Release of MilkyWay Sepearation Modified Fit v1.30
log in

Advanced search

Message boards : News : Release of MilkyWay Sepearation Modified Fit v1.30

1 · 2 · Next
Author Message
Jake Weiss
Volunteer moderator
Project developer
Project tester
Project scientist
Send message
Joined: 25 Feb 13
Posts: 189
Credit: 4,386,102
RAC: 3,631

Message 61516 - Posted: 15 Apr 2014, 16:45:15 UTC

Hi everyone,

I have just released a new version of MilkyWay Separation Modified Fit. There were small changes made that only effected the integration area used in our likelihood calculations to be more consistent with previous versions of MilkyWay@home. This is just a minor bookkeeping change so no issues are expected, but as always if any issues come up please post them here.

Thank you,

Jake W.

[TA]Assimilator1
Avatar
Send message
Joined: 22 Jan 11
Posts: 316
Credit: 40,011,286
RAC: 9

Message 61520 - Posted: 16 Apr 2014, 18:31:58 UTC
Last modified: 16 Apr 2014, 18:39:39 UTC

Does it fix the Windows 32 bit bug?
Bug report in v1.28 thread http://milkyway.cs.rpi.edu/milkyway/forum_thread.php?id=3357&postid=59922

[edit]I assume not seeing as their isn't a v1.30 win 32bit app listed in the downloads folder :(. http://milkyway.cs.rpi.edu/milkyway/download/
____________
Team AnandTech - SETI@H, Muon1 DPAD, F@H, MW@H, A@H, LHC@H, POGS, R@H, Einstein@H.

Main rig - i7 4820k @4.34 GHz, HD 7870 XT 3 GB, 16 GB DDR3 1866, Win 7 64bit, BOINC 7.6.22
2nd rig - Q9550 @3.6 GHz, HD 5850, 4GB DDR2 1066, Win 7 64bit

Len LE/GE
Send message
Joined: 8 Feb 08
Posts: 261
Credit: 104,050,322
RAC: 0

Message 61521 - Posted: 16 Apr 2014, 20:26:18 UTC

Doesn't look like it.

The commits on github don't give any hints about it and I can't see a 32bit version for windows in the download directory.

Jake Weiss
Volunteer moderator
Project developer
Project tester
Project scientist
Send message
Joined: 25 Feb 13
Posts: 189
Credit: 4,386,102
RAC: 3,631

Message 61525 - Posted: 17 Apr 2014, 4:49:19 UTC

Hey guys,

This release does not fix the win32 issues, I'm sorry. It should be possible for you guys to compile your own version of milkyway@home for win32 if you are interested. The issue with us releasing a version is it has to work with all computers running win32 even with some outdated hardware, which has proven itself to be very difficult. If you can get it running on your machine though it is possible to still get work units and run them for credits.

Jake W.

[TA]Assimilator1
Avatar
Send message
Joined: 22 Jan 11
Posts: 316
Credit: 40,011,286
RAC: 9

Message 61528 - Posted: 17 Apr 2014, 17:24:10 UTC - in response to Message 61525.
Last modified: 17 Apr 2014, 17:24:41 UTC

Hi Jake, thx for the reply.
Maybe I would compile a MW app if I knew what compiling even was ;).
When you said 'if we could get it running', did you mean with the v1.28 app or a newly complied one? If v1.28 do you know what hardware it fails on?

Len
Commits? Github?
Yea like I said it's not in the d/l dir ;)
____________
Team AnandTech - SETI@H, Muon1 DPAD, F@H, MW@H, A@H, LHC@H, POGS, R@H, Einstein@H.

Main rig - i7 4820k @4.34 GHz, HD 7870 XT 3 GB, 16 GB DDR3 1866, Win 7 64bit, BOINC 7.6.22
2nd rig - Q9550 @3.6 GHz, HD 5850, 4GB DDR2 1066, Win 7 64bit

Len LE/GE
Send message
Joined: 8 Feb 08
Posts: 261
Credit: 104,050,322
RAC: 0

Message 61541 - Posted: 18 Apr 2014, 23:14:52 UTC

The source is here on GitHub.
I found it sometimes gives interesting hints about the latest changes and helped me in the past to understand how to use the cmdline line params to tweak the apps for my gpu.

