Installer v0.41 release notes

Started by arkayn, June 03, 2013, 06:17:33 AM

Previous topic - Next topic
Lunatics are pleased to announce the availability of the v0.41 installer.

Please take care you use the right installer (32/64 bit) for your system.

Should you have any questions or problems - ask! Post in this thread or open a new one.

New in this installer:
Update of all MB applications to support MB v7
Update of all AP apps to latest revisions.

How to use the installer

For the CPU applications, you need to know what your CPU supports. Either check the Boinc startup messages or check with CPU-Z downloadable from http://www.cpuid.com
Tick or untick CPU AP and MB applications as desired. 
Tick ATI (OpenCL) AP or MB if/as desired.
Tick NVidia (OpenCL) AP or CUDA MB if/as desired.
More details further down.

NB We use default values throughout the app_info.xml. If you have previously installed optimised applications and manually edited app_info.xml (e.g. <count> for multithreading GPUs, <flops>, <cmdline> for ATI) you WILL lose those edits on running the installer and have to redo them. You may wish to make a note of them. You will also lose all other manual edits to app_info.xml such as beta applications.

Known issues - general
The installer is supposed to shut down the BOINC client whilst leaving the Manager running - noticable from the red 'disconnected' dot in the tray. After finishing the BOINC client should restart and the red dot vanish when the Manager reconnects. This may take a few minutes. On some systems however this fails to work. Please ensure BOINC has restarted.
Advanced users only: If you know you are going to edit app_info.xml, it may be easier to shutdown and restart BOINC manually, so you can get the changes into app_info.xml before the restart.


CPU MB apps - AKv8c

If you have any doubt at all about what instruction sets your CPU supports, download CPU-Z from http://www.cpuid.com/softwares/cpu-z.html. We have now incorporated a 3rd party CPU feature detection tool into the installer. Unsupported applications will be greyed out.


CPU AP app - r1797

Updated from r555/r557


NVidia GPU (CUDA) app - x41zc

For details please see the separate ReadMe_x41zc.txt.

Rule of thumb selection guide:
Cuda 2.3: PreFermi - NOT to be used on Fermi/Kepler
Cuda 3.2: mixed systems
Cuda 4.2: Fermi
Cuda 5.0: Kepler
If you have run stock and have established APR you can also pick the fastest app from there.
For compatible cards see http://developer.nvidia.com/cuda-gpus
Should run on all NVidia CUDA cards with at least 512MiB video RAM.
Actual free memory required is in the region of 200-235MiB, depending on driver, OS flavour and CUDA variant. The code will scale up if more memory is available.
Running on a 256MiB card is just possible, but you will have to free up as much VRAM as possible, e.g. by disabling Aero.
Used video RAM can be checked with GPU-Z downloadable from http://www.techpowerup.com/downloads/2244/techpowerup-gpu-z-v0-7-1/
Watch closely on GPU-Z and in BOINC manager if the application finds enough memory.
Symptoms of insufficient memory are
- in BOINC manager: tasks start up and run a few seconds then go to 'waiting  to run' and the next task tries to start
- on GPU-Z: the sensor for memory shows used VRAM ramping up and almost immediately dropping again.
Depending on BOINC version your tasks will either error out with 'too many exits' or get stuck in an infinite loop.
If you don't have enough memory, set BOINC to NNT, abort tasks and uninstall the CUDA application by rerunning the installer with CUDA MB UNticked.
If in doubt ask for help on NC:
http://setiathome.berkeley.edu/forum_forum.php?id=10
To multithread on Fermi/Kepler cards, find all instances of <count>1</count> in app_info.xml and decrease to 0.5 or 0.3 to run two or three tasks in parallel.
Again, if  you are uncertain of how to do this correctly, ask on the above mentioned forum. 'Your mileage may vary' - it depends on your specific system which count will give the highest throughput.
For full release notes and history please see the readme provided in the docs folder [after install].


NVidia (OpenCL) AP app - r1843

Please see the separate ReadMe_AstroPulse_OpenCL_NV.txt for details.
Reserving one CPU core per GPU card will improve performance.


