At last a WIN10 option that is useful, I think?

[Dingo]

There is a message on the News about this on GPUGRID>  http://www.gpugrid.net/forum_thread.php?id=4716

I have been trying to use Cygwin to backup the code on my Linux Server to my WIN 10 machine at home.  This just might be the way ?

I will also give GPUGRID a try on WSL on Windows 10 as well.

I took a screenshot but had to decrease the size to post here so not sure if you can see the details,

[attachment deleted by admin]

[Dingo]

Well just finished the code to backup my Linux server onto my Win 10 PC.  It took about five minutes to code using rsync and it is running now.  I have been screwing around with CYGWIN for a while.


Another good thing is that linux can also see and read / write to the Windows files  

Code Select Expand


ls /mnt

c d

Directory c/ points to your Windows C: drive. This command will list the Windows program files directory.

ls '/mnt/c/Program Files/'


cd '/mnt/c/Users/Public/Downloads/'
echo 'Hello' > hi.txt
/mnt/c/Windows/System32/notepad.exe hi.txt



Makes it easy to move files etc to from within the WIN 10 Linux and other Linux servers.

[Dingo]

I have a lot of time on my hands and this mornong at about 01:00 just before I was going to bed I decided to load BOINC onto Ubuntu that is part of Windows 10.  I already had it enabled as I use it to rsync from my linux server to a backup on my Win 10 machine.  If you search in google for "enable ubunto on win 10" you will see how to do it.

It is now 01:32 and I am loading the Nvidia drivers for linus to see if I can use my GTX 1080 ti on this linux and run Numberfields as well as preparing for my 6 gpu rig when I get it all together.  It is only 24% through the install of the drivers and up to the 128th package because the Ubunto on Win 10 is very basic so loading BOINC and the drivers also needs to download all the associated packages that make them work.

Before I started installing the drivers I had BOINC up and running and could use my Win 10 Boinc Manager to manage the installation by selecting the IP address from BOINC Manager.  I installed Numberfields as test and it worked.  This is the Machine:  https://numberfields.asu.edu/NumberFields/show_host_detail.php?hostid=1516927



Loading BOINC on Ubuntu so that it can be managed from another machine via BOINC Manager

apt install boinc

cd /etc/init.d
rm boinc-client
nano boinc

Insert the following and save:

#!/bin/bash
# BOINC start/stop
#
if [ $1 == start ];then
cd /usr/bin
./boinc --daemon --allow_remote_gui_rpc
fi
if [ $1 == stop ];then
  pkill boinc
fi

============================================
chmod 755 /etc/init.d/boinc
update-rc.d boinc defaults
nano /etc/boinc-client/gui_rpc_auth.cfg   ---->   insert the password
nano /usr/bin/gui_rpc_auth.cfg            ---->   insert the password

********************************************

If this is on a normal Ubuntu or Dabian install then you need to setup the firewall.
WIN 10 Ubuntu uses the Windows 10 Firewall so no need to do the following:

^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
apt -y install iptables-persistent
nano iptables.sh
**Add the following to the file**
iptables -P INPUT DROP
iptables -P FORWARD DROP
iptables -P OUTPUT ACCEPT
iptables -A INPUT -i lo -j ACCEPT
iptables -A INPUT -m state --state RELATED,ESTABLISHED -j ACCEPT
iptables -A INPUT -p tcp --dport 22 -j ACCEPT
iptables -A INPUT -p tcp --dport 31416 -j ACCEPT
iptables -A INPUT -p tcp --dport 80 -j ACCEPT
iptables -A INPUT -p tcp --dport 5901 -j ACCEPT
iptables -A INPUT -p tcp --dport 18080 -j ACCEPT
iptables -A INPUT -p icmp -j ACCEPT
iptables-save > /etc/iptables/rules.v4
iptables-save > /etc/iptables/rules.v6
***
Save the file then run it
sh iptables.sh  

**** Check the file to see if it was created OK
cat /etc/iptables/rules.v4
cat /etc/iptables/rules.v6
rm iptables.sh

^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
/etc/init.d/boinc stop
/etc/init.d/boinc stop  ---->  should already be stopped but make sure
/etc/init.d/boinc start   --------> Start boinc


Download and Install Nvidia Drivers:

I used the tazzduke's instructions below and added a bit:

Multi GPU install in Linux.

This is what worked for me in getting Linux up and running.