Profile Crunch3r
Volunteer developer
Avatar
Send message
Joined: 17 Feb 08
Posts: 363
Credit: 258,180,004
RAC: 0

Message 61549 - Posted: 19 Apr 2014, 14:37:03 UTC - in response to Message 61525.
Last modified: 19 Apr 2014, 14:39:30 UTC

Hey guys,

This release does not fix the win32 issues, I'm sorry. It should be possible for you guys to compile your own version of milkyway@home for win32 if you are interested. The issue with us releasing a version is it has to work with all computers running win32 even with some outdated hardware, which has proven itself to be very difficult. If you can get it running on your machine though it is possible to still get work units and run them for credits.

Jake W.


Howdy,

I'm not sure what that win32 "bug" is all about but with some clarification about it, i might be able to help...
____________

Join Support science! Joinc Team BOINC United now!

Jake Weiss
Volunteer moderator
Project developer
Project tester
Project scientist
Send message
Joined: 25 Feb 13
Posts: 189
Credit: 4,386,102
RAC: 3,631

Message 61552 - Posted: 19 Apr 2014, 15:29:58 UTC
Last modified: 19 Apr 2014, 15:30:39 UTC

Hey guys,

Currently, there is an issue getting the boinc libraries to compile with our code in the win32 version. I had a pretty sketchy fix to it over last summer, but in the end it ended up not working on all computers.

Thank you,

Jake W.

[TA]Assimilator1
Avatar
Send message
Joined: 22 Jan 11
Posts: 316
Credit: 40,011,286
RAC: 9

Message 61554 - Posted: 19 Apr 2014, 16:16:01 UTC

Do you know which h/w it didn't work on?
____________
Team AnandTech - SETI@H, Muon1 DPAD, F@H, MW@H, A@H, LHC@H, POGS, R@H, Einstein@H.

Main rig - i7 4820k @4.34 GHz, HD 7870 XT 3 GB, 16 GB DDR3 1866, Win 7 64bit, BOINC 7.6.22
2nd rig - Q9550 @3.6 GHz, HD 5850, 4GB DDR2 1066, Win 7 64bit

Profile Crunch3r
Volunteer developer
Avatar
Send message
Joined: 17 Feb 08
Posts: 363
Credit: 258,180,004
RAC: 0

Message 61561 - Posted: 20 Apr 2014, 16:38:04 UTC - in response to Message 61552.

Hey guys,

Currently, there is an issue getting the boinc libraries to compile with our code in the win32 version. I had a pretty sketchy fix to it over last summer, but in the end it ended up not working on all computers.

Thank you,

Jake W.


Could you be more specific about that?

I've got no problem compiling it for 32 bit nor 64 bit at all.
I'm guessing that you think you need to compile/link to the latest boinc api/lib while compiling a new MW app, which is not needed at all. Use the one the previously worked...


____________

Join Support science! Joinc Team BOINC United now!

Jake Weiss
Volunteer moderator
Project developer
Project tester
Project scientist
Send message
Joined: 25 Feb 13
Posts: 189
Credit: 4,386,102
RAC: 3,631

Message 61593 - Posted: 23 Apr 2014, 3:01:00 UTC

With the compiler we use there has been an issue where some of the 64 bit definitions in BOINC already exist in the compiler for 32 bit. It is a weird issue and likely a bug with our compiler. (One of the many we have run into and sadly updating the version of our compiler has not helped.)

Jake W.

Len LE/GE
Send message
Joined: 8 Feb 08
Posts: 261
Credit: 104,050,322
RAC: 0

Message 61601 - Posted: 23 Apr 2014, 17:48:55 UTC

Message 60211 is an example I guess.
Isn't that about something like SSE2 on a cpu that is only able to do SSE?

