New N-Body Runs
log in

Advanced search

Message boards : News : New N-Body Runs

Author Message
Jake Bauer
Project developer
Project tester
Project scientist
Send message
Joined: 20 Aug 12
Posts: 66
Credit: 406,916
RAC: 0

Message 56730 - Posted: 4 Jan 2013, 19:12:33 UTC

Hello everyone,

We have a functional release of N-Body out, and we are going to do further testing to see if we can fix some of the issues you are still having.

Version 1.06 is on its way in response to the different problems that are happening.

For now, we are going to change the run that is currently up to do something slightly different. If you have not updated, you will error immediately. We have changed the minimum version requirements. This will help us see what is wrong specifically with the current release.

Also, please use Jeff Thompson's thread about errors.

Jake

Richard Haselgrove
Send message
Joined: 4 Sep 12
Posts: 218
Credit: 448,778
RAC: 0

Message 56731 - Posted: 4 Jan 2013, 19:17:03 UTC - in response to Message 56730.

If you have not updated, you will error immediately.

Updated what, exactly, please?

Jeffery M. Thompson
Volunteer moderator
Project developer
Project tester
Project scientist
Avatar
Send message
Joined: 23 Sep 12
Posts: 145
Credit: 8,365,827
RAC: 3,817

Message 56732 - Posted: 4 Jan 2013, 19:44:44 UTC

This is checking the version of the nbody binary specifically. It is requiring a minimum version of 1.04.

mitrichr
Avatar
Send message
Joined: 21 Dec 07
Posts: 24
Credit: 4,567,143
RAC: 0

Message 56733 - Posted: 4 Jan 2013, 21:02:48 UTC - in response to Message 56730.
Last modified: 4 Jan 2013, 21:03:45 UTC

see if we can fix some of the issues you are still having...


The one issue that I have is that I cannot crunch GPU for this project because I have Nvidia GTX670's, poor or no DP.

I would venture to guess that you could have a much larger cohort of GPU crunchers w/o DP. Apparently, MW is the only project insisting upon DP, essentially meaning ATI cards.

[My credits for MW are all CPU]
____________
http://sciencespringe.wordpress.com
http://facebook.com/sciencesprings
http://twitter.com/sciencesprings

Profile Overtonesinger
Avatar
Send message
Joined: 15 Feb 10
Posts: 63
Credit: 1,836,010
RAC: 0

Message 56782 - Posted: 7 Jan 2013, 16:26:01 UTC

I have re-added the project. I still see only single-CPU NBODY tasks! :(

Native NBODY Win x64 MultiThreaded app. will be ?

When ? :)

john.kingsnorth
Send message
Joined: 20 Sep 12
Posts: 1
Credit: 686,083
RAC: 0

Message 56893 - Posted: 15 Jan 2013, 16:43:09 UTC

All my WUs are throwing computation errors =] I've just suspended the project until a new update comes out, hopefully that will fix things!

Profile microchip
Send message
Joined: 25 Feb 09
Posts: 77
Credit: 6,869,202
RAC: 0

Message 56910 - Posted: 16 Jan 2013, 8:10:22 UTC - in response to Message 56893.

All my WUs are throwing computation errors =] I've just suspended the project until a new update comes out, hopefully that will fix things!


If I may ask, do all WUs error out or just the N-Body ones? If the N-Body, why completely suspend MW? Do as I do by deselecting N-Body in the preferences on the site and crunch only Separation WUs. That's what I do until N-Body is fixed.
____________

Team Belgium

Andy
Send message
Joined: 8 Jan 12
Posts: 1
Credit: 9,857,926
RAC: 0

Message 56948 - Posted: 18 Jan 2013, 14:21:56 UTC - in response to Message 56732.

My MW tasks have gone from taking minutes to the latest one needing over 9800 hours! (Yes, 9800...not a typo.) And yet the task was due in 4 days...man, talk about demanding. :) I aborted, and am hoping this issue gets resolved soon.

Andy

Ilium
Send message
Joined: 13 Jan 13
Posts: 1
Credit: 23,717
RAC: 0

Message 56957 - Posted: 20 Jan 2013, 0:26:06 UTC - in response to Message 56948.

My MW tasks have gone from taking minutes to the latest one needing over 9800 hours! (Yes, 9800...not a typo.) And yet the task was due in 4 days...man, talk about demanding. :) I aborted, and am hoping this issue gets resolved soon.

Andy


I received one of those as well. Not something my computer can get through even if I give it all the run time.

EdwardPF
Send message
Joined: 8 Apr 10
Posts: 25
Credit: 268,525
RAC: 0

Message 56958 - Posted: 20 Jan 2013, 2:28:32 UTC - in response to Message 56957.

