
[14:16:29] Verified work/wudata_03.edr
[14:16:30] Completed 22968 out of 500000 steps  (4%)
[14:36:34] Completed 25000 out of 500000 steps  (5%)
[15:25:22] ***** Got an Activate signal (2)
[15:25:22] Killing all core threads

Folding@Home Client Shutdown.

Here's the log after changing to -smp only.

Code: Select all

# Linux SMP Console Edition ###################################################
###############################################################################

                       Folding@Home Client Version 6.29

                          http://folding.stanford.edu

###############################################################################
###############################################################################

Launch directory: /usr/local/fah
Executable: ./fah6
Arguments: -smp -verbosity 9 

[15:26:37] - Ask before connecting: No
[15:26:37] - User name: Porky (Team 52523)
[15:26:37] - User ID: 651D58F87A25D954
[15:26:37] - Machine ID: 1
[15:26:37] 
[15:26:37] Loaded queue successfully.
[15:26:37] 
[15:26:37] - Autosending finished units... [January 25 15:26:37 UTC]
[15:26:37] + Processing work unit
[15:26:37] Trying to send all finished work units
[15:26:37] Core required: FahCore_a3.exe
[15:26:37] + No unsent completed units remaining.
[15:26:37] Core found.
[15:26:37] - Autosend completed
[15:26:37] Working on queue slot 03 [January 25 15:26:37 UTC]
[15:26:37] + Working ...
[15:26:37] - Calling './FahCore_a3.exe -dir work/ -nice 19 -suffix 03 -np 2 -checkpoint 15 -verbose -lifeline 883 -version 629'

[15:26:37] 
[15:26:37] *------------------------------*
[15:26:37] Folding@Home Gromacs SMP Core
[15:26:37] Version 2.13 (Tue Aug 18 18:28:33 CEST 2009)
[15:26:37] 
[15:26:37] Preparing to commence simulation
[15:26:37] - Looking at optimizations...
[15:26:37] - Files status OK
[15:26:37] - Expanded 1797587 -> 2078149 (decompressed 115.6 percent)
[15:26:37] Called DecompressByteArray: compressed_data_size=1797587 data_size=2078149, decompressed_data_size=2078149 diff=0
[15:26:37] - Digital signature verified
[15:26:37] 
[15:26:37] Project: 6012 (Run 0, Clone 121, Gen 50)
[15:26:37] 
[15:26:37] Assembly optimizations on if available.
[15:26:37] Entering M.D.
[15:26:44] Using Gromacs checkpoints
[15:26:45] Resuming from checkpoint
[15:26:45] Verified work/wudata_03.log
[15:26:45] Verified work/wudata_03.trr
[15:26:45] Verified work/wudata_03.edr
[15:26:46] Completed 28848 out of 500000 steps  (5%)
[15:30:38] Completed 30000 out of 500000 steps  (6%)
[15:47:25] Completed 35000 out of 500000 steps  (7%)
[16:04:13] Completed 40000 out of 500000 steps  (8%)
[16:21:04] Completed 45000 out of 500000 steps  (9%)
[16:38:08] Completed 50000 out of 500000 steps  (10%)

The PPD improvement I mentioned above was with bonus points calculated, which of course I won't see for awhile. Real ppd is about 400+ on the 2 core machines and 1000+ on the quads.

_r2w_ben · Post by **_r2w_ben** » Mon Jan 25, 2010 6:20 pm

Thanks for the logs.
The first one shows -np 4 so it probably was spawning 4 threads.

ikerekes · Post by **ikerekes** » Mon Jan 25, 2010 6:24 pm

here is some result of how a3 is faring on my systems (all of them 8.04 ubuntu native smp, with one gpu2 client running parallel under wine):

Q6600 @3.3Ghz
Index 5: finished 1920.00 pts (231.031 pt/hr) 8.66 X min speed
Index 6: finished 1920.00 pts (231.047 pt/hr) 8.66 X min speed
Index 7: finished 470.00 pts (45.589 pt/hr) 14 X min speed
Index 8: folding now 470.00 pts (44.011 pt/hr) 13.5 X min speed; 53% complete

