New system many invalids and errors

Message boards : Number crunching : New system many invalids and errors
Message board moderation

To post messages, you must log in.

1 · 2 · Next

AuthorMessage
ETWhereRU

Send message
Joined: 10 Mar 12
Posts: 11
Credit: 3,000,869
RAC: 0
United States
Message 1641235 - Posted: 13 Feb 2015, 14:33:45 UTC
Last modified: 13 Feb 2015, 14:51:37 UTC

Hello All,

Replaced my system and for the three projects I am running, many are invalid. I have over the past week tried many different combinations to my app_config files and still no love. At this time I have no app_config and 90% CPU in my preferences. Below you will find my cc_config and log file. the system has a 1500 watt PSU (hopefully enough) and is water cooled and my temps are never over 40C even after all night of crunching with all GPU. BTW the GPU are a pair of AMD 295x2's. I hope you can help me!

Log:
2/13/2015 6:20:23 AM | | Starting BOINC client version 7.4.36 for windows_x86_64
2/13/2015 6:20:23 AM | | log flags: file_xfer, sched_ops, task, coproc_debug
2/13/2015 6:20:23 AM | | Libraries: libcurl/7.39.0 OpenSSL/1.0.1j zlib/1.2.8
2/13/2015 6:20:23 AM | | Data directory: C:\ProgramData\BOINC
2/13/2015 6:20:23 AM | | Running under account Gorden
2/13/2015 6:20:23 AM | | [coproc] launching child process at C:\Program Files\BOINC\boinc.exe
2/13/2015 6:20:23 AM | | [coproc] relative to directory C:\ProgramData\BOINC
2/13/2015 6:20:23 AM | | [coproc] with data directory "C:\ProgramData\BOINC"
2/13/2015 6:20:23 AM | | OpenCL: AMD/ATI GPU 0: Hawaii (driver version 1642.5 (VM), device version OpenCL 2.0 AMD-APP (1642.5), 4096MB, 4096MB available, 3583 GFLOPS peak)
2/13/2015 6:20:23 AM | | OpenCL: AMD/ATI GPU 1: Hawaii (driver version 1642.5 (VM), device version OpenCL 1.2 AMD-APP (1642.5), 3072MB, 3072MB available, 3583 GFLOPS peak)
2/13/2015 6:20:23 AM | | OpenCL: AMD/ATI GPU 2: Hawaii (driver version 1642.5 (VM), device version OpenCL 1.2 AMD-APP (1642.5), 3072MB, 3072MB available, 3583 GFLOPS peak)
2/13/2015 6:20:23 AM | | OpenCL: AMD/ATI GPU 3: Hawaii (driver version 1642.5 (VM), device version OpenCL 1.2 AMD-APP (1642.5), 3072MB, 3072MB available, 3583 GFLOPS peak)
2/13/2015 6:20:23 AM | | OpenCL CPU: Intel(R) Core(TM) i7-5960X CPU @ 3.00GHz (OpenCL driver vendor: Advanced Micro Devices, Inc., driver version 1642.5 (sse2,avx), device version OpenCL 1.2 AMD-APP (1642.5))
