climateprediction.net home page
Posts by lazlo_vii

Posts by lazlo_vii

21) Message boards : Number crunching : The Brute Squad (Message 63037)
Posted 27 Nov 2020 by lazlo_vii
Post:
I am working on bringing them up now. I threatened to write down my setup and post it here last year. Maybe I'll do that today.
22) Message boards : Number crunching : The Brute Squad (Message 63030)
Posted 27 Nov 2020 by lazlo_vii
Post:
It's on the way.

They WCG queue needs to empy first, but it is on the way:

laz@bsquad-host-1:~$ lscpu
Architecture:                    x86_64
CPU op-mode(s):                  32-bit, 64-bit
Byte Order:                      Little Endian
Address sizes:                   43 bits physical, 48 bits virtual
CPU(s):                          32
On-line CPU(s) list:             0-31
Thread(s) per core:              2
Core(s) per socket:              16
Socket(s):                       1
NUMA node(s):                    1
Vendor ID:                       AuthenticAMD
CPU family:                      23
Model:                           113
Model name:                      AMD Ryzen 9 3950X 16-Core Processor
Stepping:                        0
Frequency boost:                 enabled
CPU MHz:                         4053.346
CPU max MHz:                     3500.0000
CPU min MHz:                     2200.0000
BogoMIPS:                        6986.96
Virtualization:                  AMD-V
L1d cache:                       512 KiB
L1i cache:                       512 KiB
L2 cache:                        8 MiB
L3 cache:                        64 MiB
NUMA node0 CPU(s):               0-31
Vulnerability Itlb multihit:     Not affected
Vulnerability L1tf:              Not affected
Vulnerability Mds:               Not affected
Vulnerability Meltdown:          Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl and seccomp
Vulnerability Spectre v1:        Mitigation; usercopy/swapgs barriers and __user pointer sanitization
Vulnerability Spectre v2:        Mitigation; Full AMD retpoline, IBPB conditional, STIBP conditional, RSB filling
Vulnerability Srbds:             Not affected
Vulnerability Tsx async abort:   Not affected
Flags:                           fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxs
                                 r sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl n
                                 onstop_tsc cpuid extd_apicid aperfmperf pni pclmulqdq monitor ssse3 fma cx16 sse4_1 sse
                                 4_2 movbe popcnt aes xsave avx f16c rdrand lahf_lm cmp_legacy svm extapic cr8_legacy ab
                                 m sse4a misalignsse 3dnowprefetch osvw ibs skinit wdt tce topoext perfctr_core perfctr_
                                 nb bpext perfctr_llc mwaitx cpb cat_l3 cdp_l3 hw_pstate sme ssbd mba sev ibpb stibp vmm
                                 call fsgsbase bmi1 avx2 smep bmi2 cqm rdt_a rdseed adx smap clflushopt clwb sha_ni xsav
                                 eopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local clzero irp
                                 erf xsaveerptr wbnoinvd arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyas
                                 id decodeassists pausefilter pfthreshold avic v_vmsave_vmload vgif umip rdpid overflow_
                                 recov succor smca