Q8400 @3.3Ghz
Index 1: finished 1920.00 pts (241.450 pt/hr) 9.05 X min speed
Index 2: finished 1920.00 pts (241.737 pt/hr) 9.07 X min speed
Index 3: finished 380.00 pts (48.807 pt/hr) 9.25 X min speed
Index 4: folding now 470.00 pts (43.573 pt/hr) 13.4 X min speed; 29% complete

E5200 @3.65
Index 2: finished 1920.00 pts (122.740 pt/hr) 4.6 X min speed
.
.
Index 8: folding now 484.00 pts (22.801 pt/hr) 6.78 X min speed; 37% complete

E5200-2 @3.75Ghz
Index 7: finished 1920.00 pts (124.040 pt/hr) 4.65 X min speed
.
.
Index 2: folding now 470.00 pts (23.467 pt/hr) 7.19 X min speed; 76% complete

A-620 quad@2.6Ghz
Index 1: finished 1920.00 pts (172.468 pt/hr) 6.47 X min speed
Index 2: folding now 470.00 pts (30.446 pt/hr) 9.33 X min speed; 88% complete

AMD X2 6000+ @3.0Ghz
Index 9: finished 1920.00 pts (79.633 pt/hr) 2.99 X min speed
Index 0: finished 1920.00 pts (77.187 pt/hr) 2.89 X min speed
.
.
Index 7: finished 1920.00 pts (73.617 pt/hr) 2.76 X min speed
Index 8: folding now 470.00 pts (15.139 pt/hr) 4.64 X min speed; 51% complete

in general all the systems (quad or dual) lost about 80% in ppd

Hyperlife · Post by **Hyperlife** » Mon Jan 25, 2010 6:28 pm

ikerekes wrote:in general all the systems (quad or dual) lost about 80% in ppd

You're not including the bonus calculation. Read the SMP2 Points FAQ and use a program like HFM.NET that can calculate the bonus.

ikerekes · Post by **ikerekes** » Mon Jan 25, 2010 6:42 pm

Hyperlife wrote:
ikerekes wrote:in general all the systems (quad or dual) lost about 80% in ppd
You're not including the bonus calculation. Read the SMP2 Points FAQ and use a program like HFM.NET that can calculate the bonus.

The bonus points not clicking in until at least 10 bonus eligible unit has been returned. With this rate it will take weeks and weeks to return 10 units.
HFM.NET doesn't run on Linux and I don't have windows installed on any of my machines.
If I am loosing 80% of my productivity, how can the k=2 compensate for it?

Hyperlife · Post by **Hyperlife** » Mon Jan 25, 2010 7:05 pm

ikerekes wrote:The bonus points not clicking in until at least 10 bonus eligible unit has been returned. With this rate it will take weeks and weeks to return 10 units.

No it won't. A Q6600 at 3.3GHz can finish 10 WUs in less than 5 days. In addition, the minimum is 10 WUs per username/passkey, not machine, so as long as all of your folders work on A3 units, you'll hit 10 in a couple of days. You've already completed a few WUs already -- as long as you entered your passkey on all your rigs, you're around halfway to 10 right now. If those earlier units are A2s (and they look like they are), then you've already finished 10.

ikerekes wrote:HFM.NET doesn't run on Linux and I don't have windows installed on any of my machines.
If I am loosing 80% of my productivity, how can the k=2 compensate for it?

Do the math and you'll find out (or read the FAQ I posted above and follow the sample calculation). You're not losing 80% of your productivity -- you're just as productive as you were before. However, the majority of your points will now come from the bonus calculation.

ikerekes · Post by **ikerekes** » Mon Jan 25, 2010 7:53 pm

Hyperlife wrote:
ikerekes wrote:The bonus points not clicking in until at least 10 bonus eligible unit has been returned. With this rate it will take weeks and weeks to return 10 units.
No it won't. A Q6600 at 3.3GHz can finish 10 WUs in less than 5 days. In addition, the minimum is 10 WUs per username/passkey, not machine, so as long as all of your folders work on A3 units, you'll hit 10 in a couple of days. You've already completed a few WUs already -- as long as you entered your passkey on all your rigs, you're around halfway to 10 right now. If those earlier units are A2s (and they look like they are), then you've already finished 10.