2/13/2015 6:20:23 AM | | No NVIDIA library found
2/13/2015 6:20:23 AM | | calInit() returned 1
2/13/2015 6:20:23 AM | | Host name: Gorden-PC
2/13/2015 6:20:23 AM | | Processor: 16 GenuineIntel Intel(R) Core(TM) i7-5960X CPU @ 3.00GHz [Family 6 Model 63 Stepping 2]
2/13/2015 6:20:23 AM | | Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss htt tm pni ssse3 fma cx16 sse4_1 sse4_2 movebe popcnt aes f16c rdrandsyscall nx lm avx avx2 vmx tm2 dca pbe fsgsbase bmi1 smep bmi2
2/13/2015 6:20:23 AM | | OS: Microsoft Windows 7: Ultimate x64 Edition, Service Pack 1, (06.01.7601.00)
2/13/2015 6:20:23 AM | | Memory: 63.87 GB physical, 63.87 GB virtual
2/13/2015 6:20:23 AM | | Disk: 785.86 GB total, 518.66 GB free
2/13/2015 6:20:23 AM | | Local time is UTC -8 hours
2/13/2015 6:20:23 AM | | Config: use all coprocessors
2/13/2015 6:20:23 AM | Einstein@Home | URL http://einstein.phys.uwm.edu/; Computer ID 11745829; resource share 100
2/13/2015 6:20:23 AM | Milkyway@Home | URL http://milkyway.cs.rpi.edu/milkyway/; Computer ID 443594; resource share 100
2/13/2015 6:20:23 AM | SETI@home | URL http://setiathome.berkeley.edu/; Computer ID 6636246; resource share 100
2/13/2015 6:20:23 AM | SETI@home | General prefs: from SETI@home (last modified 08-Feb-2015 07:00:23)
2/13/2015 6:20:23 AM | SETI@home | Computer location: home
2/13/2015 6:20:23 AM | SETI@home | General prefs: no separate prefs for home; using your defaults
2/13/2015 6:20:23 AM | | Reading preferences override file
2/13/2015 6:20:23 AM | | Preferences:
2/13/2015 6:20:23 AM | | max memory usage when active: 32702.88MB
2/13/2015 6:20:23 AM | | max memory usage when idle: 58865.19MB
2/13/2015 6:20:23 AM | | max disk usage: 518.73GB
2/13/2015 6:20:23 AM | | max CPUs used: 14
2/13/2015 6:20:23 AM | | suspend work if non-BOINC CPU load exceeds 25%
2/13/2015 6:20:23 AM | | (to change preferences, visit a project web site or select Preferences in the Manager)
2/13/2015 6:20:23 AM | | Not using a proxy
2/13/2015 6:20:24 AM | SETI@home | [coproc] Assigning ATI instance 0 to 06oc12ad.16106.75387.438086664199.12.130_1
2/13/2015 6:20:24 AM | SETI@home | [coproc] Assigning ATI instance 1 to 06oc12ad.16106.75387.438086664199.12.134_0
2/13/2015 6:20:24 AM | SETI@home | [coproc] Assigning ATI instance 2 to 06oc12ad.16106.75387.438086664199.12.98_1
2/13/2015 6:20:24 AM | SETI@home | [coproc] Assigning ATI instance 3 to 06oc12ad.16106.75387.438086664199.12.60_1