laz@bsquad-host-2:~$ lscpu
Architecture:                    x86_64
CPU op-mode(s):                  32-bit, 64-bit
Byte Order:                      Little Endian
Address sizes:                   43 bits physical, 48 bits virtual
CPU(s):                          16
On-line CPU(s) list:             0-15
Thread(s) per core:              2
Core(s) per socket:              8
Socket(s):                       1
NUMA node(s):                    1
Vendor ID:                       AuthenticAMD
CPU family:                      23
Model:                           113
Model name:                      AMD Ryzen 7 3700X 8-Core Processor
Stepping:                        0
Frequency boost:                 enabled
CPU MHz:                         2199.305
CPU max MHz:                     3600.0000
CPU min MHz:                     2200.0000
BogoMIPS:                        7200.14
Virtualization:                  AMD-V
L1d cache:                       256 KiB
L1i cache:                       256 KiB
L2 cache:                        4 MiB
L3 cache:                        32 MiB
NUMA node0 CPU(s):               0-15
Vulnerability Itlb multihit:     Not affected
Vulnerability L1tf:              Not affected
Vulnerability Mds:               Not affected
Vulnerability Meltdown:          Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl and seccomp
Vulnerability Spectre v1:        Mitigation; usercopy/swapgs barriers and __user pointer sanitization
Vulnerability Spectre v2:        Mitigation; Full AMD retpoline, IBPB conditional, STIBP conditional, RSB filling
Vulnerability Srbds:             Not affected
Vulnerability Tsx async abort:   Not affected
Flags:                           fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxs
                                 r sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl n
                                 onstop_tsc cpuid extd_apicid aperfmperf pni pclmulqdq monitor ssse3 fma cx16 sse4_1 sse
                                 4_2 movbe popcnt aes xsave avx f16c rdrand lahf_lm cmp_legacy svm extapic cr8_legacy ab
                                 m sse4a misalignsse 3dnowprefetch osvw ibs skinit wdt tce topoext perfctr_core perfctr_
                                 nb bpext perfctr_llc mwaitx cpb cat_l3 cdp_l3 hw_pstate sme ssbd mba sev ibpb stibp vmm
                                 call fsgsbase bmi1 avx2 smep bmi2 cqm rdt_a rdseed adx smap clflushopt clwb sha_ni xsav
                                 eopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local clzero irp
                                 erf xsaveerptr wbnoinvd arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyas
                                 id decodeassists pausefilter pfthreshold avic v_vmsave_vmload vgif umip rdpid overflow_
                                 recov succor smca

laz@bsquad-host-3:~$ lscpu
Architecture:                    x86_64
CPU op-mode(s):                  32-bit, 64-bit
Byte Order:                      Little Endian
Address sizes:                   43 bits physical, 48 bits virtual
CPU(s):                          12
On-line CPU(s) list:             0-11
Thread(s) per core:              2
Core(s) per socket:              6
Socket(s):                       1
NUMA node(s):                    1
Vendor ID:                       AuthenticAMD
CPU family:                      23
Model:                           113
Model name:                      AMD Ryzen 5 3600 6-Core Processor
Stepping:                        0
Frequency boost:                 enabled
CPU MHz:                         2192.723
CPU max MHz:                     3600.0000
CPU min MHz:                     2200.0000
BogoMIPS:                        7186.80
Virtualization:                  AMD-V
L1d cache:                       192 KiB
L1i cache:                       192 KiB
L2 cache:                        3 MiB
L3 cache:                        32 MiB
NUMA node0 CPU(s):               0-11
Vulnerability Itlb multihit:     Not affected
Vulnerability L1tf:              Not affected
Vulnerability Mds:               Not affected
Vulnerability Meltdown:          Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl and seccomp
Vulnerability Spectre v1:        Mitigation; usercopy/swapgs barriers and __user pointer sanitization
Vulnerability Spectre v2:        Mitigation; Full AMD retpoline, IBPB conditional, STIBP always-on, RSB filling
Vulnerability Srbds:             Not affected
Vulnerability Tsx async abort:   Not affected
Flags:                           fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxs
                                 r sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl n
                                 onstop_tsc cpuid extd_apicid aperfmperf pni pclmulqdq monitor ssse3 fma cx16 sse4_1 sse
                                 4_2 movbe popcnt aes xsave avx f16c rdrand lahf_lm cmp_legacy svm extapic cr8_legacy ab
                                 m sse4a misalignsse 3dnowprefetch osvw ibs skinit wdt tce topoext perfctr_core perfctr_
                                 nb bpext perfctr_llc mwaitx cpb cat_l3 cdp_l3 hw_pstate sme ssbd mba sev ibpb stibp vmm
                                 call fsgsbase bmi1 avx2 smep bmi2 cqm rdt_a rdseed adx smap clflushopt clwb sha_ni xsav
                                 eopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local clzero irp
                                 erf xsaveerptr wbnoinvd arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyas
                                 id decodeassists pausefilter pfthreshold avic v_vmsave_vmload vgif umip rdpid overflow_
                                 recov succor smca
23) Message boards : Number crunching : New work Discussion (Message 62805)
Posted 31 Oct 2020 by lazlo_vii
Post:
With any luck I'll be able to crunch a few of them. I set WGC to "No New Tasks" on half of a Ryzen 7.
24) Message boards : Number crunching : New work Discussion (Message 62803)
Posted 31 Oct 2020 by lazlo_vii
Post:
I see over 1000 N216's in the queue. Are those bounced tasks or new?
25) Message boards : Number crunching : New work Discussion (Message 62789)
Posted 24 Oct 2020 by lazlo_vii
Post:
OpenIFS v2.19
x86_64-pc-linux-gnu
Peak working set size 11,797.17 MB
Peak swap size 13,131.91 MB
Peak disk usage 3,516.51 MB