[TA]Assimilator1
Avatar
Send message
Joined: 22 Jan 11
Posts: 316
Credit: 40,011,286
RAC: 9

Message 61603 - Posted: 23 Apr 2014, 17:57:39 UTC

Jake
Did you see my last question?
____________
Team AnandTech - SETI@H, Muon1 DPAD, F@H, MW@H, A@H, LHC@H, POGS, R@H, Einstein@H.

Main rig - i7 4820k @4.34 GHz, HD 7870 XT 3 GB, 16 GB DDR3 1866, Win 7 64bit, BOINC 7.6.22
2nd rig - Q9550 @3.6 GHz, HD 5850, 4GB DDR2 1066, Win 7 64bit

MCP_007
Send message
Joined: 27 May 13
Posts: 9
Credit: 31,962,726
RAC: 22,171

Message 61661 - Posted: 28 Apr 2014, 13:49:06 UTC

Hi,

I have over 100 tasks with validation inconclusive starting on April 10th and my points dropped from 30 k to 10 k starting on April 14th.

Any help?

Thanks in advance,

Rene

[TA]Assimilator1
Avatar
Send message
Joined: 22 Jan 11
Posts: 316
Credit: 40,011,286
RAC: 9

Message 61662 - Posted: 28 Apr 2014, 17:27:03 UTC - in response to Message 61661.
Last modified: 28 Apr 2014, 17:27:18 UTC

Seeing as v1.3 was released on the 15th I don't see how your problem could be related to the new app.
More of a concern than the 'inconclusive' tasks is the huge number of invalid tasks, you definitely have a problem! Have you updated graphics drivers lately?

Btw I noticed your not crunching MW v1.02 WUs, how come?
____________
Team AnandTech - SETI@H, Muon1 DPAD, F@H, MW@H, A@H, LHC@H, POGS, R@H, Einstein@H.

Main rig - i7 4820k @4.34 GHz, HD 7870 XT 3 GB, 16 GB DDR3 1866, Win 7 64bit, BOINC 7.6.22
2nd rig - Q9550 @3.6 GHz, HD 5850, 4GB DDR2 1066, Win 7 64bit

MCP_007
Send message
Joined: 27 May 13
Posts: 9
Credit: 31,962,726
RAC: 22,171

Message 61664 - Posted: 29 Apr 2014, 3:16:44 UTC - in response to Message 61662.

Hi,

For the invalid tasks I asked two times here, but unfortunately I got no answer. All tasks with bpl in them failed. The graphics driver is 13.251-131206a-166389C-ATI (which was the newest one until some days ago). I stopped overclocking my 3770K and the Radeon HD7770 for a while, but still got the invalid bpl WUs ...

I will install the 14.4 driver tonight and see what happens...

V1.02 WUs? I do not know how to enable them ... :-(


Thanks,

Rene

mac intel
Send message
Joined: 22 Mar 11
Posts: 5
Credit: 5,397
RAC: 0

Message 61665 - Posted: 29 Apr 2014, 7:44:53 UTC

hello, i´ve got error messages like this

ps_modfit_15_3s_130_wrap_1_1398336302_1323456
http://milkyway.cs.rpi.edu/milkyway/workunit.php?wuid=545908915

Stderr Ausgabe <core_client_version>7.3.15</core_client_version> <![CDATA[ <message> process exited with code 1 (0x1, -255) </message> <stderr_txt> <search_application> milkyway_separation 1.30 Darwin x86_64 double OpenCL </search_application> Reading preferences ended prematurely BOINC GPU type suggests using OpenCL vendor 'NVIDIA Corporation' Setting process priority to 0 (13): Permission denied Error loading Lua script 'astronomy_parameters.txt': [string "number_parameters: 4..."]:1: '<name>' expected near '4' Switching to Parameter File Using SSE4.1 path Found 1 platform Platform 0 information: Name: Apple Version: OpenCL 1.2 (Dec 8 2013 21:07:05) Vendor: Apple Extensions: cl_APPLE_SetMemObjectDestructor cl_APPLE_ContextLoggingFunctions cl_APPLE_clut cl_APPLE_query_kernel_names cl_APPLE_gl_sharing cl_khr_gl_event Profile: FULL_PROFILE Didn't find preferred platform Using device 0 on platform 0 Found 2 CL devices Device 'HD Graphics 4000' (Intel:0x1024400) (CL_DEVICE_TYPE_GPU) Board: Driver version: 1.2(Feb 7 2014 18:36:33) Version: OpenCL 1.2 Compute capability: 0.0 Max compute units: 16 Clock frequency: 1200 Mhz Global mem size: 1073741824 Local mem size: 65536 Max const buf size: 65536 Double extension: (none) Device doesn't support double precision Failed to calculate likelihood <background_integral> nan </background_integral> <stream_integral> nan nan nan </stream_integral> <background_likelihood> nan </background_likelihood> <stream_only_likelihood> nan nan nan </stream_only_likelihood> <search_likelihood> nan </search_likelihood> 09:35:40 (10322): called boinc_finish </stderr_txt> ]]>