Thank you for the clarification Hyperlife!

I have to admit, since I am new to the bonus system (I never had a powerful enough machine to participate in the bigadv program) I thought that the bonus eligibility tied to the machine not only to the username/passkey combo.
That bring up an other question though. If I have a machine which is powerful enough to return an a2 in lets say 28-30 hours but might not be able to return an a3 unit in the preferred time period, how I keep it out from the a3 program, not to jeopardize my >80% return requirement.
(I have a 2Ghz lappy, with failed HD, so I am running notfred on it still brings in about 2 wu's in 3 days).

may be these clarifications and questions belongs to a separate thread, all the smp2 deserves his own board

PS. My sincere apologies that I gave PPD numbers instead of the elapsed times (the ppd numbers were readily available from qd).
That gave the wrong idea. I couldn't care less about the actual points (I have more than 13 million) but I do care about the number of wu's returned and how fast I can return them.

Hyperlife · Post by **Hyperlife** » Mon Jan 25, 2010 8:06 pm

ikerekes wrote:That bring up an other question though. If I have a machine which is powerful enough to return an a2 in lets say 28-30 hours but might not be able to return an a3 unit in the preferred time period, how I keep it out from the a3 program, not to jeopardize my >80% return requirement.
(I have a 2Ghz lappy, with failed HD, so I am running notfred on it still brings in about 2 wu's in 3 days).

For now, you can remove the -advmethods flag to avoid A3 WUs. (In the future, A3s may be sent to all SMP folders, though, so this may be a temporary solution only.) However, A3 WUs should fold in about the same time as the A2 WUs, so if your A2 WUs are meeting the preferred deadlines, the A3s should meet them as well.

ikerekes · Post by **ikerekes** » Mon Jan 25, 2010 9:21 pm

Hyperlife wrote: However, A3 WUs should fold in about the same time as the A2 WUs, so if your A2 WUs are meeting the preferred deadlines, the A3s should meet them as well.

Yes but they are not! that was my whole point, sorry to confusing the issue with the ppd numbers.

Maybe this will explain it better:

