1. To donate NVIDIA hardware.
It's excessively hard to develop w/o ability to run what you developed.
PM me to arrange donation.

Have no spare hardware but have money to share?
2. To donate money for NVIDIA hardware aquisition.
PM Mike or me to arrange donation.

Have no spare hardware or money but have testing skills and familiar with application tuning?
3. Run special test builds to help to adopt existing software to wide range of hardware devices of NVIDIA platform and help others to setup such builds for test.
Builds to be published in this thread or dedicated threads on Lunatics site (currently out of reach).

Have nothing but will to help?
4. Run special test builds on your own host and ask for help to install them on own host. There are enough peoples around who could help with initial setup.
Post links to build's processing results to help in testing.

All these ways of help will lead to improvement of existing SETI software for NVIDIA in all ways, performance and usability both including. And all such help will be accepted with gratitude.

---

SETI apps news
We're not gonna fight them. We're gonna transcend them.

I'm always interested in #3 and #4 though have only two Windows hosts; please ask me anytime (I think you have my e-mail.) I also have some spare cards (ie 2xGTX480, 2xGTX460, GTX570, 9800GT) but I think the postage to Russia would be more than they are worth!

---

I'm always interested in #3 and #4 though have only two Windows hosts; please ask me anytime (I think you have my e-mail.) I also have some spare cards (ie 2xGTX480, 2xGTX460, GTX570, 9800GT) but I think the postage to Russia would be more than they are worth!

Wouldn`t cost anything.
I can arrange that.

---

With each crime and every kindness we birth our future.

but I think the postage to Russia would be more than they are worth!

Yes, that's the real issue that already prohibited one of donations. Usual send via post too costly to be useful indeed.

Regarding #3 - stay tuned I'll post builds with additional counters and output to explore what FREMI+ midrange GPUs are.

---

SETI apps news
We're not gonna fight them. We're gonna transcend them.

I'm always interested in #3 and #4 though have only two Windows hosts; please ask me anytime (I think you have my e-mail.) I also have some spare cards (ie 2xGTX480, 2xGTX460, GTX570, 9800GT) but I think the postage to Russia would be more than they are worth!

Wouldn`t cost anything.
I can arrange that.

Great! Mike is our lifesaver as usual in such questions.

---

SETI apps news
We're not gonna fight them. We're gonna transcend them.

but I think the postage to Russia would be more than they are worth!

Yes, that's the real issue that already prohibited one of donations. Usual send via post too costly to be useful indeed.

Regarding #3 - stay tuned I'll post builds with additional counters and output to explore what FREMI+ midrange GPUs are.

. . I am running a GT730 in my other rig and that would qualify for low end Nvidia GPU testing. I am always willing to try new things. The machine is a Core2 Duo E7600 so kind of low end as well.

[Edit] Also, with the success of SoG on Guppis does that suggest that Arecibo VLAR WUs might be crunchable on Nvidia GPUs now?

.

I have a GT 720 M and willing to help.

Here: https://cloud.mail.ru/public/8h9C/QZkh4ejzP is the first of builds with increased output for better understanding NV GPU load at PulseFind.

Please run it mostly with GUPPI tasks.
And please leave app at default config (only add options listed below) or at that config (in case of driver restarts) where app can process but GUI lags are big enough.

Additional options to enable are:
-use_sleep -v 6

With these options (and even w/o them) app will generate quite a big stderr.txt file.
Please don't post its content here. Just zip it, upload somewhere and post the link. I need stderr for GBT VLAR task (better to copy stderr.txt from slot directory to safe location before task finish (~90% done) cause after completion BOINC will truncate stderr and make it unusable).


Most interesting for me are NV GPUs that experience GUI lags or driver restarts at default config.

---

SETI apps news
We're not gonna fight them. We're gonna transcend them.

Here: https://cloud.mail.ru/public/8h9C/QZkh4ejzP is the first of builds with increased output for better understanding NV GPU load at PulseFind.

Please run it mostly with GUPPI tasks.
And please leave app at default config (only add options listed below) or at that config (in case of driver restarts) where app can process but GUI lags are big enough.

