GNU/Linux 64 bit S5R3 "power users" App 4.28 available

Bernd Machenschalk
Bernd Machenschalk
Moderator
Administrator
Joined: 15 Oct 04
Posts: 3789
Credit: 170535336
RAC: 53219
Topic 193498

This App is from codebase of the 4.27 App but built as a native 64-bit App.

Actually I couldn't see any noticeable speedup compared to the 4.27, but this may vary depending on the CPU type.

I'm rather interested in finding out whether this buid system offers sufficient compatibility for tha current 64 bit Linux Distros.

You may want to look into the app_info.xml first. I tried to make migration easy from most previous Apps, but that might not suit everyone. Furthermore it has actulally not been tested that the checkpoint of 32 bit Apps is read in correctly by the 64 bit App.

Only run this if you're sure of what you're doing.

Find it on the Power User Apps page.

BM

Mikie Tim T
Mikie Tim T
Joined: 22 Jan 05
Posts: 103
Credit: 202854338
RAC: 21066

GNU/Linux 64 bit S5R3 "power users" App 4.28 available

I installed this app, and although I'm typically a credit whore, I thought I'd take one for the team and see if the new app would resume from checkpoint from the 4.27 i686 app. When I restarted BOINC after moving the files into place, I got a computation error on both workunits, and this in the messages:

Fri 08 Feb 2008 01:31:14 PM CST||Starting BOINC client version 5.10.28 for x86_64-pc-linux-gnu
Fri 08 Feb 2008 01:31:14 PM CST||log flags: task, file_xfer, sched_ops
Fri 08 Feb 2008 01:31:14 PM CST||Libraries: libcurl/7.17.1 OpenSSL/0.9.8g zlib/1.2.3
Fri 08 Feb 2008 01:31:14 PM CST||Data directory: /home/michael/BOINC
Fri 08 Feb 2008 01:31:14 PM CST|Einstein@Home|Found app_info.xml; using anonymous platform
Fri 08 Feb 2008 01:31:14 PM CST||Processor: 2 AuthenticAMD AMD Turion(tm) 64 X2 Mobile Technology TL-56 [Family 15 Model 72 Stepping 2]
Fri 08 Feb 2008 01:31:14 PM CST||Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt rdtscp lm 3dnowext 3dnow pni cx16 lahf_lm cmp_legacy svm cr8_legacy
Fri 08 Feb 2008 01:31:14 PM CST||OS: Linux: 2.6.20-16-generic
Fri 08 Feb 2008 01:31:14 PM CST||Memory: 1003.03 MB physical, 511.99 MB virtual
Fri 08 Feb 2008 01:31:14 PM CST||Disk: 19.53 GB total, 10.73 GB free
Fri 08 Feb 2008 01:31:14 PM CST||Local time is UTC -6 hours
Fri 08 Feb 2008 01:31:14 PM CST|Einstein@Home|URL: http://einstein.phys.uwm.edu/; Computer ID: 1098481; location: home; project prefs: default
Fri 08 Feb 2008 01:31:14 PM CST||General prefs: from http://burp.boinc.dk/ (last modified 28-Jun-2007 07:38:50)
Fri 08 Feb 2008 01:31:14 PM CST||Host location: none
Fri 08 Feb 2008 01:31:14 PM CST||General prefs: using your defaults
Fri 08 Feb 2008 01:31:14 PM CST||Preferences limit memory usage when active to 752.27MB
Fri 08 Feb 2008 01:31:14 PM CST||Preferences limit memory usage when idle to 902.72MB
Fri 08 Feb 2008 01:31:14 PM CST||Preferences limit disk usage to 3.73GB
Fri 08 Feb 2008 01:31:21 PM CST|Einstein@Home|Restarting task h1_0753.85_S5R2__5_S5R3a_0 using einstein_S5R3 version 427
Fri 08 Feb 2008 01:31:25 PM CST|Einstein@Home|Restarting task h1_0753.95_S5R2__71_S5R3a_0 using einstein_S5R3 version 427
Fri 08 Feb 2008 01:31:42 PM CST|Einstein@Home|Computation for task h1_0753.85_S5R2__5_S5R3a_0 finished
Fri 08 Feb 2008 01:31:42 PM CST|Einstein@Home|Output file h1_0753.85_S5R2__5_S5R3a_0_0 for task h1_0753.85_S5R2__5_S5R3a_0 absent
Fri 08 Feb 2008 01:31:43 PM CST|Einstein@Home|Computation for task h1_0753.95_S5R2__71_S5R3a_0 finished
Fri 08 Feb 2008 01:31:43 PM CST|Einstein@Home|Output file h1_0753.95_S5R2__71_S5R3a_0_0 for task h1_0753.95_S5R2__71_S5R3a_0 absent

So, since I lost about 17 hours of processing to this, I would heartily recommend others to allow your current results to finish before trying this. Will let you know tomorrow how the new app validates. This laptop is running Kubuntu 7.04 64 bit, by the way.

