Posts by MarkJ

21) Message boards : Number crunching : Linux CUDA 'Special' App finally available, featuring Low CPU use (Message 1877916)
Posted 12 Jul 2017 by Profile MarkJProject Donor
Post:
I found the version of wx3 I was looking for here, https://launchpad.net/~costamagnagianfranco/+archive/ubuntu/locutusofborg-ppa/+build/9539157
Tried it with a new download of bonic-master and ended up at the same place I stopped the last time.
I started a new thread if someone has anything to offer, https://boinc.berkeley.edu/dev/forum_thread.php?id=11719
Doesn't look as though there's much activity over there.

Not sure if you have tried dropping Gianfranco an email. He is one of the Debian maintainers.

I know when DA announced a test version on BOINC-alpha Gianfranco asked for things to get fixed up. I think the whole dev tree is stuffed at the moment.
22) Message boards : Number crunching : Setting up a Linux machine to crunch CUDA80 for Windows users (Message 1877389)
Posted 8 Jul 2017 by Profile MarkJProject Donor
Post:
For those that like Debian I managed to update my GPU crunchers to Stretch using pretty much the same instructions to install apart from changing Jessie to Stretch.

One gotcha was my firewall script wouldn't execute any more. I used to have it in rc.local. I followed the suggestions on the Debian wiki which run it as a service which still seems to work, however you also have to set execute permissions on a couple of files. I didn't have to do that under Jessie.

Also tried clean-installing Stretch on iGPU machines and can't get a desktop (mate) any more. I've raised a bug with Debian for that one.
23) Message boards : Number crunching : To checkpoint or not -- the wear and tear of SSD drives (Message 1877387)
Posted 8 Jul 2017 by Profile MarkJProject Donor
Post:
Q: Should the optimized app make checkpoints and report progress? How often? Based on time or number of internal iterations? Does it need a command line option for that?

I'm running the optimised app on GTX1060's (3GB) and they rip through a work unit in 5 mins or less. The machines all have HDD with a 2 minute checkpoint interval. I wouldn't bother with the complexity of checkpoints, but a percentage progress is always good to see.

Thanks for the great app. By the way if you're interested the Asteroids at home CUDA app needs optimising :-)
24) Message boards : Number crunching : Ok, what the heck is going on with limiting availability of video cards?? At least at Micro Center? (Message 1877385)
Posted 8 Jul 2017 by Profile MarkJProject Donor
Post:
I'm counting my blessings for getting the 1080's and 1060's when I did. Dang.

Me too. I got a second pair of 1060's last month before the price jumped. I can see why they had a limit of 2 per customer.

Now might be a good time to sell off the GTX970's that my 1060's replaced.
25) Message boards : Number crunching : Ryzen on Linux (Message 1876776)
Posted 4 Jul 2017 by Profile MarkJProject Donor
Post:
The question is should i upgrade to kernel 4.11 or 4.12 ?

Mike

Ryzen support is in 4.9.10 onwards, so 4.10 or 4.11 is more than enough. The main thing to keep updated is the bios to get microcode updates and other fixes.

I'm running Debian with the 4.9.30 kernel which seems to be the latest they have at the moment.
26) Message boards : Number crunching : Debian Project Warns: Turn off Hyperthreading with Skylake and Kaby Lake (Message 1875372)
Posted 27 Jun 2017 by Profile MarkJProject Donor
Post:
Thanks for the post. I have 8 of them. All got patched this morning.

Debian have a package called Intel-microcode available in both Stretch and jessie-backports. From what I understand the microcode is a partial fix and you also need a BIOS update.
27) Message boards : Number crunching : Validation pending.... (Message 1872518)
Posted 12 Jun 2017 by Profile MarkJProject Donor
Post:
The size of your local cache has no significant impact on the number of "pending" results. The number pending is totally dependent on the relative turnarounds of you and your wingmen.