Q6600 @3.3Ghz
Index 6: finished 1920.00 pts (231.047 pt/hr) 8.66 X min speed
server: 171.67.108.24:8080; project: 2671
Folding: run 27, clone 29, generation 199; benchmark 0; misc: 500, 200
issue: Sun Jan 24 10:47:04 2010; begin: Sun Jan 24 10:47:26 2010
end: Sun Jan 24 19:06:02 2010; due: Wed Jan 27 10:47:26 2010 (3 days)
preferred: Wed Jan 27 10:47:26 2010 (3 days)
core URL: http://www.stanford.edu/~pande/Linux/AMD64/Core_a2.fah
core number: 0xa2; core name: GROCVS
CPU: 16,0 AMD64; OS: 4,0 Linux
client type: 3 Advmethods
smp cores: 4; cores to use: 4
tag: P2671R27C29G199 elapsed time: 8.25 hours
flops: 1062938618 (1062.938618 megaflops)
memory: 2013 MB
assignment info (le): Sun Jan 24 10:46:53 2010; B9B56FEA
CS: 171.67.108.25; P limit: 524286976
user: ikerekes; team: 50619; ID: C158A7683FA77A37; mach ID: 1
work/wudata_06.dat file size: 4825506; WU type: Folding@Home
Index 7: finished 470.00 pts (45.589 pt/hr) 14 X min speed
server: 130.237.232.140:8080; project: 6012
Folding: run 0, clone 25, generation 166; benchmark 0; misc: 500, 600
issue: Sun Jan 24 19:06:10 2010; begin: Sun Jan 24 19:06:19 2010
end: Mon Jan 25 05:24:53 2010; due: Sat Jan 30 19:06:19 2010 (6 days)
preferred: Wed Jan 27 19:06:19 2010 (3 days)
core URL: http://www.stanford.edu/~pande/Linux/AMD64/Core_a3.fah
core number: 0xa3; core name: GRO-A3
CPU: 16,0 AMD64; OS: 4,0 Linux
client type: 3 Advmethods
smp cores: 4; cores to use: 4
tag: P6012R0C25G166 elapsed time: 10.25 hours
flops: 1063034264 (1063.034264 megaflops)
memory: 2013 MB
assignment info (le): Sun Jan 24 19:05:58 2010; 901B1675
CS: 130.237.162.125; P limit: 524286976
user: ikerekes; team: 50619; ID: C158A7683FA77A37; mach ID: 1
work/wudata_07.dat file size: 1796825; WU type: Folding@Home
Index 8: folding now 470.00 pts (44.253 pt/hr) 13.6 X min speed; 75% complete
server: 130.237.232.140:8080; project: 6012
Folding: run 0, clone 171, generation 4; benchmark 0; misc: 500, 600
issue: Mon Jan 25 05:24:53 2010; begin: Mon Jan 25 05:25:03 2010
expect: Mon Jan 25 16:02:17 2010; due: Sun Jan 31 05:25:03 2010 (6 days)
preferred: Thu Jan 28 05:25:03 2010 (3 days)
core URL: http://www.stanford.edu/~pande/Linux/AMD64/Core_a3.fah (V2.13)
core number: 0xa3; core name: GRO-A3
CPU: 16,0 AMD64; OS: 4,0 Linux
client type: 3 Advmethods
smp cores: 4; cores to use: 4
tag: P6012R0C171G4 estimated elapsed time: 10.75 hours
flops: 1063257827 (1063.257827 megaflops)
memory: 2013 MB
assignment info (le): Mon Jan 25 05:24:44 2010; 9004E57F
CS: 130.237.162.125; P limit: 524286976
user: ikerekes; team: 50619; ID: C158A7683FA77A37; mach ID: 1

Q8400 @3.3Ghz
Index 2: finished 1920.00 pts (241.737 pt/hr) 9.07 X min speed
server: 171.64.65.56:8080; project: 2677
Folding: run 1, clone 33, generation 75; benchmark 0; misc: 500, 200
issue: Sun Jan 24 16:12:33 2010; begin: Sun Jan 24 16:13:03 2010
end: Mon Jan 25 00:09:36 2010; due: Wed Jan 27 16:13:03 2010 (3 days)
preferred: Wed Jan 27 16:13:03 2010 (3 days)
core URL: http://www.stanford.edu/~pande/Linux/AMD64/Core_a2.fah
core number: 0xa2; core name: GROCVS
CPU: 16,0 AMD64; OS: 4,0 Linux
client type: 3 Advmethods
smp cores: 4; cores to use: 4
tag: P2677R1C33G75 elapsed time: <8hours
flops: 1063500279 (1063.500279 megaflops)
memory: 2013 MB
assignment info (le): Sun Jan 24 16:12:24 2010; B9B6969F
CS: 171.67.108.25; P limit: 524286976
user: ikerekes; team: 50619; ID: AFC7C8150E8A6401; mach ID: 1
work/wudata_02.dat file size: 4842426; WU type: Folding@Home
Index 3: finished 380.00 pts (48.807 pt/hr) 9.25 X min speed
server: 130.237.232.140:8080; project: 6013
Folding: run 0, clone 0, generation 189; benchmark 0; misc: 500, 600
issue: Mon Jan 25 00:09:37 2010; begin: Mon Jan 25 00:09:54 2010
end: Mon Jan 25 07:57:03 2010; due: Thu Jan 28 00:09:54 2010 (3 days)
preferred: Thu Jan 28 00:09:54 2010 (3 days)
core URL: http://www.stanford.edu/~pande/Linux/AMD64/Core_a3.fah
core number: 0xa3; core name: GRO-A3
CPU: 16,0 AMD64; OS: 4,0 Linux
client type: 3 Advmethods
smp cores: 4; cores to use: 4
tag: P6013R0C0G189 elapsed time: <8hours
flops: 1063501379 (1063.501379 megaflops)
memory: 2013 MB
assignment info (le): Mon Jan 25 00:09:27 2010; 9004AF54
CS: 130.237.162.125; P limit: 524286976
user: ikerekes; team: 50619; ID: AFC7C8150E8A6401; mach ID: 1
work/wudata_03.dat file size: 4301861; WU type: Folding@Home
Index 4: folding now 470.00 pts (43.446 pt/hr) 13.3 X min speed; 51% complete
server: 130.237.232.140:8080; project: 6012
Folding: run 0, clone 31, generation 175; benchmark 0; misc: 500, 600
issue: Mon Jan 25 07:57:04 2010; begin: Mon Jan 25 07:57:13 2010
expect: Mon Jan 25 18:46:18 2010; due: Sun Jan 31 07:57:13 2010 (6 days)
preferred: Thu Jan 28 07:57:13 2010 (3 days)
core URL: http://www.stanford.edu/~pande/Linux/AMD64/Core_a3.fah (V2.13)
core number: 0xa3; core name: GRO-A3
CPU: 16,0 AMD64; OS: 4,0 Linux
client type: 3 Advmethods
smp cores: 4; cores to use: 4
tag: P6012R0C31G175 estimated elapsed time: 10.75 hours
flops: 1063513915 (1063.513915 megaflops)
memory: 2013 MB
assignment info (le): Mon Jan 25 07:56:55 2010; 9004C2C4
CS: 130.237.162.125; P limit: 524286976
user: ikerekes; team: 50619; ID: AFC7C8150E8A6401; mach ID: 1

