Welcome to MilkyWay@home

Posts by bhorlor

1) Message boards : Number crunching : What's Everybody Doing with their Double Precision These Days? (Message 76408)
Posted 4 Oct 2023 by Profile bhorlor
Post:
Good call. I had not selected that application. As soon as I did I started downloading GPU WUs. I wish it was easier to tell which ones are intended for CPUs and which are for GPUs (only 2 are explicit), but at least this got me what I wanted.
Thanks!
2) Message boards : Number crunching : What's Everybody Doing with their Double Precision These Days? (Message 76403)
Posted 2 Oct 2023 by Profile bhorlor
Post:
Now that the GPU support here is done, what projects are people doing with their AMD GPUs? It looks like MW@H was the only BOINC project that took advantage of the double precision, and WCG and Einstein rarely send any WUs.
3) Message boards : News : Nbody WU Flush (Message 73620)
Posted 20 May 2022 by Profile bhorlor
Post:
I see, CPU separations vs. GPU separations. My original interpretation was that CPUs are worthless here.
Thanks for the clarification.
4) Message boards : News : Nbody WU Flush (Message 73609)
Posted 20 May 2022 by Profile bhorlor
Post:
I think you should bring every single one of your cpu cores here RIGHT NOW to help clear them out!!
I'd love to but the big CPUs have got GPUs running Separation. If I try to get Nbody for those CPUs, the server gives me CPU seperations! Which are tremendously slow compared to GPUs and utterly pointless. I really don't understand why we don't have more options in the server preferences. Einstein manages.

Something is making the whole server sluggish, it's taking forever just to post this message, not sure what's changed.

Although I do notice the number of seperations waiting to go out is 30000 instead of the usual 10000. Don't tell me we're going to get a huge mass of those to clear too (although that should be easier as we can do them on GPUs).

Why are they pointless?
5) Questions and Answers : Preferences : Multiple GPU WUs simultaneously (Message 72894)
Posted 16 Apr 2022 by Profile bhorlor
Post:
Well, poop.
4/16/2022 1:41:30 PM | Milkyway@Home | Found app_config.xml

Yep, that did it. I thought that file went in the main directory, not the individual project folder.
6) Questions and Answers : Preferences : Multiple GPU WUs simultaneously (Message 72891)
Posted 16 Apr 2022 by Profile bhorlor
Post: 
4/16/2022 12:23:51 PM | Milkyway@Home | Computation for task de_modfit_70_bundle5_3s_south_pt2_3_1649967544_1323994_0 finished
4/16/2022 12:23:51 PM | Milkyway@Home | Starting task de_modfit_78_bundle5_3s_south_pt2_2_1647295262_6578924_3
4/16/2022 12:24:03 PM | Milkyway@Home | Computation for task de_modfit_70_bundle5_3s_south_pt2_3_1649967544_1323998_0 finished
4/16/2022 12:24:03 PM | Milkyway@Home | Starting task de_modfit_73_bundle5_3s_south_pt2_2_1649967544_885559_2
4/16/2022 12:24:41 PM |  | Re-reading cc_config.xml
4/16/2022 12:24:41 PM |  | log flags: file_xfer, sched_ops, task
4/16/2022 12:24:52 PM | Milkyway@Home | Computation for task de_modfit_70_bundle5_3s_south_pt2_2_1649967544_1367267_0 finished
4/16/2022 12:24:52 PM | Milkyway@Home | Starting task de_modfit_85_bundle5_3s_south_pt2_2_1649367564_2914905_2
4/16/2022 12:24:59 PM | Milkyway@Home | Computation for task de_modfit_70_bundle5_3s_south_pt2_3_1649967544_1323995_0 finished
4/16/2022 12:24:59 PM | Milkyway@Home | Starting task de_modfit_79_bundle5_3s_south_pt2_2_1647295262_11414174_3

And I only edit using Notepad++, and the file extensions, where appropriate, are .xml. The log didn't show that it's reading the app_config file at all.

I thought to check the log after restarting BOINC to see if that provided any illumination:

4/16/2022 12:28:10 PM |  | Starting BOINC client version 7.16.11 for windows_x86_64
4/16/2022 12:28:10 PM |  | log flags: file_xfer, sched_ops, task
4/16/2022 12:28:10 PM |  | Libraries: libcurl/7.47.1 OpenSSL/1.0.2s zlib/1.2.8
4/16/2022 12:28:10 PM |  | Data directory: C:\ProgramData\BOINC
4/16/2022 12:28:10 PM |  | Running under account [redacted]
4/16/2022 12:28:10 PM |  | CUDA: NVIDIA GPU 0: GeForce RTX 3090 (driver version 456.71, CUDA version 11.1, compute capability 8.6, 4096MB, 3520MB available, 18893 GFLOPS peak)
4/16/2022 12:28:10 PM |  | OpenCL: NVIDIA GPU 0: GeForce RTX 3090 (driver version 456.71, device version OpenCL 1.2 CUDA, 24576MB, 3520MB available, 18893 GFLOPS peak)
4/16/2022 12:28:13 PM |  | Windows processor group 0: 32 processors
4/16/2022 12:28:13 PM |  | Host name: [redacted]
4/16/2022 12:28:13 PM |  | Processor: 32 AuthenticAMD AMD Ryzen 9 5950X 16-Core Processor [Family 25 Model 33 Stepping 0]
4/16/2022 12:28:13 PM |  | Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 htt pni ssse3 fma cx16 sse4_1 sse4_2 movebe popcnt aes f16c rdrandsyscall nx lm avx avx2 svm sse4a osvw ibs skinit wdt tce topx page1gb rdtscp fsgsbase bmi1 smep bmi2
4/16/2022 12:28:13 PM |  | OS: Microsoft Windows 10: Professional x64 Edition, (10.00.19043.00)
4/16/2022 12:28:13 PM |  | Memory: 31.92 GB physical, 36.78 GB virtual
4/16/2022 12:28:13 PM |  | Disk: 222.94 GB total, 29.62 GB free
4/16/2022 12:28:13 PM |  | Local time is UTC -4 hours
4/16/2022 12:28:13 PM |  | WSL detected:
4/16/2022 12:28:13 PM |  | [Debian]: Linux Debian (Debian GNU/Linux 9.13 (stretch) [4.4.0-19041-Microsoft])
4/16/2022 12:28:13 PM |  | [Ubuntu] (default): Linux Ubuntu (Ubuntu 20.04.2 LTS [4.4.0-19041-Microsoft])
4/16/2022 12:28:13 PM |  | No general preferences found - using defaults
4/16/2022 12:28:13 PM |  | Reading preferences override file
4/16/2022 12:28:13 PM |  | Preferences:
4/16/2022 12:28:13 PM |  | max memory usage when active: 21896.54 MB
4/16/2022 12:28:13 PM |  | max memory usage when idle: 29413.26 MB
4/16/2022 12:28:13 PM |  | max disk usage: 31.65 GB
4/16/2022 12:28:13 PM |  | max CPUs used: 21
4/16/2022 12:28:13 PM |  | (to change preferences, visit a project web site or select Preferences in the Manager)
4/16/2022 12:28:13 PM |  | Setting up project and slot directories
4/16/2022 12:28:13 PM |  | Checking active tasks
4/16/2022 12:28:13 PM | Einstein@Home | URL http://einstein.phys.uwm.edu/; Computer ID 12925904; resource share 100
4/16/2022 12:28:13 PM | Milkyway@Home | URL http://milkyway.cs.rpi.edu/milkyway/; Computer ID 843393; resource share 100
4/16/2022 12:28:13 PM | Rosetta@home | URL https://boinc.bakerlab.org/rosetta/; Computer ID 3936402; resource share 25
4/16/2022 12:28:13 PM |  | Setting up GUI RPC socket
4/16/2022 12:28:13 PM |  | Checking presence of 266 project files
7) Questions and Answers : Preferences : Multiple GPU WUs simultaneously (Message 72860)
Posted 15 Apr 2022 by Profile bhorlor
Post:
Tried it and there's no effect. I'm really starting to think Milkyway doesn't want me to use my 3090.
8) Questions and Answers : Preferences : Multiple GPU WUs simultaneously (Message 72837)
Posted 15 Apr 2022 by Profile bhorlor
Post:
I've been looking at how to run multiple WUs on my GPU with MW like I do with Einstein. I've copied a bunch of people's app_config and it doesn't change anything: I still only run 1 GPU process at a time. Einstein required me to do it through the website (app_config didn't work by itself until I changed my computing preferences), but I don't see a similar way on MW to do it. My app_config is below, though I've tried a bunch of different things:
name=milkyway, milkyway@home, Milkyway@Home, MilkyWay@Home
concurrent=0, 1, 2, 5, 10, 50
gpu_usage=0.5, 0.25, 0.125, 0.1, 0.05
cpu_usage=0.5, 0.25, 0.125, 0.1, 0.05

Not every combination, obviously, but a bunch, and they all result in me doing a single GPU WU.

<app_config>
<app>
<name>milkyway</name>
<max_concurrent>50</max_concurrent>
<gpu_versions>
<gpu_usage>0.125</gpu_usage>
<cpu_usage>0.05</cpu_usage>
</gpu_versions>
</app>
</app_config>
9) Questions and Answers : Windows : Using Multiple GPUs (Message 69915)
Posted 12 Jun 2020 by Profile bhorlor
Post:
I have a 2080 ti & a 1660 super in my computer. I have used the use_all_gpu option so that both will run. However, even though it says Running (0.955 CPUs + 1 NVIDIA GPU(Device 0)) and ...(Device 1)), it's actually using the 2080 only (though twice, it seems). I only noticed because I'm monitoring the temperatures and the 2080's goes up while the 1660's stays flat.

If I tell it to ignore_nvidia_dev 0, then I can get it to use the 1660, but I can't seem to find a way to get it to run both simultaneously. If it ignores the 2080, then the 1660 will run just fine (CUDA utilization 80%+, power draw increases, temperature increases).

I've included the event log for having use_all_gpus enabled. I can confirm through thermals, power from the power supply, and Windows task manager that only the 2080 is being used.
Telling it to ignore either GPU gives an extra line in the log stating it's being ignored.