cc_config:

<cc_config>
<log_flags>
<file_xfer>1</file_xfer>
<sched_ops>1</sched_ops>
<task>1</task>
<android_debug>0</android_debug>
<app_msg_receive>0</app_msg_receive>
<app_msg_send>0</app_msg_send>
<async_file_debug>0</async_file_debug>
<benchmark_debug>0</benchmark_debug>
<checkpoint_debug>0</checkpoint_debug>
<coproc_debug>1</coproc_debug>
<cpu_sched>0</cpu_sched>
<cpu_sched_debug>0</cpu_sched_debug>
<cpu_sched_status>0</cpu_sched_status>
<dcf_debug>0</dcf_debug>
<disk_usage_debug>0</disk_usage_debug>
<file_xfer_debug>0</file_xfer_debug>
<gui_rpc_debug>0</gui_rpc_debug>
<heartbeat_debug>0</heartbeat_debug>
<http_debug>0</http_debug>
<http_xfer_debug>0</http_xfer_debug>
<mem_usage_debug>0</mem_usage_debug>
<network_status_debug>0</network_status_debug>
<notice_debug>0</notice_debug>
<poll_debug>0</poll_debug>
<priority_debug>0</priority_debug>
<proxy_debug>0</proxy_debug>
<rr_simulation>0</rr_simulation>
<rrsim_detail>0</rrsim_detail>
<sched_op_debug>0</sched_op_debug>
<scrsave_debug>0</scrsave_debug>
<slot_debug>0</slot_debug>
<state_debug>0</state_debug>
<statefile_debug>0</statefile_debug>
<suspend_debug>0</suspend_debug>
<task_debug>0</task_debug>
<time_debug>0</time_debug>
<trickle_debug>0</trickle_debug>
<unparsed_xml>0</unparsed_xml>
<work_fetch_debug>0</work_fetch_debug>
</log_flags>
<options>
<abort_jobs_on_exit>0</abort_jobs_on_exit>
<allow_multiple_clients>0</allow_multiple_clients>
<allow_remote_gui_rpc>0</allow_remote_gui_rpc>
<client_version_check_url>http://boinc.berkeley.edu/download.php?xml=1</client_version_check_url>
<client_new_version_text></client_new_version_text>
<client_download_url>http://boinc.berkeley.edu/download.php</client_download_url>
<disallow_attach>0</disallow_attach>
<dont_check_file_sizes>0</dont_check_file_sizes>
<dont_contact_ref_site>0</dont_contact_ref_site>
<dont_use_vbox>0</dont_use_vbox>
<exit_after_finish>0</exit_after_finish>
<exit_before_start>0</exit_before_start>
<exit_when_idle>0</exit_when_idle>
<fetch_minimal_work>0</fetch_minimal_work>
<fetch_on_update>0</fetch_on_update>
<force_auth>default</force_auth>
<http_1_0>0</http_1_0>
<http_transfer_timeout>300</http_transfer_timeout>
<http_transfer_timeout_bps>10</http_transfer_timeout_bps>
<max_event_log_lines>2000</max_event_log_lines>
<max_file_xfers>8</max_file_xfers>
<max_file_xfers_per_project>2</max_file_xfers_per_project>
<max_stderr_file_size>0</max_stderr_file_size>
<max_stdout_file_size>0</max_stdout_file_size>
<max_tasks_reported>0</max_tasks_reported>
<ncpus>-1</ncpus>
<network_test_url>http://www.google.com/</network_test_url>
<no_alt_platform>0</no_alt_platform>
<no_gpus>0</no_gpus>
<no_info_fetch>0</no_info_fetch>
<no_priority_change>0</no_priority_change>
<os_random_only>0</os_random_only>
<proxy_info>
<socks_server_name></socks_server_name>
<socks_server_port>80</socks_server_port>
<http_server_name></http_server_name>
<http_server_port>80</http_server_port>
<socks5_user_name></socks5_user_name>
<socks5_user_passwd></socks5_user_passwd>
<http_user_name></http_user_name>
<http_user_passwd></http_user_passwd>
<no_proxy></no_proxy>
</proxy_info>
<rec_half_life_days>10.000000</rec_half_life_days>
<report_results_immediately>0</report_results_immediately>
<run_apps_manually>0</run_apps_manually>
<save_stats_days>30</save_stats_days>
<skip_cpu_benchmarks>0</skip_cpu_benchmarks>
<simple_gui_only>0</simple_gui_only>
<start_delay>0.000000</start_delay>
<stderr_head>0</stderr_head>
<suppress_net_info>0</suppress_net_info>
<unsigned_apps_ok>0</unsigned_apps_ok>
<use_all_gpus>1</use_all_gpus>
<use_certs>0</use_certs>
<use_certs_only>0</use_certs_only>
<vbox_window>0</vbox_window>
</options>
</cc_config>



I for any suggestions to stop the invalids!

Thank You

Regards
ET
ID: 1641235 · Report as offensive
ETWhereRU

Send message
Joined: 10 Mar 12
Posts: 11
Credit: 3,000,869
RAC: 0
United States
Message 1641244 - Posted: 13 Feb 2015, 14:38:45 UTC - in response to Message 1641235.  
Last modified: 13 Feb 2015, 14:48:00 UTC

Me again,

Just noticed that 3 out of 4 of my GPU are using openGL 1.2 and the other 2.0?? Also the memory of the others is low?? I also thought these card were good for a bit over 5 teraflops per GPU??? GPU-Z show 4G memory per card, seems that the older openCL 1.2 is reporting it wrong. Any thoughts?

Thank You

Regards
ET
ID: 1641244 · Report as offensive
Profile Zalster Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 27 May 99
Posts: 5517
Credit: 528,817,460
RAC: 242
United States
Message 1641247 - Posted: 13 Feb 2015, 14:40:41 UTC - in response to Message 1641235.  

ET

I'm guessing you are running stock applications since you said 3 projects are all getting invalids?