ATI (OpenCL) MB app - r1843

Please see separate ReadMe_MultiBeam_OpenCL_ATI.txt for details, especially on driver requirements.
Reserving one CPU core per GPU card will improve performance.
We offer the MB application in two different variants, depending on card:
plain - for most HD4xxx owners and those with HD5xxx, HD6xxx cards and  HD7xxx cards who suffer from driver incompatibilities.
HD5 - for most if not all owners of HD5xxx GPUs and HD6xxx GPUs. For most HD7xxx cards ('Tahiti' core cards may prefer the plain app).

ATI (OpenCL) AP app - r1843
ATI (Brook) AP app - r1854

Please see separate ReadMe_AstroPulse_OpenCL_ATI.txt or ReadMe_AstroPulse_Brook.txt for details, especially on driver requirements.
Reserving one CPU core per GPU card will improve performance.

For cards without OpenCL support (from HD2xxxx on) use the 'hybrid' r1854 Astropulse application. NB this is a combined CPU/GPU application and will only use the GPU for some of the calculations, while doing the rest on a CPU core.
Minimum driver version Catalyst 9.x.
The Hybrid AP application uses about 80% of a CPU core and does only about 20% of the calculations on the GPU. Parameters have been choosen to reflect this usage. However present BOINC can not reserve '0.8' of a CPU core and therefore reserves none. This means your CPUs will be overcommitted and runtimes and DCF (if applicable) will suffer. You do increase the overall output though.



This time the thanks go to Joe (Josef W. Segur) for the optimised CPU apps, to Raistmer (with additional input from Urs Echternacht) for OpenCL apps and to Jason (jason_gee) for CUDA apps.
Thanks guys. There would be no need for an optimised app installer if there weren't optimised apps in the first place!

We would also like to take this opportunity to thank our active alpha testers (in no particular order and apologies to anybody we've forgotten):
Claggy, arkayn and Mike.

Additional thanks go to BilBg for thoroughly beta-testing the installer and chasing several bugs out of it.


The Lunatics Installer team AD MMXIII

Hello,

I have a thing or two to report on the installer v.41 that could be improved. The app_info.xml could be tidied up a bit.

1) Application versions:
Frist off I understand that v7.00 and v6.10 application versions are used. v6.09 and v6.08 could be removed alltogether.
Keep:
<version_num>700</version_num>
<version_num>610</version_num>

remove:
<version_num>609</version_num>
<version_num>608</version_num>

2) 64 vs 32 bit versions:
Currently both 64bit and 32bit versions have entry. If the user installs for a specific environment, isn't it enough that the 64bit installer ads only x64 entries and the 32bit x86 entries

<platform>windows_x86_64</platform>
vs.
<platform>windows_intelx86</platform>

3) Plan class entries:
The user opts the installer to install one plan class. However, they all end up in the xml file with reference to the selected optimized executable. Wouldn't it be enough to have entries with only those plan_classes that has been selected to be installed?

<plan_class>cuda22</plan_class>
<plan_class>cuda23</plan_class>
<plan_class>cuda32</plan_class>
<plan_class>cuda42</plan_class>
<plan_class>cuda_fermi</plan_class>
<plan_class>cuda50</plan_class>


June 08, 2013, 12:11:24 AM #2 Last Edit: June 08, 2013, 12:13:04 AM by Claggy
You might not have existing work from the 6.08 and 6.09 Cuda apps, But other people may, are you wanting the installer to loose existing work from everyone else's hosts?,
If they had existing work from the Cuda22 plan class, then that plan class is needed, if they also had work from the Cuda23, Cuda32, Cuda42, Cuda50 plan classes then they are needed too,
do you know if everyone else's work is branded as <platform>windows_intelx86</platform> or <platform>windows_x86_64</platform> ?

Claggy

Claggy is exactly right. All those plan_class entries are there to catch and transition as many cached workunits as possible at the moment of installation, no matter what the user's hardware might have caused to be sent by the server.