[TA]Assimilator1
Avatar
Send message
Joined: 22 Jan 11
Posts: 316
Credit: 40,011,286
RAC: 9

Message 61666 - Posted: 29 Apr 2014, 17:06:08 UTC - in response to Message 61664.
Last modified: 29 Apr 2014, 17:10:06 UTC

Are you trying to run it on your Intel graphics?
If so, no wonder, it doesn't have double precision, as BOINC is telling you.

Device doesn't support double precision

Hi,

For the invalid tasks I asked two times here, but unfortunately I got no answer. All tasks with bpl in them failed. The graphics driver is 13.251-131206a-166389C-ATI (which was the newest one until some days ago). I stopped overclocking my 3770K and the Radeon HD7770 for a while, but still got the invalid bpl WUs ...

I will install the 14.4 driver tonight and see what happens...

V1.02 WUs? I do not know how to enable them ... :-(

Thanks,

Rene


Is that cat 13.2? That might be why, although the issue with any Cat drivers after 13.1 usually affects 5800s & 6900s not 7700s & MW v1.02 WUs not mod fit ones, hmmm not sure TBH but I suggest you start your own thread about it, you'll get more exposure then, so to speak ;).

Btw you didn't answer me, did you change the driver just before you got all those invalid tasks?
Usually new drivers are bad news for MW.

Re v1.02, you must of disabled them because their on by default.
Go to your online MW preferences & tick MilkyWay@Home (under 'Run only the selected applications')
____________
Team AnandTech - SETI@H, Muon1 DPAD, F@H, MW@H, A@H, LHC@H, POGS, R@H, Einstein@H.

Main rig - i7 4820k @4.34 GHz, HD 7870 XT 3 GB, 16 GB DDR3 1866, Win 7 64bit, BOINC 7.6.22
2nd rig - Q9550 @3.6 GHz, HD 5850, 4GB DDR2 1066, Win 7 64bit

mac intel
Send message
Joined: 22 Mar 11
Posts: 5
Credit: 5,397
RAC: 0

Message 61667 - Posted: 30 Apr 2014, 6:47:33 UTC

[quoteAre you trying to run it on your Intel graphics?
If so, no wonder, it doesn't have double precision, as BOINC is telling you.

Device doesn't support double precision[/quote]

hello, i´ve got the an nvidia geforce gtx650and the hd4000 build in.
but the os x 1.28/1.30 app select the hd4000 and give these error.

do i need an extra file to select the right card or does the app does´t support the gtx650?

Jake Weiss
Volunteer moderator
Project developer
Project tester
Project scientist
Send message
Joined: 25 Feb 13
Posts: 189
Credit: 4,386,102
RAC: 3,631

Message 61668 - Posted: 30 Apr 2014, 15:28:23 UTC

MCP_007,

So it looks like you're running a version you compiled on your own. (Specifically version 1.24). You should download the newest code off github and recompile it with the changes. This should fix your validate errors.

Thank you,

Jake W

1 · 2 · Next
Post to thread

Message boards : News : Release of MilkyWay Sepearation Modified Fit v1.30


Main page · Your account · Message boards


Copyright © 2016 AstroInformatics Group