E5200 @3.65
Index 2: finished 1920.00 pts (122.740 pt/hr) 4.6 X min speed
server: 171.64.65.56:8080; project: 2669
Folding: run 2, clone 27, generation 136; benchmark 0; misc: 500, 200
issue: Tue Jan 19 21:17:11 2010; begin: Wed Jan 20 01:06:34 2010
end: Wed Jan 20 16:45:08 2010; due: Sat Jan 23 01:06:34 2010 (3 days)
preferred: Sat Jan 23 01:06:34 2010 (3 days)
core URL: http://www.stanford.edu/~pande/Linux/AMD64/Core_a2.fah
core number: 0xa2; core name: GROCVS
CPU: 16,0 AMD64; OS: 4,0 Linux
smp cores: 2; cores to use: 2
tag: P2669R2C27G136 elapsed time: 15.7 hours
memory: 2013 MB
client type: 3 Advmethods
assignment info (le): Wed Jan 20 01:06:04 2010; B9B0732B
CS: 171.67.108.25; P limit: 524286976
user: ikerekes; team: 50619; ID: E33A357DEA94053F; mach ID: 1
work/wudata_02.dat file size: 4836103; WU type: Folding@Home
Index 8: folding now 484.00 pts (22.781 pt/hr) 6.78 X min speed; 48% complete
server: 130.237.232.140:8080; project: 6014
Folding: run 0, clone 65, generation 11; benchmark 0; misc: 500, 600
issue: Mon Jan 25 03:10:33 2010; begin: Mon Jan 25 03:10:42 2010
expect: Tue Jan 26 00:25:27 2010; due: Sun Jan 31 03:10:42 2010 (6 days)
preferred: Thu Jan 28 03:10:42 2010 (3 days)
core URL: http://www.stanford.edu/~pande/Linux/AMD64/Core_a3.fah (V2.13)
core number: 0xa3; core name: GRO-A3
CPU: 16,0 AMD64; OS: 4,0 Linux
smp cores: 2; cores to use: 2
tag: P6014R0C65G11 estimated elapsed time 21.25 hours
flops: 1061349142 (1061.349142 megaflops)
memory: 2013 MB
client type: 3 Advmethods
assignment info (le): Mon Jan 25 03:10:23 2010; 900485FC
CS: 130.237.162.125; P limit: 524286976
user: ikerekes; team: 50619; ID: E33A357DEA94053F; mach ID: 1
work/wudata_08.dat file size: 1800871; WU type: Folding@Home