Speaking personally, I currently have something like 300 sah_v6 tasks pending - many people will have more. Some wingmates will be crunching, some will throw errors, some will walk away and leave the WUs to time out. Once all of us report clean pending lists (which I suspect will not happen for several months), we can lose the v6 entries, but not until then.

June 08, 2013, 12:53:51 AM #4 Last Edit: June 08, 2013, 01:01:29 AM by Caprio
Quote from: Claggy on June 08, 2013, 12:11:24 AM
You might not have existing work from the 6.08 and 6.09 Cuda apps, But other people may, are you wanting the installer to loose existing work from everyone else's hosts?,
If they had existing work from the Cuda22 plan class, then that plan class is needed, if they also had work from the Cuda23, Cuda32, Cuda42, Cuda50 plan classes then they are needed too,
do you know if everyone else's work is branded as <platform>windows_intelx86</platform> or <platform>windows_x86_64</platform> ?

Claggy

I am suprised if there is still 6.08 or 6.09 work out there. I think that it is over a year since I saw any older work thatn 6.09 for Cuda. With respect to backwards compatibility, wouldn't it be covered if installer was modified to allow multiple plan_class selection.

I would assume that platform selection goes with the installation version of Boinc. A 2-choice radio-button on the installer could let the user choose.

Quote from: Richard Haselgrove on June 08, 2013, 12:21:42 AM
Speaking personally, I currently have something like 300 sah_v6 tasks pending - many people will have more. Some wingmates will be crunching, some will throw errors, some will walk away and leave the WUs to time out. Once all of us report clean pending lists (which I suspect will not happen for several months), we can lose the v6 entries, but not until then.

I bet those v6 entries are 6.10 right? I fully agree the latest 6.10 should be supported. The last WU of 6.08 and 6.09 are to my knowledge distributed over a year ago. WU usually times out in a matter of a month or so.

June 08, 2013, 01:23:13 AM #6 Last Edit: June 08, 2013, 01:33:19 AM by Claggy
Quote from: Caprio on June 08, 2013, 12:56:34 AM
Quote from: Richard Haselgrove on June 08, 2013, 12:21:42 AM
Speaking personally, I currently have something like 300 sah_v6 tasks pending - many people will have more. Some wingmates will be crunching, some will throw errors, some will walk away and leave the WUs to time out. Once all of us report clean pending lists (which I suspect will not happen for several months), we can lose the v6 entries, but not until then.

I bet those v6 entries are 6.10 right? I fully agree the latest 6.10 should be supported. The last WU of 6.08 and 6.09 are to my knowledge distributed over a year ago. WU usually times out in a matter of a month or so.
v6 MB Wu's can be done by any of the Stock apps, (there is no such thing as a 6.08 or 6.09 Wu) on Windows that is 6.03 for the CPU, 6.10 for Fermi GPUs,
6.09 for legacy Nvidia GPUs running at least 190.38 drivers and at least 384Mb Free, and 6.08 for legacy Nvidia GPUs running at least 177.xx drivers and ~255Mb,

Setiathome Applications

Quote
                                  SETI@home Enhanced

Platform                         Version                  Installation time

Linux/x86                        6.03                     21 Aug 2008, 22:10:45 UTC
Windows/x86                      6.03                     21 Aug 2008, 22:10:45 UTC
Windows/x86                      6.08 (cuda)              21 Jan 2009, 1:43:04 UTC
Windows/x86                      6.09 (cuda23)             9 Dec 2009, 17:26:45 UTC
Windows/x86                      6.10 (cuda_fermi)         8 Jun 2010, 22:50:03 UTC
Mac OS/X 10.3+                   6.05                     14 Sep 2011, 18:09:58 UTC
SPARC/Solaris                    5.12                      1 May 2006, 22:43:19 UTC
Linux/x86_64                     5.12                      1 May 2006, 22:43:19 UTC
SPARC/Solaris                    5.12                      1 May 2006, 22:25:36 UTC
Mac OS X/Intel                   6.05                     14 Sep 2011, 18:09:58 UTC
Linux/x86_64                     5.28                      8 Oct 2007, 21:55:08 UTC

Claggy

Powered by EzPortal