My last 2 "big ones" errored out with disk space exceeded (boinc has it set to 50% - of 80Gb) and gave no credit (:-) ) the last one reported 0.00 sec CPU and 0.00 sec elapsed while in truth it was 42 Hrs elapsed (about 15%) and 120+ hrs CPU. so for now I'm not running nbody 'till I hear better reports.

Ed F

GaryG
Avatar
Send message
Joined: 29 Aug 12
Posts: 31
Credit: 40,781,945
RAC: 0

Message 56959 - Posted: 20 Jan 2013, 4:40:10 UTC

I just aborted two, one for 1368 hours and one for 8071 hours, both received today with a 1/31 deadline. The estimated time on the longer jobs appears to be about 3+ times the actual runtime on my system, one with 195 hours estimated ran in 54+. Even factoring that in, I could not complete these two by the deadline. I have also noticed several jobs with estimated time in minutes that have run 15 to 30+ hours. Maybe as more are completed the estimates can be made better.
____________

Profile Michael Yusko
Send message
Joined: 27 May 11
Posts: 20
Credit: 26,241,785
RAC: 23,354

Message 56996 - Posted: 23 Jan 2013, 3:22:05 UTC

Normal Milkyway@Home 1.02 runs a workunit every 1 minute and 36 seconds using 92% of my GPU. N-Body takes ~5hours???? using less than 1% of my GPU??? Is this supposed to be normal?
____________
-mpyusko

Intel Xeon E5450 @ 3.0GHz
AMD Radeon HD6950 2GB @ 820MHz/1250MHz

Jake Bauer
Project developer
Project tester
Project scientist
Send message
Joined: 20 Aug 12
Posts: 66
Credit: 406,916
RAC: 0

Message 57001 - Posted: 23 Jan 2013, 13:44:23 UTC - in response to Message 56996.

We don't have a GPU application released yet for N-Body. This is normal.

Jake

Richard Haselgrove
Send message
Joined: 4 Sep 12
Posts: 218
Credit: 448,778
RAC: 0

Message 57003 - Posted: 23 Jan 2013, 13:52:08 UTC - in response to Message 57001.

We don't have a GPU application released yet for N-Body.

Hence my question about the Linux (opencl_amd_ati) and (opencl_nvidia) plan classes in the 'stopped' thread.

Profile Michael Yusko
Send message
Joined: 27 May 11
Posts: 20
Credit: 26,241,785
RAC: 23,354

Message 57007 - Posted: 24 Jan 2013, 0:11:41 UTC - in response to Message 57001.

According to what I am seeing in BOINC Manager, it says N-Body is using .05 CPUs/1 ATI. (It tells me it is utilizing the GPU) If it is not using the GPU like you say, then it is trying to simultaneously compute on a CPU thread against another project task. Hmmmmmm.
____________
-mpyusko

Intel Xeon E5450 @ 3.0GHz
AMD Radeon HD6950 2GB @ 820MHz/1250MHz

Alinator
Send message
Joined: 7 Jun 08
Posts: 464
Credit: 56,639,936
RAC: 0

Message 57011 - Posted: 24 Jan 2013, 12:22:54 UTC - in response to Message 57007.
Last modified: 24 Jan 2013, 12:29:28 UTC

According to what I am seeing in BOINC Manager, it says N-Body is using .05 CPUs/1 ATI. (It tells me it is utilizing the GPU) If it is not using the GPU like you say, then it is trying to simultaneously compute on a CPU thread against another project task. Hmmmmmm.


Interesting. I guess that partially answers a question Richard had about what the project shows for Linux apps.

Can you verify if it's actually doing anything on the GPU, or just doing 'straight' CPU calculation concurrently with the other task?

In any event, the way it appears to be running would be normal for a typical GPU task (sharing the CPU control thread with another CPU task) in a default host config where BOINC was allowed to utilize all the host resources as it sees fit.

However, I suppose it also indicates the project team does intend to try and get a true GPU app rolled out at some point (if it wasn't doing anything on the GPU at the moment).

<edit> Your account data came up finally, and it looks like the tasks were only running on the CPU.

Richard Haselgrove
Send message
Joined: 4 Sep 12
Posts: 218
Credit: 448,778
RAC: 0

Message 57012 - Posted: 24 Jan 2013, 13:48:03 UTC - in response to Message 57011.

According to what I am seeing in BOINC Manager, it says N-Body is using .05 CPUs/1 ATI. (It tells me it is utilizing the GPU) If it is not using the GPU like you say, then it is trying to simultaneously compute on a CPU thread against another project task. Hmmmmmm.

Interesting. I guess that partially answers a question Richard had about what the project shows for Linux apps.