If you get more would you please let us know how many you are running at once and post the output of "free -mw && cat /proc/meminfo"? I would like to know a little bit more what the kernel is actually doing. I may have to reconfigure things to add more swap space in a worst case scenario. Thanks!
26) Message boards : Number crunching : New work Discussion (Message 62778)
Posted 9 Oct 2020 by lazlo_vii
Post:
It looks like the 5 month WU's are running at 10% a day on the 3700X. It would be a little faster if I was only doing 3 of them but it's still a good rate.
27) Message boards : Number crunching : New work Discussion (Message 62771)
Posted 9 Oct 2020 by lazlo_vii
Post:
So I got excited and configured one of my machines to do the new N216 models. When it didn't get any work I thought "Oh well, gone already." Then I realized I had issue commands to the wrong system and it didn't get any work because it's queue was full. Oops. Once I had the right IP address for boinccmd I got 4 of the new 5 month models. I'll try to have them back within 14 days but I can't promise because I have never run 4 N216's on one Zen2 chiplet. I was able to do three 4 months WU's in 7-8 days though so it shouldn't be that bad.
28) Message boards : Number crunching : Updated BOINC Clients 7.16.11 - Windows 64-bit and Mac OS X (64-bit Intel) (Message 62737)
Posted 23 Sep 2020 by lazlo_vii
Post:
One thing about these ARP tasks is that they can be suspended/stopped/re-started endless times without having a fit and dying with a computation error.

Very endearing.


Another FYI, they recommend using "Keep no GPU in memory when Suspended" with ARP because they only write check points every 12.5% of the way through. On my Ryzen 7 systems that can up to an hour of work lost otherwise and even more on my other boxes.

But to your point about being able to restart them, yes, it is nice not having to wait five minutes before rebooting.
29) Message boards : Number crunching : Updated BOINC Clients 7.16.11 - Windows 64-bit and Mac OS X (64-bit Intel) (Message 62731)
Posted 19 Sep 2020 by lazlo_vii
Post:
Just an FYI, the ARP tasks hate SMT and eat L3 cache as much as CPDN N216 tasks. So do one per core or 2MB of L3 (which ever is lower) for best turn around times.
30) Questions and Answers : Unix/Linux : Building from source on 20.04 (Message 62711)
Posted 10 Sep 2020 by lazlo_vii
Post:
$ sudo systemctl disable boinc-client.service
Synchronizing state of boinc-client.service with SysV service script with /lib/systemd/systemd-sysv-install.
Executing: /lib/systemd/systemd-sysv-install disable boinc-client
Removed /etc/systemd/system/multi-user.target.wants/boinc-client.service.


That should solve any problems you have with BOINC auto starting in Ubuntu. After that you can use start or stop instead of disable to control it manually.
31) Questions and Answers : Unix/Linux : Building from source on 20.04 (Message 62691)
Posted 8 Sep 2020 by lazlo_vii
Post:
In Ubuntu 20.04 and beyond all packages are install from the Snap Store. This could be the root cause of your issue if something you have installed is packaged with the wrong version of the files you need to compile against. On the other hand, it's really simple to launch a LXD container running Ubuntu 18.04 so you could use the container as your build environment and see if resulting binary would run on 20.04.
32) Message boards : climateprediction.net Science : AFlame PROJECT (Message 62690)
Posted 8 Sep 2020 by lazlo_vii
Post:
Why not use the "Notices" tab in the BOINC client to let them know when the libraries are missing? Another route could be to compile any needed 32bit code as a bunch of static libs and bundle them with the executables needed to run the WU's.
33) Message boards : Number crunching : Welcome back/checking if everything is working? (Message 62672)
Posted 3 Sep 2020 by lazlo_vii
Post:
Out of interest whats the power consumption of this machine with all threads running. I still have a dual slot2 xeon machine... 700mhz. The lights used to dim when that was fired up. Gone are the days of cheap electricity. I think my I7 machine is about 26 quid a month to run. Thats English pounds.........


Not sure yet, will see if it goes up much though psu is more efficient than old machine. I haven't tried running with more than ten threads yet as that is all I got through testing. And with a very slow broad band will probably stick to that sort of number.



