| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Support
Forum: BOINC Agent Support
Thread: less points with BOINC 5.8.8 for FightAIDS@home |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 9
|
|
| Author |
|
|
sbrandner
Cruncher Joined: Apr 1, 2006 Post Count: 2 Status: Offline Project Badges: 
|
I am using BOINC 5.8.8 on a Linux system. Usually I calculated for FightAIDS@home but when I was claiming points I found out that I
got less points for the same workunit. Workunit Name Status Sent Time Time Due / Return Time CPU Time (hours) Claimed/ Granted BOINC Credit faah1391_ d162n055_ x2BPW_ 01-- Valid 03/02/2007 20:44:53 03/03/2007 08:09:51 8.65 102.4 / 14.5 faah1391_ d162n055_ x2BPW_ 01-- Valid 03/02/2007 20:43:36 03/03/2007 08:45:03 7.71 32.8 / 29.0 faah1391_ d162n055_ x2BPW_ 01-- Valid 03/02/2007 20:39:38 03/03/2007 06:28:13 5.23 25.1 / 29.0 I got 14.5 points while all the other got 29 points. If I get half of the points I will not except further work for FightAIDS@home. Does anyone have an idea why I get less points. |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Suggest u upgrade to 5.8.15 for linux which has a better benchmark. The 102.4 claim has been deemed an 'Outlier', thus 50% of the remaining average score is awarded. The BOINC rules are explained here
----------------------------------------This has nothing specifically to do with FA@H. The rule is for all WCG projects. What machine does this run on as 8.65 hours does not strike me as particularly fast, yet the claim being of a much faster one. All members who are still running BOINC Linux version less than 5.8.15 are recommended to upgrade to 5.8.15 for the as mentioned better benchmarking.... get it here: http://boinc.berkeley.edu/download_all.php or directly: http://boinc.berkeley.edu/dl/boinc_5.8.15_windows_intelx86.exe Added: If in doubt, see this thread discussion by other members.
WCG
----------------------------------------Please help to make the Forums an enjoyable experience for All! [Edit 3 times, last edit by Sekerob at Mar 5, 2007 10:03:34 AM] |
||
|
|
sbrandner
Cruncher Joined: Apr 1, 2006 Post Count: 2 Status: Offline Project Badges:
|
Ok, I upgraded to 5.8.15.
I guess the reason for the long calculation time is that I have a dual core processor and each workunit runs on one processor core only which has less performance. I do not know if the benchmark takes care about that. Maybe this is the reason for the 'Outlier'. |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
No, on a dual core it is really 2x100% crunching power, on a hyperthreaded machine it's 115% tops. The benchmark is an evaluation of the computational capacity 'per core'. Add the Whet/Dhrystone values u see in message tab after a benchmark, divide by 480 and you get the 'per-core' CPU hourly claim. If 10 per core, with 2 jobs parallel would give 20 per CPU hour in total.
----------------------------------------
WCG
----------------------------------------Please help to make the Forums an enjoyable experience for All! [Edit 1 times, last edit by Sekerob at Mar 5, 2007 11:54:40 AM] |
||
|
|
D_T_G
Advanced Cruncher Poland Joined: Jan 25, 2007 Post Count: 56 Status: Offline Project Badges: 
|
BOINC Linux 5.8.15 (...) better benchmarking Benchmarks are rather at the same level, still worse than under Windows OS, though clamed credits are much higher on 5.8.15. At least for me and 2 others linux boinc users I know. |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
If Whet/Dhry Stone is same, but hourly claim is up, can only think that a different devisor was introduced. Really interested to see a few Linux benchmark values and a few jobs with their hours and claim...maybe can reconstitute what it is. Stock BOINC 5.8.15 only, else it becomes a waste of time.