1. Install Linux
2. Check for updates and then update.
3. Reboot (doesn't hurt to do this)
4. Update to latest kernel
5. Reboot (doesn't hurt to do this)
6. Add NVIDIA PPA to the repository
   sudo add-apt-repository ppa:graphics-drivers/ppa
   sudo apt-get update
7. Look at this page to see what version to install  
8. It will check and give you options on which NVIDIA Driver to install.  https://launchpad.net/~graphics-drivers/+archive/ubuntu/ppa
  currently:  sudo apt-get install nvidia-418
9. reboot
10    Check Installation status
   sudo lsmod | grep nvidia  or try  sudo nvidia-smi
   If there is no output, then your installation has probably failed. It is also possible that the driver is not available in your system's driver database. You can run the following command to check if your system is running on the open source driver nouveau. If the output is negative for nouveau, then all is well with your installation.
     sudo lsmod | grep nouveau

11. Prevent automatic update:
      sudo apt-mark hold nvidia-418

12. Are you running into issues with the new drivers, you can easily remove it.
      a. Remove the graphics-drivers PPA --->   sudo add-apt-repository -r ppa:graphics-drivers/ppa
       b. Enter the following command to completely remove the driver
 
sudo apt-get purge nvidia*

13. Reboot

14. Add Boinc PPA to the repository
  ## shouldn't need this   sudo add-apt-repository ppa:costamagnagianfranco/boinc
   ##                      sudo apt-get update
15. Install Boinc sudo apt install boinc  ****  see above so that you can use the GUI BOINC Manager on another machine.
16. Add GPU's one at a time, reboot, repeat till all cards are recognised.
17 sudo apt install mesa-utils
18  glxinfo  # this will show what nvidia cards are installed
19  to check individual GPU's use:
     cat /proc/driver/nvidia/gpus/0/information
     cat /proc/driver/nvidia/gpus/1/information
     cat /proc/driver/nvidia/gpus/2/information
    etc



I am currently on 8 above loading the drivers which are downloading very slow at about 32 kb/s so I am going to bed as it is 02:06 and I need my beauty sleep.  I will leave this running and check back tomorrow when I get up which will be late.

EDIT:  OK I am still up and doing this but will stop soon, as it's 02:31

[chooka03]

LOL
You sound like me Dingo! Lose sleep or can't sleep when there's BOINC things to be done. (or in my case usually fixing hardware/software issues)

[Dingo]

Well got this on BOINC Manager today so I am not going to go any further getting my 1080Ti recognised by Win10 Ubuntu:  I cannot get the driver recognised by the system and I think it is because it is not really Ubuntu as it runs the backend dtuff off windows I think ?

I did see that it will probably work on a stand alone Ubuntu system like my GPU Rack.  Hopefully Tuesday or Wednesday it will be up and running.

NumberFields@home: GPU app status update
So there have been some new developments over the last week. It's both good and bad.

First of all, some history. The reason I waited so long to develop a GPU app is because the calculation was heavily dependent on multi-precision libraries (gmp) and number theoretic libraries (pari/gp). Both of these use dynamically allocated memory which is a big no-no in GPUs. I found a multi-precision library online that I could use by hard coding the precision to the maximum required (about 750 bits), thereby removing the dependence on memory allocations. The next piece of the puzzle was to code up a polynomial discriminant function. After doing this, I could finally compile a kernel for the GPU. That is the history for the current GPU app. It is about 20 to 30 times faster than the current cpu version (depends on WU and cpu/gpu speeds).

But then I got thinking... my GPU polynomial discriminant algorithm is different from the one in the PARI library (theirs works for any degree and mine is specialized to degree 10). So to do a true apples-to-apples comparison, I replaced the PARI algorithm with mine in the cpu version of the code. I was shocked by what I found... the cpu version was now about 10x faster than it used to be. I never thought I was capable of writing an algorithm that would be 10x faster than a well established library function. WTF? Now I'm kicking myself in the butt for not having done this sooner!

This brings mixed emotions. On one side, it is great that I now have a cpu version that is 10x faster. But it also means that my GPU code is total crap. With all the horsepower in a present day GPU I would expect it to be at least 10x faster than the equivalent cpu version. Compared with the new cpu version, the gpu is only 2 to 3 times faster. That is unacceptable.

So the new plan is as follows:
1. Deploy new cpu executables. Since it's 10x faster, I will need to drop the credit by a factor of 10. (Credits/hour will remain the same for the cpu but will obviously drop for the GPU)
2. Develop new and improved GPU kernels.

I don't blame the GPU users for jumping ship at this point. Frankly, the inefficiency of the current GPU app just makes it not worth it (for them or the project).

For what it's worth, I did have openCL versions built. Nvidia version works perfectly. The AMD version is buggy for some reason, as is the windows version. Since I will be changing the kernels anyways, there is no point in debugging them yet.