There are issues with the drivers for the ATI. But since I don't own any I'm not too familiar. I think TBar, Oz and some of the others are very familiar with the problem.

Some of them will probably be around later and will be able to point you to the correct drivers and have some ideas that will probably help you out.

Since it's still very early in the day I would give them till much later in the day to go thru here and review your post


Zalster
ID: 1641247 · Report as offensive
ETWhereRU

Send message
Joined: 10 Mar 12
Posts: 11
Credit: 3,000,869
RAC: 0
United States
Message 1641254 - Posted: 13 Feb 2015, 14:50:01 UTC - in response to Message 1641247.  

Good Morning,

Yes completely stock, downloaded about a week ago from Boinc.
ID: 1641254 · Report as offensive
Profile Zalster Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 27 May 99
Posts: 5517
Credit: 528,817,460
RAC: 242
United States
Message 1641270 - Posted: 13 Feb 2015, 15:18:16 UTC - in response to Message 1641254.  

It probably won't help you for any other projects, but if you install Lunatics optimized apps for Seti that might resolve the issue.


http://setiathome.berkeley.edu/forum_thread.php?id=71867&postid=1596404

first decide if you want to use your CPU or not, once you install, you want to install the ATI applications and skip over the nvidia section.

Or you can wait until one of the others comes around and comments on some ideas

Zalster
ID: 1641270 · Report as offensive
Profile Mr. Kevvy Crowdfunding Project Donor*Special Project $250 donor
Volunteer moderator
Volunteer tester
Avatar

Send message
Joined: 15 May 99
Posts: 3776
Credit: 1,114,826,392
RAC: 3,319
Canada
Message 1641273 - Posted: 13 Feb 2015, 15:27:23 UTC - in response to Message 1641244.  

a bit over 5 teraflops per GPU???


Egads... 5 teraflops would result in a RAC of one million. I think that may be a tad overstated. ;^)
ID: 1641273 · Report as offensive
ETWhereRU

Send message
Joined: 10 Mar 12
Posts: 11
Credit: 3,000,869
RAC: 0
United States
Message 1641280 - Posted: 13 Feb 2015, 15:37:57 UTC - in response to Message 1641273.  
Last modified: 13 Feb 2015, 15:42:48 UTC

Good Morning,

LOL, AMD states 11.5 Tflops per card... yes I also understand that a lot of these specs from the manufactures are optimistic to say the least. Even taking into consideration AMD's fudge factor I would hope to get a total 16 Tflops (compared to 23 Tflops spec) total from the 4 GPU's. How can I tell for sure the true Tflops?

At this point just want no errors and then will go for speed ;-)

Best Regards,
ET
ID: 1641280 · Report as offensive
Profile Mr. Kevvy Crowdfunding Project Donor*Special Project $250 donor
Volunteer moderator
Volunteer tester
Avatar

Send message
Joined: 15 May 99
Posts: 3776
Credit: 1,114,826,392
RAC: 3,319
Canada
Message 1641282 - Posted: 13 Feb 2015, 15:43:57 UTC
Last modified: 13 Feb 2015, 15:44:31 UTC

The "Cobblestone": One GFlOp running 24x7 will earn 200 credits per day. Your current RAC at the present is exactly 2,000 = 10 GFlOps.

Just to show how far computing has come, when I was a young'un in the 1980's the ultimate supercomputer was the Cray X-MP, worth about $15 million at the time. It does 400 MFlOps, so if you put it to work on a BOINC project it would have a rather underwhelming RAC of.... 80.
ID: 1641282 · Report as offensive
ETWhereRU

Send message
Joined: 10 Mar 12
Posts: 11
Credit: 3,000,869
RAC: 0
United States
Message 1641283 - Posted: 13 Feb 2015, 15:51:37 UTC - in response to Message 1641270.  

Hello,

Would like to get stock BOINC working, I know I am not the only with AMD cards and if I can get it working it will help others with similar AMD multi-GPU setups. Would like to hold off with Lunatics solution until it is determined BOINC will just not work.

Thank You!