Because you have less in the cache they turn around faster. Sure they take the same amount to time to run but they spend less time in a "ready to start" state. Think of those people who have 5 and 10 day caches, that's around 4.5 to 9.5 days before they even start them.

The drawback is of course when there is an outage (like every Tuesday) you could run out of work but that's what other projects are for.

I agree with you regarding the error and invalid rates, they are more important than how many are pending.
28) Message boards : Number crunching : Setting up a Linux machine to crunch CUDA80 for Windows users (Message 1872517)
Posted 12 Jun 2017 by Profile MarkJProject Donor
Post:
We don't tend to reinstall very often. For what Tbar, Petri and Jason do it would make sense but for the average to slightly advanced users we want to install it and have it run. There is the occasional app update to take care of but that is pretty much it. I take the KISS principle, if I can get most of the bits from a repo then that's one less thing for me to stuff up.

I've switched my entire farm from Windows to Linux apart from a daily driver laptop. A process that is repeatable and reliable is the way to go. If I got a new machine today I could have it up and running in say an hour with exactly the same setup as all the others apart from hardware differences.
29) Message boards : Number crunching : Validation pending.... (Message 1872418)
Posted 11 Jun 2017 by Profile MarkJProject Donor
Post:
I've got 2761 when I checked just now. That's out of 4595 so more than 50% are waiting. I run pretty small caches. A couple of my machines are running the cuda80 app which is quick.
30) Message boards : Number crunching : Setting up a Linux machine to crunch CUDA80 for Windows users (Message 1872415)
Posted 11 Jun 2017 by Profile MarkJProject Donor
Post:
Post the step by step guide here.

Its on the second page of this thread. It can also be found here

It doesn't cover the copying app_info and apps to the project folder but if you need that part I sure I can knock up something. I normally use a USB thumb drive and just drag and drop the files using the gui.
31) Message boards : Number crunching : Setting up a Linux machine to crunch CUDA80 for Windows users (Message 1872383)
Posted 11 Jun 2017 by Profile MarkJProject Donor
Post:
Why pick the most complicated way of doing it?

You could do a bog-standard Debian install and get everything from the repos apart from the CUDA app. You get a more up to date BOINC client (7.6.33) which has improved CUDA detection and a more up to date CUDA driver (375.66). The only complex bit would be doing the app_info and apps into the project folder.
32) Message boards : Number crunching : Restarting BOINC Manager (Message 1872209)
Posted 10 Jun 2017 by Profile MarkJProject Donor
Post:
I"m running Debian. I added the boincmgr to the desktop. In its properties I specified the password to the core client. That is the same password thats in /var/lib/boinc-client/gui_rpc_auth.cfg unless its blank.
33) Message boards : Number crunching : New Linux Trojans installs crypto currency mining software on Raspberry Pi (Message 1871948)
Posted 9 Jun 2017 by Profile MarkJProject Donor
Post:
You wouldn't think a Pi would produce much in the way of crypto currency. I have a bunch of Pi's and the first thing I do after imaging them is to change their password.

I did read one tutorial where the guy wanted to run it as a web server so he deleted the default user account. That sounds like a good idea now. For some reason though he didn't put a firewall on it.
34) Message boards : Number crunching : Linux CUDA 'Special' App finally available, featuring Low CPU use (Message 1871739)
Posted 8 Jun 2017 by Profile MarkJProject Donor
Post:
For those that went with Debian they've just put the 375.66 drivers into Jessie-backports. All I did to upgrade was:

1. Login as root
2. In a terminal type "service boinc-client stop"
3. Type "apt update" to get the latest list of packages
4. Type "apt upgrade -y" to update it. It pops up a warning about older drivers. Just press enter to continue
5. After its finished (it takes about 5 mins to compile) type "sync" to flush to disk
6. Type "reboot" to reboot the machine

My original instructions for setting up a Linux CUDA cruncher using Debian can be found here
35) Message boards : Number crunching : Do we have more data than volunteers? (Message 1871455)
Posted 6 Jun 2017 by Profile MarkJProject Donor
Post:
Speculation is pointless, I have a good reason to do this and SETI@Home serves the purpose. I may do more, as this is only a few racks in one row of several in my data center. It's only temporary, will end when it ends. Enjoy the fireworks.