Can you verify if it's actually doing anything on the GPU, or just doing 'straight' CPU calculation concurrently with the other task?

In any event, the way it appears to be running would be normal for a typical GPU task (sharing the CPU control thread with another CPU task) in a default host config where BOINC was allowed to utilize all the host resources as it sees fit.

However, I suppose it also indicates the project team does intend to try and get a true GPU app rolled out at some point (if it wasn't doing anything on the GPU at the moment).

<edit> Your account data came up finally, and it looks like the tasks were only running on the CPU.

What you see in BOINC Manager, in an information line like ".05 CPUs/1 ATI", are simply figures copied from BOINC's <app_version> specification downloaded to your computer when the task was issued. And that, in turn, is copied from the application specification that was written to the BOINC server by the administrator when the application was deployed.

In this case, all it's telling us is that a whole ATI card has been 'reserved' by the BOINC client on your machine for the exclusive use of the N-Body app, and that BOINC won't schedule any other task (from this project or any other project) to use that ATI resource throughout the duration of the N-Body run.

Whether or not the application is written in such a way that its code can execute on an ATI card is entirely a matter for the application developer - and from what Jake Bauer wrote ("We don't have a GPU application released yet for N-Body"), this app will make no use whatsoever of the card which BOINC has made available for it.

That, from what I'm seeing, can only be explained by a mis-communication between the developer(s) who wrote the app, and the administrator(s) who deployed it on the BOINC server.

Profile Michael Yusko
Send message
Joined: 27 May 11
Posts: 20
Credit: 26,241,785
RAC: 23,354

Message 57015 - Posted: 24 Jan 2013, 18:29:37 UTC - in response to Message 57012.
Last modified: 24 Jan 2013, 18:38:56 UTC

My preferences are to allow MilkyWay@Home GPU computation only. I utilize both CPU threads on this machine for Seti@Home. It should never accept any CPU tasks from MilkWay@Home on this machine. As you stated it appears it was improperly deployed.

This runs on a Slackware 14.0 64-bit system. I also have a 2 Debian Testing machines, 1 64-bit (Dual Xeon) and 1 32-bit (Pentium M). Neither have a GPU capable of project computing. I'd be happy to provide any feedback and information I can, just tell me what you want to know.

I've thought about picking up a decent GPU for the Xeon but being a server, I'm limited to PCIe 1x and there aren't any DP GPUs available. (I think the top GPU I can get for it would be a HD5450) It has a Voodoo3 2000 PCI in it right now because that is far better than the Rage XL residing onboard. But I digress.

As I said, if you want Linux feedback, let me know.

Edit:

bash-4.2$ top -bn 1 top - 13:35:04 up 13 min, 7 users, load average: 5.48, 4.88, 2.74 Tasks: 196 total, 4 running, 191 sleeping, 0 stopped, 1 zombie Cpu(s): 15.3%us, 9.2%sy, 40.1%ni, 27.7%id, 7.5%wa, 0.0%hi, 0.2%si, 0.0%st Mem: 8160416k total, 3073960k used, 5086456k free, 143096k buffers Swap: 31053864k total, 0k used, 31053864k free, 1306536k cached PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 26681 mpyusko 30 10 19284 8796 992 R 113 0.1 3:23.35 milkyway_nbody_ 26283 mpyusko 39 19 57132 43m 1872 S 18 0.5 0:28.84 setiathome-5.28 26284 mpyusko 39 19 80216 64m 1848 S 18 0.8 0:28.66 setiathome-5.28


