Projects of the Month for March 2018 - Page 3 - Overclock.net - An Overclocking Community

Forum Jump: 

Projects of the Month for March 2018

Reply
 
Thread Tools
post #21 of 36 (permalink) Old 03-08-2018, 01:44 AM
New to Overclock.net
 
mmonnin's Avatar
 
Join Date: Nov 2012
Posts: 5,826
Rep: 282 (Unique: 135)
I'm currently on 384.111 in Linux 17.04 and its all working fine but with two NV cards.

GPUGrid uses CUDA and is only going to look for NV cards. Couple differences there.

What if a 2nd client was setup for either the AMD or NV card? Tasks could be separated with a Collatz location for the separate host or with exclude GPU in cc_config. It does look like its Collatz not being able to use the right task/files for the right GPU.


mmonnin is online now  
Sponsored Links
Advertisement
 
post #22 of 36 (permalink) Old 03-08-2018, 02:20 AM
BOINC Cruncher
 
Diffident's Avatar
 
Join Date: Apr 2014
Location: Bawlmer, MD
Posts: 1,562
Rep: 98 (Unique: 66)
Quote: Originally Posted by tictoc View Post
Is anyone running Collatz in Linux on the 390.25 NVIDIA drivers? I am instantly failing units on my 1070s with the following errors:

Stderr output of failed task:
Code:
201 (0xc9) EXIT_MISSING_COPROC
BOINC Log of failed task:
Code:
[Collatz Conjecture] Missing coprocessor for task collatz_sieve_4103907764157913497600_52776558133248_0; aborting
Beginning of BOINC log:
Code:
04-Mar-2018 18:29:28 [---] Starting BOINC client version 7.8.4 for x86_64-pc-linux-gnu
04-Mar-2018 18:29:28 [---] log flags: file_xfer, sched_ops, task
04-Mar-2018 18:29:28 [---] Libraries: libcurl/7.58.0 OpenSSL/1.1.0g zlib/1.2.11 libidn2/2.0.4 libpsl/0.19.1 (+libidn2/2.0.4) nghttp2/1.30.0
04-Mar-2018 18:29:28 [---] Data directory: /var/lib/boinc
04-Mar-2018 18:29:29 [---] CUDA: NVIDIA GPU 0: GeForce GTX 1070 (driver version 390.25, CUDA version 9.1, compute capability 6.1, 4096MB, 3984MB available, 6852 GFLOPS peak)
04-Mar-2018 18:29:29 [---] CUDA: NVIDIA GPU 1: GeForce GTX 1070 (driver version 390.25, CUDA version 9.1, compute capability 6.1, 4096MB, 3984MB available, 6803 GFLOPS peak)
04-Mar-2018 18:29:29 [---] OpenCL: NVIDIA GPU 0: GeForce GTX 1070 (driver version 390.25, device version OpenCL 1.2 CUDA, 8118MB, 3984MB available, 6852 GFLOPS peak)
04-Mar-2018 18:29:29 [---] OpenCL: NVIDIA GPU 1: GeForce GTX 1070 (driver version 390.25, device version OpenCL 1.2 CUDA, 8120MB, 3984MB available, 6803 GFLOPS peak)
04-Mar-2018 18:29:29 [---] OpenCL: AMD/ATI GPU 0: AMD Radeon R9 200 Series (driver version 2527.3, device version OpenCL 1.2 AMD-APP (2527.3), 4086MB, 4086MB available, 5120 GFLOPS peak)
04-Mar-2018 18:29:29 [---] OpenCL: AMD/ATI GPU 1: AMD Radeon R9 200 Series (driver version 2527.3, device version OpenCL 1.2 AMD-APP (2527.3), 4086MB, 4086MB available, 5120 GFLOPS peak)
04-Mar-2018 18:29:30 [---] Host name: TickTockArch
04-Mar-2018 18:29:30 [---] Processor: 40 GenuineIntel Genuine Intel(R) CPU @ 2.30GHz [Family 6 Model 62 Stepping 2]
04-Mar-2018 18:29:30 [---] 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 ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc cpuid aperfmperf pni pclmulqdq dtes64 monitor ds_cpl vmx smx est tm2 ssse3 cx16 xtpr pdcm pcid dca sse4_1 sse4_2 x2apic popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm epb pti intel_ppin tpr_shadow vnmi flexpriority ept vpid fsgsbase smep erms xsaveopt dtherm ida arat pln pts
04-Mar-2018 18:29:30 [---] OS: Linux Arch: Arch Linux [4.15.6-1-ARCH]
04-Mar-2018 18:29:30 [---] Memory: 125.94 GB physical, 8.00 GB virtual
04-Mar-2018 18:29:30 [---] Disk: 234.01 GB total, 154.64 GB free
04-Mar-2018 18:29:30 [---] Local time is UTC -7 hours
04-Mar-2018 18:29:30 [---] VirtualBox version: 5.2.8r120774
04-Mar-2018 18:29:30 [---] Config: allow multiple clients
04-Mar-2018 18:29:30 [---] Config: GUI RPC allowed from any host
04-Mar-2018 18:29:30 [---] Config: use all coprocessors
I took a quick look at the top participants, and at least one person is running the 390.25 driver and completing tasks. My 290 is running Collatz without issue. I have a feeling it has to do with the mixed GPUs in my system, and Collatz somehow losing it's mind even though it does send me opencl_nvidia_gpu tasks for the 1070. I have no issues running other projects on my 1070 (currently running GPUGrid).
The task not being able to find the card usually happens after a driver update, rebooting should fix it.