Additional options to enable are:
-use_sleep -v 6

With these options (and even w/o them) app will generate quite a big stderr.txt file.
Please don't post its content here. Just zip it, upload somewhere and post the link. I need stderr for GBT VLAR task (better to copy stderr.txt from slot directory to safe location before task finish (~90% done) cause after completion BOINC will truncate stderr and make it unusable).


Most interesting for me are NV GPUs that experience GUI lags or driver restarts at default config.

. . This is my first attempt to fill this role. I am about to install Lunatics 0.45 Beta(3). Will that conflict with doing this for you? And how do I actually run this?

I have done some testing with the older r3430 low_perfGPU version.
Increasing the -period_iterations_ number initially only prevented driver crashes.
Beyond about 4, that was solved, but the tasks would eventually "hang".
Further increasing the pin did not help.
However, increasing the -use_sleep_ex to 3, while reducing the pin back to 4 seems to have fixed the problem.
This might be old news, but i thought, i mention it anyway.

Was it original build or one with hang fix I posted later?
For original -use_sleep_ex 3 will be not increase but decrease in sleep. Default for low-path was 5.
New default is 1 (for build with fix and up until new change will (if) be needed).

---

SETI apps news
We're not gonna fight them. We're gonna transcend them.

Was it original build or one with hang fix I posted later?

Yes.

I have done some testing with the older r3430 low_perfGPU version.
Increasing the -period_iterations_ number initially only prevented driver crashes.
Beyond about 4, that was solved, but the tasks would eventually "hang".
Further increasing the pin did not help.
However, increasing the -use_sleep_ex to 3, while reducing the pin back to 4 seems to have fixed the problem.
This might be old news, but i thought, i mention it anyway.

. . All good. Never hurts to review relevant details. But I was concerned about how to get it into service, I have only ever run installer apps before, I have never manually installed an app.

Was it original build or one with hang fix I posted later?
For original -use_sleep_ex 3 will be not increase but decrease in sleep. Default for low-path was 5.
New default is 1 (for build with fix and up until new change will (if) be needed).

. . I have downloaded your files but Win 10 does not see it as an archive or compressed file.

. . I have downloaded your files but Win 10 does not see it as an archive or compressed file.

Which files from where?

. . I have downloaded your files but Win 10 does not see it as an archive or compressed file.

Which files from where?

. . The download Raistmer posted the link to a few messages ago ...

. . I have downloaded your files but Win 10 does not see it as an archive or compressed file.

Which files from where?

. . The download Raistmer posted the link to a few messages ago ...

Raistmer usually uses 7z compression these days (he used to use rar...). Unlike zip, Microsoft hasn't appropriated 7z from the original developer.

Get the tools from http://www.7-zip.org/ (free, open source)

. . I have downloaded your files but Win 10 does not see it as an archive or compressed file.

Which files from where?

. . The download Raistmer posted the link to a few messages ago ...

Raistmer usually uses 7z compression these days (he used to use rar...). Unlike zip, Microsoft hasn't appropriated 7z from the original developer.

Get the tools from http://www.7-zip.org/ (free, open source)

. . Thanks for that

Most interesting for me are NV GPUs that experience GUI lags or driver restarts at default config.

. . This is my first attempt to fill this role. I am about to install Lunatics 0.45 Beta(3). Will that conflict with doing this for you? And how do I actually run this?

. . Not a great success.

. . The first run through was taking a very long time before I had to go to bed to I set the command line to try and speed up the next WUs.

. . The first WU ending up taking 7 hours and 18 mins. :(

. . The new command line was : -use_sleep_ex 1 -sbs 384 -v

. . I should not have tried the larger buffer size because when I arose all subsequent Guppi WUs had errored out.

. . Sorry but I was hoping that it would speed things up enough to get more info.

. . I have now returned Lunatics 0.45 Beta(3) to running CUDA50.

. . The new command line was : -use_sleep_ex 1 -sbs 384 -v

Option to try is -v 6 .
6 is important here.

. . Sorry but I was hoping that it would speed things up enough to get more info.

Nope. Debug/verbose build will speedup things ultimately but because of info and new optimizations I can gather from its output, not because it faster per se. Increased verbosity comes with performance degradation - rule of thumb (for human operation too btw ;) ).

BTW, most likely failures in task come from not correct -v option use.
Will make it more foolproof as result.

---

SETI apps news
We're not gonna fight them. We're gonna transcend them.