Best Regards,
Gorden
ID: 1641283 · Report as offensive
ETWhereRU

Send message
Joined: 10 Mar 12
Posts: 11
Credit: 3,000,869
RAC: 0
United States
Message 1641286 - Posted: 13 Feb 2015, 15:58:16 UTC - in response to Message 1641282.  

Hello Kevvy,

Thanks for the how to calculate. Since I am borking about 50% of my results will the real number be closer to 20?

Best Regards,
ET
ID: 1641286 · Report as offensive
Profile Mr. Kevvy Crowdfunding Project Donor*Special Project $250 donor
Volunteer moderator
Volunteer tester
Avatar

Send message
Joined: 15 May 99
Posts: 3776
Credit: 1,114,826,392
RAC: 3,319
Canada
Message 1641292 - Posted: 13 Feb 2015, 16:04:43 UTC - in response to Message 1641286.  
Last modified: 13 Feb 2015, 16:04:55 UTC

Here is you on BOINCStats which shows your actual daily production not just RAC which is averaged. Daily. prod. there total is 21,283 which is >100 GFlOps... this is more reasonable.
ID: 1641292 · Report as offensive
ETWhereRU

Send message
Joined: 10 Mar 12
Posts: 11
Credit: 3,000,869
RAC: 0
United States
Message 1641300 - Posted: 13 Feb 2015, 16:22:11 UTC - in response to Message 1641292.  
Last modified: 13 Feb 2015, 16:35:56 UTC

Hello,

Thanks for the information! This is usually just 8 hours a day with me stealing a bit of CPU/GPU time for work/games. Will leave running over the next three days with no user activity to see just how much crunching it can do. Hope we can get my app_config tuned. From reading I should be able to double maybe triple work the GPU's and get some more crunching efficiency.

Concerning your last reply, yes it is incredible how far computing power has come. I remember when I thought a TRS-80 was fast 8-) Now we can to go to our local electronics store and build a system that has the processing power that was not even dreamed of 5 years ago. In another 5... could we hit YFlops (Yotta 10-24)?

Thanks Again!!

Best Regards,
ET
ID: 1641300 · Report as offensive
woohoo
Volunteer tester

Send message
Joined: 30 Oct 13
Posts: 972
Credit: 165,671,404
RAC: 5
United States
Message 1641334 - Posted: 13 Feb 2015, 17:42:01 UTC

I'm going to guess that you're running Catalyst 14.12. I didn't have any luck with 14.12 because BOINC would only see one of my three gpus. I'm on 14.11.2 Beta right now and it's working fine now. Other available versions are 13.12, 14.4 and 14.9 although for those that play games you might need at least 14.9 to get the most stability.
ID: 1641334 · Report as offensive
ETWhereRU

Send message
Joined: 10 Mar 12
Posts: 11
Credit: 3,000,869
RAC: 0
United States
Message 1641348 - Posted: 13 Feb 2015, 18:09:49 UTC - in response to Message 1641334.  
Last modified: 13 Feb 2015, 18:10:52 UTC

Hello,

Yes you are correct I am on 14.12 drivers

I had to add this line to my cc_config and now all my GPU's are being seen and used.

<use_all_gpus>1</use_all_gpus>



I am looking at BOINC manager and right now the I have 4 Einstein@Home running (0.5CPU's + 1 AMD/ATI GPU) on devices 0 through 3. If someone can show me how I can paste a screen shot if that will help.

Best Regards,
ET
ID: 1641348 · Report as offensive
Phil Burden

Send message
Joined: 26 Oct 00
Posts: 264
Credit: 22,303,899
RAC: 0
United Kingdom
Message 1641356 - Posted: 13 Feb 2015, 18:25:47 UTC - in response to Message 1641348.  

Hello,

Yes you are correct I am on 14.12 drivers

I had to add this line to my cc_config and now all my GPU's are being seen and used.

<use_all_gpus>1</use_all_gpus>



I am looking at BOINC manager and right now the I have 4 Einstein@Home running (0.5CPU's + 1 AMD/ATI GPU) on devices 0 through 3. If someone can show me how I can paste a screen shot if that will help.

Best Regards,
ET


You might want to try 13.12 drivers, I had issues (as have others) with later drivers. Just a thought.