Team Boinc! - Forum Folding War Team
Gentoo Linux - Telemetry Free Since 1999!
Washington Capitals 2018 Stanley Cup Champions
Boinc Desktop
(15 items)
CrunchAholic
(10 items)
NAS
(7 items)
CPU
AMD Ryzen 7 1700X
Motherboard
ASRock X370 Taichi
GPU
AMD Radeon VII
RAM
HyperX Predator
Hard Drive
Samsung 970 EVOPlus
Power Supply
EVGA 850 G2
Cooling
Watercool Heatkiller IV PRO AM4 CPU Block
Cooling
EKWB Coolstream XE 240 Radiator
Cooling
EKWB Coolstream PE 360 Radiator
Cooling
Watercool Heatkiller 150mm Tube Res
Case
Thermaltake View 31 Tempered Glass RGB Edition
Operating System
Gentoo Linux
Monitor
Alienware AW3418DW
Keyboard
Coolermaster Masterkeys Pro S
Mouse
Roccat Kone EMP
CPU
Intel Xeon E5-2670
CPU
Intel Xeon E5-2670
Motherboard
Asrock Rack EP2C602
GPU
EVGA 1070ti FTW
GPU
EVGA 1070ti FTW
RAM
64 GBs Samsung M939B1K70CHD-CH9 PC3-10600R
Hard Drive
Samsung HD322HJ
Power Supply
Corsair AX1200
Case
Rosewill RSV-L4000
Operating System
Gentoo Linux
CPU
i3-4130T
Motherboard
GIGABYTE GA-H97N-WIFI
RAM
CORSAIR CMX16GX3M2A1600C11
Hard Drive
Western Digital Red 10 TB x3
Hard Drive
Western Digital Red 8 TB
Case
Fractal Design Node 804
Operating System
Gentoo Linux
▲ hide details ▲


Diffident is offline  
post #23 of 36 (permalink) Old 03-08-2018, 01:21 PM
Retired Staff
 
Tex1954's Avatar
 
Join Date: Feb 2011
Location: In a house older than me...
Posts: 4,462
Rep: 100 (Unique: 66)
Quote: Originally Posted by Diffident View Post
I have 64G of ECC, so... I have enough.

I never had any problems before, most of my points for NFS are from my 2p.
I just upgraded my 2P to Linux Mint 18.3 and had some problems. Had to load build-essential libraries to make it work right... not sure what you are running, but good luck!