Mikie Tim T
Mikie Tim T
Joined: 22 Jan 05
Posts: 103
Credit: 202854338
RAC: 21066

In case you want to do a

In case you want to do a postmortem on the two results that came up with Computation Error, here they are:

92014606
91881250

Melvyn Bobo Slacke
Melvyn Bobo Slacke
Joined: 22 Jan 05
Posts: 32
Credit: 1692164
RAC: 0

Got this one: Process

Got this one: Process creation (../../projects/einstein.phys.uwm.edu/einstein_S5R3_4.28_x86_64-unknown-linux-gnu) failed: Error -1
But this is on a 32bit box.

PovAddict
PovAddict
Joined: 31 Mar 05
Posts: 45
Credit: 1437507
RAC: 847

RE: But this is on a 32bit

Message 78517 in response to message 78516

Quote:
But this is on a 32bit box.


64-bit apps can't run on 32-bit machines.

rroonnaalldd
rroonnaalldd
Joined: 12 Dec 05
Posts: 116
Credit: 537221
RAC: 0

RE: This App is from

Quote:

This App is from codebase of the 4.27 App but built as a native 64-bit App.

Actually I couldn't see any noticeable speedup compared to the 4.27, but this may vary depending on the CPU type.

I'm rather interested in finding out whether this buid system offers sufficient compatibility for tha current 64 bit Linux Distros.

You may want to look into the app_info.xml first. I tried to make migration easy from most previous Apps, but that might not suit everyone. Furthermore it has actulally not been tested that the checkpoint of 32 bit Apps is read in correctly by the 64 bit App.

Only run this if you're sure of what you're doing.

Find it on the Power User Apps page.

...installed on Suse64-10.2 in VMware-Server-1.0.4 and now crunching...
I hope to get results in 12hours.

rroonnaalldd
rroonnaalldd
Joined: 12 Dec 05
Posts: 116
Credit: 537221
RAC: 0

RE: ...installed on

Message 78519 in response to message 78518

Quote:

...installed on Suse64-10.2 in VMware-Server-1.0.4 and now crunching...
I hope to get results in 12hours.

First unit done on E6320 and also seen no speedup.

Mikie Tim T
Mikie Tim T
Joined: 22 Jan 05
Posts: 103
Credit: 202854338
RAC: 21066

Just finished up the first

Just finished up the first two results, which show a noticable slowdown over 4.27, more than can be attributed to the cyclical nature of the results. At least they validate successfully, though.

91937954
91936831

So, doesn't seem to like Turion X2 processors, at least.

th3
th3
Joined: 24 Aug 06
Posts: 208
Credit: 2208434
RAC: 0

RE: First unit done on

Message 78521 in response to message 78519

Quote:

First unit done on E6320 and also seen no speedup.


A bit slower than your 4.27 results?

Comparing with your 4.27 result at the same frequency and using Hewsons Ready Reckoner, the 4.27 result is only 2 from a highest runtime sequence number, and the 4.28 is 10 from a lowest runtime sequence number, still 4.28 is slower.

rroonnaalldd
rroonnaalldd
Joined: 12 Dec 05
Posts: 116
Credit: 537221
RAC: 0

RE: RE: First unit done

Message 78522 in response to message 78521

Quote:
Quote:

First unit done on E6320 and also seen no speedup.

A bit slower than your 4.27 results?

Comparing with your 4.27 result at the same frequency and using Hewsons Ready Reckoner, the 4.27 result is only 2 from a highest runtime sequence number, and the 4.28 is 10 from a lowest runtime sequence number, still 4.28 is slower.

by the way, which linux-version wins the ranking of speed now?

Hefto99
Hefto99
Joined: 11 Feb 05
Posts: 9
Credit: 182794092
RAC: 0

I have noticed significant

I have noticed significant speed decrease as well. this WU took 51ksec in comparison with 28k-30ksec needed for 4.27

Another WU currently in progress looks similar (49% in almost 7 hours).

Distro is openSuSE 10.3

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.