I will say that I am not interested in competing against anyone. I just happen to like SETI@Home, even though I know the odds of it finding what it is looking for are infinitesimal.

Oh don't get me wrong I'm in this for science and am more than happy you've dedicated your machines to the search.

I'm just a little envious of your data center compared to my small farm. Unfortunately I'm constrained by available space, power and finances.
36) Message boards : Number crunching : Do we have more data than volunteers? (Message 1871223)
Posted 5 Jun 2017 by Profile MarkJProject Donor
Post:
I'll take up some of the slack. Just brought 7000+ cores online on SETI@Home. The reason why is hilarious, but unfortunately I can't share it here.

Something like management want to stress test the disaster recovery site? Sounds like more cores than the Atlas cluster that Einstein use.

They aren't the latest kit but a year or two old judging by the Xeon v2 and v3's. Hate to think what the electricity bill would be like.

I need a shed like Wiggo, my loft gets too hot even in winter 😩
37) Message boards : Number crunching : CES 2017 -- AMD RYZEN CPU (Message 1871092)
Posted 4 Jun 2017 by Profile MarkJProject Donor
Post:
I've got a couple of Ryzen 1700's running such as this one running Linux. My current quandary is if I should replace the 1700's with 1700x or not. Apart from the initial cost of buying another CPU there is an extra 30 watts for the 1700x for about a 10 to 12% gain in productivity.

I did some experiments with the Einstein gravity wave tuning run on them. Running 16 at a time average run time was 33,000 to 37,000 seconds, when I limit it to 8 via an app_config they came in at 20,000 seconds. The app was their AVX and it got the Lo work units.

My current experiment is to see if leaving a single thread free improves performance or not. I'm running a single GPU with the Seti CUDA80 app so it can make use of the available thread but certainly doesn't need its own as it sleeps a fair bit (according to top).
38) Message boards : Number crunching : AMD's Upcoming Ryzen R9 "Threadripper" 16-core, 32-thread CPU (Message 1868334)
Posted 20 May 2017 by Profile MarkJProject Donor
Post:
For a no-cost solution reducing the deadline would help reduce the database bloat that needs frequent compression? While it may have an effect on the slower rigs they could run a smaller cache so they can meet the deadline.

Currently over at Einstein they're doing a Gravity wave tuning run with a 5 day deadline. Now that's obviously not practical for normal use but it certainly clears the tasks a lot quicker. Maybe 10 to 21 days is a better number to use.
39) Message boards : Number crunching : Help with Intel GPU on Arch Linuc (Message 1866694)
Posted 11 May 2017 by Profile MarkJProject Donor
Post:
I am running Debian. When I set up mine I had to get beignet 1.3.0 (beignet-opencl-icd) from Jessie backports to support the HD Graphics 530. I also needed to get boinc-client-opencl from there. You don't need mesa.

Given you're running Arch I am not sure how up to date their repos are.

As mentioned above running the iGPU will slow down the rest of the CPU tasks so its usually not worth the effort.
40) Message boards : Number crunching : You have to love early Xmas and B'day presents (even if you have to pay for them yourself). (Message 1864716)
Posted 30 Apr 2017 by Profile MarkJProject Donor
Post:
I've converted the farm to Linux. Up until Friday the two GTX1060's were running the Petri special and then I decommissioned the machines. I'm selling them off to get a couple of Ryzen's and the two GTX1060's will go in them.

In the mean time I have a couple of i3's running a GTX970 each with Petri special. I've ordered another pair of 1060's which will replace the 970's in order to cut down the electricity use a bit. Its nowhere near Christmas but that's my present to myself :-)


Previous 20 · Next 20


 
©2017 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.