E5200-2 @3.75Ghz

Index 7: finished 1920.00 pts (124.040 pt/hr) 4.65 X min speed
server: 171.64.65.56:8080; project: 2677
Folding: run 3, clone 33, generation 74; benchmark 0; misc: 500, 200
issue: Thu Jan 21 03:33:34 2010; begin: Thu Jan 21 09:38:21 2010
end: Fri Jan 22 01:07:05 2010; due: Sun Jan 24 09:38:21 2010 (3 days)
preferred: Sun Jan 24 09:38:21 2010 (3 days)
core URL: http://www.stanford.edu/~pande/Linux/x86/Core_a2.fah
core number: 0xa2; core name: GROCVS
CPU: 1,0 x86; OS: 4,0 Linux
client type: 3 Advmethods
smp cores: 2; cores to use: 2
tag: P2677R3C33G74 elapsed time : 13.5 hours
flops: 1061571922 (1061.571922 megaflops)
memory: 2013 MB
assignment info (le): Thu Jan 21 09:37:47 2010; B9B2399C
CS: 171.67.108.25; P limit: 524286976
user: ikerekes; team: 50619; ID: 30070C37EF9C9A64; mach ID: 1
work/wudata_07.dat file size: 4843821; WU type: Folding@Home
Index 2: folding now 470.00 pts (23.509 pt/hr) 7.2 X min speed; 88% complete
server: 130.237.232.140:8080; project: 6012
Folding: run 0, clone 22, generation 154; benchmark 0; misc: 500, 600
issue: Sun Jan 24 19:49:59 2010; begin: Sun Jan 24 19:51:39 2010
expect: Mon Jan 25 15:51:12 2010; due: Sat Jan 30 19:51:39 2010 (6 days)
preferred: Wed Jan 27 19:51:39 2010 (3 days)
core URL: http://www.stanford.edu/~pande/Linux/AMD64/Core_a3.fah (V2.13)
core number: 0xa3; core name: GRO-A3
CPU: 16,0 AMD64; OS: 4,0 Linux
client type: 3 Advmethods
smp cores: 2; cores to use: 2
tag: P6012R0C22G154 estimated elapsed time 20 hours
flops: 1061148597 (1061.148597 megaflops)
memory: 2013 MB
assignment info (le): Sun Jan 24 19:49:50 2010; 901B6C3D
CS: 130.237.162.125; P limit: 524286976
user: ikerekes; team: 50619; ID: 30070C37EF9C9A64; mach ID: 1
work/wudata_02.dat file size: 1797251; WU type: Folding@Home