Thanks in advance for any help.

6/12/2020 10:41:14 AM | | Starting BOINC client version 7.16.7 for windows_x86_64
6/12/2020 10:41:14 AM | | Libraries: libcurl/7.47.1 OpenSSL/1.0.2s zlib/1.2.8
6/12/2020 10:41:14 AM | | Data directory: C:\ProgramData\BOINC
6/12/2020 10:41:14 AM | | Running under account [redacted]
6/12/2020 10:41:14 AM | | CUDA: NVIDIA GPU 0: GeForce RTX 2080 Ti (driver version 446.14, CUDA version 11.0, compute capability 7.5, 4096MB, 3539MB available, 14231 GFLOPS peak)
6/12/2020 10:41:14 AM | | CUDA: NVIDIA GPU 1: GeForce GTX 1660 SUPER (driver version 446.14, CUDA version 11.0, compute capability 7.5, 4096MB, 3555MB available, 5111 GFLOPS peak)
6/12/2020 10:41:14 AM | | OpenCL: NVIDIA GPU 0: GeForce RTX 2080 Ti (driver version 446.14, device version OpenCL 1.2 CUDA, 11264MB, 3539MB available, 14231 GFLOPS peak)
6/12/2020 10:41:14 AM | | OpenCL: NVIDIA GPU 0: GeForce RTX 2080 Ti (driver version 446.14, device version OpenCL 1.2 CUDA, 11264MB, 3539MB available, 14231 GFLOPS peak)
6/12/2020 10:41:14 AM | | OpenCL: NVIDIA GPU 1: GeForce GTX 1660 SUPER (driver version 446.14, device version OpenCL 1.2 CUDA, 6144MB, 3555MB available, 5111 GFLOPS peak)
6/12/2020 10:41:14 AM | | OpenCL: NVIDIA GPU 1: GeForce GTX 1660 SUPER (driver version 446.14, device version OpenCL 1.2 CUDA, 6144MB, 3555MB available, 5111 GFLOPS peak)
6/12/2020 10:41:14 AM | | Version change (7.16.5 -> 7.16.7)
6/12/2020 10:41:14 AM | | Windows processor group 0: 24 processors
6/12/2020 10:41:14 AM | | Host name: [redacted]
6/12/2020 10:41:14 AM | | Processor: 24 AuthenticAMD AMD Ryzen 9 3900X 12-Core Processor [Family 23 Model 113 Stepping 0]
6/12/2020 10:41:14 AM | | Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 htt pni ssse3 fma cx16 sse4_1 sse4_2 movebe popcnt aes f16c rdrandsyscall nx lm avx avx2 svm sse4a osvw ibs skinit wdt tce topx page1gb rdtscp fsgsbase bmi1 smep bmi2
6/12/2020 10:41:14 AM | | OS: Microsoft Windows 10: Core x64 Edition, (10.00.18363.00)
6/12/2020 10:41:14 AM | | Memory: 31.95 GB physical, 36.70 GB virtual
6/12/2020 10:41:14 AM | | Disk: 222.94 GB total, 81.93 GB free
6/12/2020 10:41:14 AM | | Local time is UTC -4 hours
6/12/2020 10:41:14 AM | | No WSL found.
6/12/2020 10:41:14 AM | | Config: use all coprocessors
6/12/2020 10:41:14 AM | | No general preferences found - using defaults
6/12/2020 10:41:14 AM | | Reading preferences override file
6/12/2020 10:41:14 AM | | Preferences:
6/12/2020 10:41:14 AM | | max memory usage when active: 16357.44 MB
6/12/2020 10:41:14 AM | | max memory usage when idle: 29443.39 MB
6/12/2020 10:41:14 AM | | max disk usage: 83.62 GB
6/12/2020 10:41:14 AM | | max CPUs used: 18
6/12/2020 10:41:14 AM | | suspend work if non-BOINC CPU load exceeds 25%
6/12/2020 10:41:14 AM | | (to change preferences, visit a project web site or select Preferences in the Manager)
6/12/2020 10:41:14 AM | | Setting up project and slot directories
6/12/2020 10:41:14 AM | | Checking active tasks
6/12/2020 10:41:14 AM | Milkyway@Home | URL http://milkyway.cs.rpi.edu/milkyway/; Computer ID 843393; resource share 100
6/12/2020 10:41:14 AM | Rosetta@home | URL https://boinc.bakerlab.org/rosetta/; Computer ID 3936402; resource share 25
6/12/2020 10:41:14 AM | | Setting up GUI RPC socket
6/12/2020 10:41:14 AM | | Checking presence of 103 project files
6/12/2020 10:41:14 AM | | Running CPU benchmarks
6/12/2020 10:41:45 AM | | Benchmark results:
6/12/2020 10:41:45 AM | | Number of CPUs: 18
6/12/2020 10:41:45 AM | | 5393 floating point MIPS (Whetstone) per CPU
6/12/2020 10:41:45 AM | | 19135 integer MIPS (Dhrystone) per CPU




©2024 Astroinformatics Group