bash-4.2$ clinfo Number of platforms: 1 Platform Profile: FULL_PROFILE Platform Version: OpenCL 1.2 AMD-APP (1084.4) Platform Name: AMD Accelerated Parallel Processing Platform Vendor: Advanced Micro Devices, Inc. Platform Extensions: cl_khr_icd cl_amd_event_callback cl_amd_offline_devices Platform Name: AMD Accelerated Parallel Processing Number of devices: 2 Device Type: CL_DEVICE_TYPE_GPU Device ID: 4098 Board name: AMD Radeon HD 6900 Series Device Topology: PCI[ B#1, D#0, F#0 ] Max compute units: 22 Max work items dimensions: 3 Max work items[0]: 256 Max work items[1]: 256 Max work items[2]: 256 Max work group size: 256 Preferred vector width char: 16 Preferred vector width short: 8 Preferred vector width int: 4 Preferred vector width long: 2 Preferred vector width float: 4 Preferred vector width double: 2 Native vector width char: 16 Native vector width short: 8 Native vector width int: 4 Native vector width long: 2 Native vector width float: 4 Native vector width double: 2 Max clock frequency: 820Mhz Address bits: 32 Max memory allocation: 268435456 Image support: Yes Max number of images read arguments: 128 Max number of images write arguments: 8 Max image 2D width: 16384 Max image 2D height: 16384 Max image 3D width: 2048 Max image 3D height: 2048 Max image 3D depth: 2048 Max samplers within kernel: 16 Max size of kernel argument: 1024 Alignment (bits) of base address: 2048 Minimum alignment (bytes) for any datatype: 128 Single precision floating point capability Denorms: No Quiet NaNs: Yes Round to nearest even: Yes Round to zero: Yes Round to +ve and infinity: Yes IEEE754-2008 fused multiply-add: Yes Cache type: None Cache line size: 0 Cache size: 0 Global memory size: 1073741824 Constant buffer size: 65536 Max number of constant args: 8 Local memory type: Scratchpad Local memory size: 32768 Kernel Preferred work group size multiple: 64 Error correction support: 0 Unified memory for Host and Device: 0 Profiling timer resolution: 1 Device endianess: Little Available: Yes Compiler available: Yes Execution capabilities: Execute OpenCL kernels: Yes Execute native function: No Queue properties: Out-of-Order: No Profiling : Yes Platform ID: 0x00007f3e17fe7ee0 Name: Cayman Vendor: Advanced Micro Devices, Inc. Device OpenCL C version: OpenCL C 1.2 Driver version: 1084.4 Profile: FULL_PROFILE Version: OpenCL 1.2 AMD-APP (1084.4) Extensions: cl_khr_fp64 cl_amd_fp64 cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_khr_3d_image_writes cl_khr_byte_addressable_store cl_khr_gl_sharing cl_ext_atomic_counters_32 cl_amd_device_attribute_query cl_amd_vec3 cl_amd_printf cl_amd_media_ops cl_amd_media_ops2 cl_amd_popcnt Device Type: CL_DEVICE_TYPE_CPU Device ID: 4098 Board name: Max compute units: 2 Max work items dimensions: 3 Max work items[0]: 1024 Max work items[1]: 1024 Max work items[2]: 1024 Max work group size: 1024 Preferred vector width char: 16 Preferred vector width short: 8 Preferred vector width int: 4 Preferred vector width long: 2 Preferred vector width float: 4 Preferred vector width double: 2 Native vector width char: 16 Native vector width short: 8 Native vector width int: 4 Native vector width long: 2 Native vector width float: 4 Native vector width double: 2 Max clock frequency: 3203Mhz Address bits: 64 Max memory allocation: 2147483648 Image support: Yes Max number of images read arguments: 128 Max number of images write arguments: 8 Max image 2D width: 8192 Max image 2D height: 8192 Max image 3D width: 2048 Max image 3D height: 2048 Max image 3D depth: 2048 Max samplers within kernel: 16 Max size of kernel argument: 4096 Alignment (bits) of base address: 1024 Minimum alignment (bytes) for any datatype: 128 Single precision floating point capability Denorms: Yes Quiet NaNs: Yes Round to nearest even: Yes Round to zero: Yes Round to +ve and infinity: Yes IEEE754-2008 fused multiply-add: Yes Cache type: Read/Write Cache line size: 64 Cache size: 32768 Global memory size: 8356265984 Constant buffer size: 65536 Max number of constant args: 8 Local memory type: Global Local memory size: 32768 Kernel Preferred work group size multiple: 1 Error correction support: 0 Unified memory for Host and Device: 1 Profiling timer resolution: 1 Device endianess: Little Available: Yes Compiler available: Yes Execution capabilities: Execute OpenCL kernels: Yes Execute native function: Yes Queue properties: Out-of-Order: No Profiling : Yes Platform ID: 0x00007f3e17fe7ee0 Name: Pentium(R) Dual-Core CPU E6700 @ 3.20GHz Vendor: GenuineIntel Device OpenCL C version: OpenCL C 1.2 Driver version: 1084.4 (sse2) Profile: FULL_PROFILE Version: OpenCL 1.2 AMD-APP (1084.4) Extensions: cl_khr_fp64 cl_amd_fp64 cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_khr_int64_base_atomics cl_khr_int64_extended_atomics cl_khr_3d_image_writes cl_khr_byte_addressable_store cl_khr_gl_sharing cl_ext_device_fission cl_amd_device_attribute_query cl_amd_vec3 cl_amd_printf cl_amd_media_ops cl_amd_media_ops2 cl_amd_popcnt

____________
-mpyusko

Intel Xeon E5450 @ 3.0GHz
AMD Radeon HD6950 2GB @ 820MHz/1250MHz


Post to thread

Message boards : News : New N-Body Runs


Main page · Your account · Message boards


Copyright © 2016 AstroInformatics Group