A-620 quad@2.6Ghz

Index 1: finished 1920.00 pts (172.468 pt/hr) 6.47 X min speed
server: 171.64.65.56:8080; project: 2669
Folding: run 7, clone 147, generation 155; benchmark 0; misc: 500, 200
issue: Sun Jan 24 10:22:26 2010; begin: Sun Jan 24 10:23:14 2010
end: Sun Jan 24 21:31:11 2010; due: Wed Jan 27 10:23:14 2010 (3 days)
preferred: Wed Jan 27 10:23:14 2010 (3 days)
core URL: http://www.stanford.edu/~pande/Linux/AMD64/Core_a2.fah
core number: 0xa2; core name: GROCVS
CPU: 16,0 AMD64; OS: 4,0 Linux
client type: 3 Advmethods
smp cores: 4; cores to use: 4
tag: P2669R7C147G155 elapsed time: 11.2 hours
memory: 2048 MB
assignment info (le): Sun Jan 24 10:22:17 2010; B9B6388E
CS: 171.67.108.25; P limit: 524286976
user: ikerekes; team: 50619; ID: F839BB0C338CB810; mach ID: 1
work/wudata_01.dat file size: 4833238; WU type: Folding@Home
Index 2: ready for upload 470.00 pts (30.511 pt/hr) 9.35 X min speed
server: 130.237.232.140:8080; project: 6012
Folding: run 0, clone 61, generation 125; benchmark 0; misc: 500, 600
issue: Sun Jan 24 21:30:33 2010; begin: Sun Jan 24 21:31:21 2010
end: Mon Jan 25 12:55:37 2010; due: Sat Jan 30 21:31:21 2010 (6 days)
preferred: Wed Jan 27 21:31:21 2010 (3 days)
core URL: http://www.stanford.edu/~pande/Linux/AMD64/Core_a3.fah
core number: 0xa3; core name: GRO-A3
CPU: 16,0 AMD64; OS: 4,0 Linux
client type: 3 Advmethods
smp cores: 4; cores to use: 4
tag: P6012R0C61G125 elapsed time: 15.4 hours
flops: 1062759155 (1062.759155 megaflops)
memory: 2048 MB
assignment info (le): Sun Jan 24 21:30:23 2010; 901B75AC
CS: 130.237.162.125; upload failures: 3; P limit: 524286976
user: ikerekes; team: 50619; ID: F839BB0C338CB810; mach ID: 1
work/wudata_02.dat file size: 1797146; WU type: Folding@Home
Index 3: folding now 470.00 pts (30.214 pt/hr) 9.26 X min speed; 1% complete
server: 130.237.232.140:8080; project: 6012
Folding: run 0, clone 260, generation 0; benchmark 0; misc: 500, 600
issue: Mon Jan 25 12:54:47 2010; begin: Mon Jan 25 12:55:44 2010
expect: Tue Jan 26 04:29:04 2010; due: Sun Jan 31 12:55:44 2010 (6 days)
preferred: Thu Jan 28 12:55:44 2010 (3 days)
core URL: http://www.stanford.edu/~pande/Linux/AMD64/Core_a3.fah (V2.13)
core number: 0xa3; core name: GRO-A3
CPU: 16,0 AMD64; OS: 4,0 Linux
client type: 3 Advmethods
smp cores: 4; cores to use: 4
tag: P6012R0C260G0 estimated elapsed time: 15.4 hours
flops: 1063158812 (1063.158812 megaflops)
memory: 2048 MB
assignment info (le): Mon Jan 25 12:54:37 2010; 90041C0E
CS: 130.237.162.125; P limit: 524286976
user: ikerekes; team: 50619; ID: F839BB0C338CB810; mach ID: 1
work/wudata_03.dat file size: 1133559; WU type: Folding@Home
AMD X2 6000+ @3.0Ghz
Index 2: finished 1920.00 pts (79.946 pt/hr) 3 X min speed
server: 171.64.65.56:8080; project: 2669
Folding: run 4, clone 148, generation 121; benchmark 0; misc: 500, 200
issue: Mon Jan 18 03:50:52 2010; begin: Mon Jan 18 03:51:09 2010
end: Tue Jan 19 03:52:07 2010; due: Thu Jan 21 03:51:09 2010 (3 days)
preferred: Thu Jan 21 03:51:09 2010 (3 days)
core URL: http://www.stanford.edu/~pande/Linux/x86/Core_a2.fah
core number: 0xa2; core name: GROCVS
CPU: 1,0 x86; OS: 4,0 Linux
client type: 3 Advmethods
smp cores: 2; cores to use: 2
tag: P2669R4C148G121 elapsed time: 24 hours
flops: 1059397700 (1059.397700 megaflops)
memory: 878 MB
assignment info (le): Mon Jan 18 03:50:43 2010; B9BFFF44
CS: 171.67.108.25; P limit: 524286976
user: ikerekes; team: 50619; ID: 6DFE5442D293FE6D; mach ID: 1
work/wudata_02.dat file size: 4845172; WU type: Folding@Home
Index 8: folding now 470.00 pts (15.041 pt/hr) 4.61 X min speed; 59% complete
server: 130.237.232.140:8080; project: 6012
Folding: run 0, clone 46, generation 169; benchmark 0; misc: 500, 600
issue: Sun Jan 24 19:04:34 2010; begin: Sun Jan 24 19:04:42 2010
expect: Tue Jan 26 02:19:33 2010; due: Sat Jan 30 19:04:42 2010 (6 days)
preferred: Wed Jan 27 19:04:42 2010 (3 days)
core URL: http://www.stanford.edu/~pande/Linux/AMD64/Core_a3.fah (V2.13)
core number: 0xa3; core name: GRO-A3
CPU: 16,0 AMD64; OS: 4,0 Linux
client type: 3 Advmethods
smp cores: 2; cores to use: 2
tag: P6012R0C46G169 estimated elapsed time: 31.25 hours
flops: 1059202345 (1059.202345 megaflops)
memory: 1004 MB
assignment info (le): Sun Jan 24 19:04:24 2010; 901B17DB
CS: 130.237.162.125; P limit: 524286976
user: ikerekes; team: 50619; ID: 6DFE5442D293FE6D; mach ID: 1
work/wudata_08.dat file size: 1797834; WU type: Folding@Home