----------------------------------------
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
D_T_G
Advanced Cruncher Poland Joined: Jan 25, 2007 Post Count: 56 Status: Offline Project Badges:
|
OK, I can start:
----------------------------------------5.8.11: 2007-02-16 11:08:35 [---] Starting BOINC client version 5.8.11 for i686-pc-linux-gnu 2007-02-16 11:08:35 [---] Processor: 2 GenuineIntel Intel(R) Core(TM)2 CPU 6400 @ 2.13GHz [fpu vme de pse tsc msr p ae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe nx lm constant_tsc pni monito r ds_cpl vmx est tm2 ssse3 cx16 xtpr lahf_lm] 2007-02-16 11:09:37 [---] 1898 floating point MIPS (Whetstone) per CPU 2007-02-16 11:09:37 [---] 4671 integer MIPS (Dhrystone) per CPU ======================================================= 2007-02-16 11:12:18 [---] 1890 floating point MIPS (Whetstone) per CPU 2007-02-16 11:12:18 [---] 4293 integer MIPS (Dhrystone) per CPU ======================================================= 2007-02-16 11:13:50 [---] 1892 floating point MIPS (Whetstone) per CPU 2007-02-16 11:13:50 [---] 4682 integer MIPS (Dhrystone) per CPU 5.8.15: 2007-02-28 09:23:48 [---] Starting BOINC client version 5.8.15 for i686-pc-linux-gnu 2007-02-28 09:23:48 [---] Version change (5.8.11 -> 5.8.15) 2007-02-28 09:24:50 [---] 1887 floating point MIPS (Whetstone) per CPU 2007-02-28 09:24:50 [---] 4489 integer MIPS (Dhrystone) per CPU My last valid faah: [Me]: faah1363_ d107n683_ x2BPW_ 01-- Valid 03/05/2007 02:29:39 03/05/2007 09:03:47 4.26 52.5 / 53.4 [others:] faah1363_ d107n683_ x2BPW_ 01-- Valid 02/25/2007 02:29:13 02/26/2007 05:24:07 20.63 54.3 / 53.4 faah1363_ d107n683_ x2BPW_ 01-- Valid 02/25/2007 02:25:57 02/25/2007 12:53:20 8.56 24.0 / 53.4 faah1391_ d163n724_ x2BPW_ 00-- Valid 03/03/2007 11:26:16 03/04/2007 01:36:59 10.44 36.9 / 32.3 faah1391_ d163n724_ x2BPW_ 00-- Valid 03/03/2007 11:26:08 03/04/2007 07:54:28 12.36 27.7 / 32.3 [me:] faah1391_ d163n724_ x2BPW_ 00-- Valid 03/03/2007 11:25:18 03/03/2007 22:12:52 4.97 64.4 / 32.3 faah1369_ d119n519_ x2BPW_ 02-- Valid 03/01/2007 23:20:02 03/02/2007 05:43:26 3.98 22.8 / 33.6 faah1369_ d119n519_ x2BPW_ 02-- Valid 03/01/2007 23:18:41 03/02/[me: as you can se the above is simmiliar cpu time, but I have twice bigger clamed] 2007 11:35:18 3.83 49.6 / 33.6 faah1369_ d119n519_ x2BPW_ 02-- Valid 03/01/2007 23:17:05 03/02/2007 08:40:25 4.76 28.4 / 33.6 This is propably from 5.8.11 (but not sure): faah1359_ d099n371_ x2BPW_ 02-- Valid 03/01/2007 22:19:56 03/02/2007 09:59:30 8.39 28.9 / 26.6 [me:]faah1359_ d099n371_ x2BPW_ 02-- Valid 02/21/2007 22:17:49 02/22/2007 17:09:11 4.13 53.0 / 26.6 faah1359_ d099n371_ x2BPW_ 02-- Valid 02/21/2007 22:16:25 02/22/2007 07:29:28 7.17 24.2 / 26.6 I can't provide any results from 5.4.11, but as far as I remember clamed credits were about 30 for 4h faah. |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Well, (Whet+Dhry)*Duration / Claim = Divisor. Depending on the WU it is either about 492 or 513. Your hourly claim is around 12.95.
----------------------------------------Why the divisor is not the 480 exact - always on my P4 - i dont know, but on my (windows) C2D it varies from 480 to 484. Anyway, Dr. Anderson seems to have tweaked the algorithm. With a before of about 30 and now 55-65, things look to be in much better shape on Linux.....one day if flop counting becomes a reality, all credits will be exactly identical no matter what platform.
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
D_T_G
Advanced Cruncher Poland Joined: Jan 25, 2007 Post Count: 56 Status: Offline Project Badges:
|
OMG the benchmarks on the 5.8.17 are extremely better :D
----------------------------------------I have just tried it, see: 2007-03-20 08:26:22 [---] Starting BOINC client version 5.8.17 for i686-pc-linux-gnu 2007-03-20 08:26:22 [---] log flags: task, file_xfer, sched_ops, task_debug 2007-03-20 08:26:22 [---] Libraries: libcurl/7.16.0 OpenSSL/0.9.8d zlib/1.2.3 (...) 2007-03-20 08:26:22 [---] Version change (5.8.15 -> 5.8.17) (...) 2007-03-20 08:27:24 [---] Benchmark results: 2007-03-20 08:27:24 [---] Number of CPUs: 2 2007-03-20 08:27:24 [---] 2208 floating point MIPS (Whetstone) per CPU 2007-03-20 08:27:24 [---] 6391 integer MIPS (Dhrystone) per CPU And compare it to my previous post! :O EDIT: And the divisor is ~ 504 :) ---------------------------------------- [Edit 1 times, last edit by D_T_G at Mar 20, 2007 10:56:02 AM] |
|||
|
||||
|
|