P.
ID: 1641356 · Report as offensive
woohoo
Volunteer tester

Send message
Joined: 30 Oct 13
Posts: 972
Credit: 165,671,404
RAC: 5
United States
Message 1641367 - Posted: 13 Feb 2015, 18:41:49 UTC

In the past I've only used the use_all_gpus line when I had dissimilar gpus. So I suppose on one of my rigs that has identical 290Xs I could try the switch with 14.12. Another rig I have with one 290X and one 295X2 shouldn't really need the switch but I could try it. I know that the combination of 14.11.2 and my gpus and multiple wus on Einstein produces invalids, I might give 14.12 another whirl, although I'm not sure if I should try it with multiple wus or not. Or maybe try 13.12.
ID: 1641367 · Report as offensive
ETWhereRU

Send message
Joined: 10 Mar 12
Posts: 11
Credit: 3,000,869
RAC: 0
United States
Message 1641389 - Posted: 13 Feb 2015, 19:26:07 UTC - in response to Message 1641367.  

Hello,

Cant change over right now as I am in the middle of a work project 8-( Also I really don't want to have my system be just a cruncher and lose my ability to play games. On Monday I will give it a try if no other option presents itself. The other data point, I am getting valid results from the GPU, It seems logical if it was a driver issue all would be invalid??? I think (and can be completely off base) that the different OpenGL version attached to the other 3 GPU and incorrect memory size maybe a clue to the issue. I am downloading the AMD SDK to see if I can get all GPU's to run 2.0

Thank You all for the suggestions.

Best Regards,
ET
ID: 1641389 · Report as offensive
woohoo
Volunteer tester

Send message
Joined: 30 Oct 13
Posts: 972
Credit: 165,671,404
RAC: 5
United States
Message 1641417 - Posted: 13 Feb 2015, 20:34:28 UTC
Last modified: 13 Feb 2015, 20:35:00 UTC

I was just in the middle of trying to go to 13.12 with multiple wus until I hit a snag, probably due to the evolution of .NET, Windows 8.1 and Catalyst.

So I'm now on 14.12 on all of my rigs, and it turns out that the reason the <use_all_gpus>1</use_all_gpus> line is needed is because the first gpu is detected as opencl 2.0 and the remaining are assigned 1.2 so at that point BOINC thinks they're inferior. I'll have to run this setup for a few days to see if it generates any more invalids.

Keep in mind that I've disabled crossfire. This doesn't affect me in the traditional sense because I don't play games, but it crossfire synchronizes everything and I don't like that. In my rig with two identical gpus, the top card runs hotter than the bottom card so I want to give the two cards the ability to perform independently of each other. In my rig with three gpus, two of the gpus have a lower clock speed and less pcie bandwidth than the last gpu, so I want to let that last gpu cruch faster. In your setup everything is identical, pcie bandwidth is equal, clock speeds are the same and temperatures are cool so you should be able to get away with crossfire. Now what would be a real bummer would be if your system was generating invalids because of the two different opencl versions. In that case you could go to an earlier driver.
ID: 1641417 · Report as offensive
ETWhereRU

Send message
Joined: 10 Mar 12
Posts: 11
Credit: 3,000,869
RAC: 0
United States
Message 1641478 - Posted: 13 Feb 2015, 22:37:08 UTC - in response to Message 1641417.  
Last modified: 13 Feb 2015, 22:39:42 UTC

Hello,

I also disable cross fire while crunching, if I want to play just pause Boinc, enable crossfire and no issue. That makes sense about the cc_config option and it is the same as I am seeing. I cant take down my system now as it is doing some video conversions however this afternoon will load up the AMD SDK as it has opengl 2.0.

Best Regards,
ET
ID: 1641478 · Report as offensive
woohoo
Volunteer tester

Send message
Joined: 30 Oct 13
Posts: 972
Credit: 165,671,404
RAC: 5
United States
Message 1641514 - Posted: 13 Feb 2015, 23:09:31 UTC

I just installed the sdk on my three rigs and every gpu after the first still shows up as just opengl 1.2
ID: 1641514 · Report as offensive
1 · 2 · Next

Message boards : Number crunching : New system many invalids and errors


 
©2024 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.