In summary:
Q6600 @3.3Ghz
tag: P2671R27C29G199 elapsed time: 8.25 hours
tag: P6012R0C25G166 elapsed time: 10.25 hours
tag: P6012R0C171G4 estimated elapsed time: 10.75 hours

Q8400 @3.3Ghz
tag: P2677R1C33G75 elapsed time: <8hours
tag: P6013R0C0G189 elapsed time: <8hours

tag: P6012R0C31G175 estimated elapsed time: 10.75 hours

E5200 @3.65
tag: P2669R2C27G136 elapsed time: 15.7 hours
tag: P6014R0C65G11 estimated elapsed time 21.25 hours

E5200-2 @3.75Ghz
tag: P2677R3C33G74 elapsed time : 13.5 hours
tag: P6012R0C22G154 estimated elapsed time 20 hours

A-620 quad@2.6Ghz
tag: P2669R7C147G155 elapsed time: 11.2 hours
tag: P6012R0C61G125 elapsed time: 15.4 hours
tag: P6012R0C260G0 estimated elapsed time: 15.4 hours

AMD X2 6000+ @3.0Ghz
tag: P2669R4C148G121 elapsed time: 24 hours
tag: P6012R0C46G169 estimated elapsed time: 31.25 hours

Of course the duo's are worst than the quads, and the AMD is worst than Intel but there is a definite drop in the elapsed time on all the P6012.

Hyperlife · Post by **Hyperlife** » Mon Jan 25, 2010 10:01 pm

The preferred deadline on 6012 is 3 days (72 hours), and your slowest rig shows a completion time of 31.25 hours. Looks like you're OK.

Yes, the times are a bit longer, but not long enough to miss the preferred deadline. Since different projects tend to have different sizes and different processing times, don't worry if your WU per day count drops -- you're still doing the equivalent number of hours of science.

kg4icg · Post by **kg4icg** » Tue Jan 26, 2010 6:07 am

Darn, I'm knocking A3's out in 6 hours, but then again i'm using a i7-920 clocked to 3.2 -smp8 with 6 gb ddr3 in which it's using less than 2400mb for the fold

Folding Forum

Project: 6012 (Run 0, Clone 32, Gen 168)

Project: 6012 (Run 0, Clone 32, Gen 168)

Re: Project: 6012 (Run 0, Clone 32, Gen 168)

Re: Project: 6012 (Run 0, Clone 32, Gen 168)

Re: Project: 6012 (Run 0, Clone 32, Gen 168)

Re: Project: 6012 (Run 0, Clone 32, Gen 168)

Re: Project: 6012 (Run 0, Clone 32, Gen 168)

Re: Project: 6012 (Run 0, Clone 32, Gen 168)

Re: Project: 6012 (Run 0, Clone 32, Gen 168)

Re: Project: 6012 (Run 0, Clone 32, Gen 168)

Re: Project: 6012 (Run 0, Clone 32, Gen 168)

Re: Project: 6012 (Run 0, Clone 32, Gen 168)

Re: Project: 6012 (Run 0, Clone 32, Gen 168)

Re: Project: 6012 (Run 0, Clone 32, Gen 168)

Re: Project: 6012 (Run 0, Clone 32, Gen 168)

Re: Project: 6012 (Run 0, Clone 32, Gen 168)