You will get better performance on CPDN tasks if you stick to one WU per core. Even then the N216's slow down my 3700X if I do more than 6 at once. The CPU itself should max out at about 65W. The RAM, fans, HDD, etc will add a bit as well. Most consumer HDD will pull 30W when powering on or waking up from sleep and idle at 5-10W. SSD's pull about 5W under full load. If you have an up to date BIOS for your Ryzen you can check the settings to see if you enable Eco-Mode for the CPU. That will drop it's high end clock speed about 5% but also drop max power usage to 45W.

EDIT: Another big power draw comes from wireless stuff. So avoid bluetooth and WiFi if you want to save money.
34) Message boards : Number crunching : Welcome back/checking if everything is working? (Message 62667)
Posted 2 Sep 2020 by lazlo_vii
Post:
Everything I can check at the moment seems to work normally. Having ditched my core2duo for a Ryzen7 I have lost a couple of tasks however.


Which CPU/MB combo did you go with?
35) Message boards : Number crunching : Welcome back/checking if everything is working? (Message 62664)
Posted 2 Sep 2020 by lazlo_vii
Post:
Welcome back to you as well!

I am glad to see you have things sorted out in Oxford.
36) Message boards : Number crunching : Big models (Message 62651)
Posted 9 Aug 2020 by lazlo_vii
Post:
Will these larger models also use more RAM? If so can we get a hint as to the numbers we should expect?
37) Message boards : Number crunching : Big models (Message 62646)
Posted 8 Aug 2020 by lazlo_vii
Post:
193 MB is huge if you are on a satellite or cell phone connection. This is not an issue for me in currently but in the past it would have been a show stopping hurdle. At this point I would support it because I have bandwidth to spare.

To be honest, CPU cycles are what I struggle with these days. I know that climate changes will kill 90% of all life on Earth if we don't stop it. I also know that human civilization as we know it will not survive COVID-19 unless we find multiple ways to treat it's symptoms and immunize against it. If we can't do that then humanity's governments will go insane and climate change will be on our Christmas Wish List. So I have every CPU I own (Two ARM CPU's, one Intel, three AMD) and one more (ARM) that I ordered today working on WCG OpenPandemic tasks. Don't get me wrong. I have a great faith in each and every one of you as a person. It's just that, as a species, I think we a dumber than a bag of hammers.

So until I can stop worrying about the knock on effects of a species so small I can't even it I will only do work for CPDN when there is a a solution to COVID or a lull in the work flow.
38) Message boards : Cafe CPDN : Where did everybody go (Message 62610)
Posted 14 Jul 2020 by lazlo_vii
Post:
I don't know about everyone else, but I have been crunching for WGC on their Open Pandemic COVID tasks. It seems like a good cause now that the Linux work here has dried up for the time being.


I’ve also been running OpenPandemic, alone with the odd WU’s from the African Rainfall Project.


Same here with Milky Way on GPU thrown in for good measure.



The OPN tasks will be ported to GPU in the near future. No ETA yet.
39) Message boards : Cafe CPDN : Where did everybody go (Message 62606)
Posted 13 Jul 2020 by lazlo_vii
Post:
I don't know about everyone else, but I have been crunching for WGC on their Open Pandemic COVID tasks. It seems like a good cause now that the Linux work here has dried up for the time being.
40) Message boards : Number crunching : New work Discussion (Message 62589)
Posted 19 Jun 2020 by lazlo_vii
Post:
What did you do? Upgrade to 20.04 without a clean install? Your computers show you have 20.04 on both computers and the tasks downloaded today.


Yes, did a standard upgrade via the command line. However this stopped the manager from working though not the client due to some wxwidgets conflict. This also causes make command to fall over when I roll my own manager on the laptop though not the desktop. I have found a work around with Jord's self extracting manager and client for the laptop to get it going again. Now I just have a longer wait till I can clean things up.


FYI, you can open a terminal and issue commands to start/stop/enable/disable the BIONC client before and after upgrades:

sudo systemctl start boinc-client
sudo systemctl stop boinc-client
sudo systemctl disable boinc-client
sudo systemctl enable boinc-client


You can also use the BOINC command line interface to manage the the client. See:

man boinccmd


Previous 20 · Next 20

©2024 climateprediction.net