Tex1954 is offline  
Sponsored Links
Advertisement
 
post #24 of 36 (permalink) Old 03-09-2018, 09:02 PM - Thread Starter
2+2=5
 
tictoc's Avatar
 
Join Date: Feb 2011
Posts: 4,474
@ku4jb Any idea how often dDM generates new tasks? There were tasks all of February, but it looks like January was dry. The site is such a mess that there's no easy way to see how often new tasks are generated.


tictoc is offline  
post #25 of 36 (permalink) Old 03-09-2018, 09:13 PM
New to Overclock.net
 
mmonnin's Avatar
 
Join Date: Nov 2012
Posts: 5,826
Rep: 282 (Unique: 135)
mmonnin is online now  
post #26 of 36 (permalink) Old 03-09-2018, 10:41 PM - Thread Starter
2+2=5
 
tictoc's Avatar
 
Join Date: Feb 2011
Posts: 4,474
Quote: Originally Posted by mmonnin View Post
I'm currently on 384.111 in Linux 17.04 and its all working fine but with two NV cards.

GPUGrid uses CUDA and is only going to look for NV cards. Couple differences there.

What if a 2nd client was setup for either the AMD or NV card? Tasks could be separated with a Collatz location for the separate host or with exclude GPU in cc_config. It does look like its Collatz not being able to use the right task/files for the right GPU.
I used one of the other clients that I already had setup, and Collatz is now running on my 1070 without issue. Before I aborted the AMD task that it grabbed when the client first communicated with the server, the secondary client was working on my 290 and my 1070, so I'm not sure why Collatz has issues on my primary client.


tictoc is offline  
post #27 of 36 (permalink) Old 03-10-2018, 06:07 AM
New to Overclock.net
 
mmonnin's Avatar
 
Join Date: Nov 2012
Posts: 5,826
Rep: 282 (Unique: 135)
Since DDM keeps running out of work and there is a task limit I tried attaching it to some more clients. That just made all tasks abort with a 'Client detached' error. On all clients, the 1st one and the extras. None of the extra ones show up as hosts on the profile either. I've never seen that happen before.

http://www.distributeddatamining.org...ultid=42508457


mmonnin is online now  
post #28 of 36 (permalink) Old 03-10-2018, 08:10 AM - Thread Starter
2+2=5
 
tictoc's Avatar
 
Join Date: Feb 2011
Posts: 4,474
Quote: Originally Posted by mmonnin View Post
Since DDM keeps running out of work and there is a task limit I tried attaching it to some more clients. That just made all tasks abort with a 'Client detached' error. On all clients, the 1st one and the extras. None of the extra ones show up as hosts on the profile either. I've never seen that happen before.

http://www.distributeddatamining.org...ultid=42508457
That is a new one for me. About the only way around that, if the dDM server is not assigning a new hostid to the second client, would be to run it in a few VMs. That's not as clean and simple as multiple clients, but it would work.

Tasks are available again for dDM.


tictoc is offline  
post #29 of 36 (permalink) Old 03-10-2018, 09:02 AM
New to Overclock.net
 
ku4jb's Avatar
 
Join Date: Apr 2010
Posts: 307
Rep: 25 (Unique: 24)
Quote: Originally Posted by tictoc View Post
@ku4jb Any idea how often dDM generates new tasks? There were tasks all of February, but it looks like January was dry. The site is such a mess that there's no easy way to see how often new tasks are generated.
I have no idea. Only pattern seems to be when they're low on tasks, "resends" come in at night for a few tasks each day.

As mentioned, I think ocn has crushed this latest batch. Not sure posting in forums would get a response either...I envision a small basement office with nobody around and the only sound is a buzzing noise from a flickering light..


ku4jb is offline  
post #30 of 36 (permalink) Old 03-10-2018, 10:19 AM
New to Overclock.net
 
mmonnin's Avatar
 
Join Date: Nov 2012
Posts: 5,826
Rep: 282 (Unique: 135)
I noticed the queue was down to 178 tasks, below the tasks in progress limit. Out of work again. Seems to come quicker and quicker each day.


mmonnin is online now  
Reply

Quick Reply
Message:
Options

Register Now

In order to be able to post messages on the Overclock.net - An Overclocking Community forums, you must first register.
Please enter your desired user name, your email address and other required details in the form below.
User Name:
If you do not want to register, fill this field only and the name will be used as user name for your post.
Password
Please enter a password for your user account. Note that passwords are case-sensitive.
Password:
Confirm Password:
Email Address
Please enter a valid email address for yourself.
Email Address:

Log-in



Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 
Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may post new threads
You may post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off