It seems that the GTX 970 is about as fast as the GTX 670 for crunching multibeam work units. The 970 will of course use a little bit less power.

I am amazed the GTX 970 would not increase in performance over the previous generations.

---

How many work units per card are you running?

Yes, We're hitting a number of limits in the application, rather than the GPU. For development purposes, Finding [& understanding] all those limits on my 980 is taking a significant amount of time, as some of those include how chatty the application is, scaling of the pretty small datasets, and some underlying system considerations.

For the time being, the best way is to up the process priority & run multiple instances (usually 2-3 per GPU), and the combined throughput should be around 2x a 670. (the second task on a maxwell GPU seems to scale very well, even with an old Core2Duo driving it)

---

"Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions.

can any post a smile "how to" to have your gpu work on more than one work unit at a time.

thanks

---

Simple, create a text file called app_config.xml with Notepad and put it into the Seti directory. The contents would have something like this:



<app_config>
<app>
<name>setiathome_v7</name>
<gpu_versions>
<gpu_usage>.33</gpu_usage>
<cpu_usage>.10</cpu_usage>
</gpu_versions>
</app>

<app>
<name>astropulse_v6</name>
<gpu_versions>
<gpu_usage>.5</gpu_usage>
<cpu_usage>.50</cpu_usage>
</gpu_versions>
</app>

<app>
<name>astropulse_v7</name>
<gpu_versions>
<gpu_usage>.33</gpu_usage>
<cpu_usage>0.50</cpu_usage>
</gpu_versions>
</app>
</app_config>

But in your case, I would reduce the GPU tasks to two per card with a 0.5 entry for gpu_usage.

Cheers, Keith
P.S. They is a whole complete wiki for config.xml at the BOINC site for the syntax of the files.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

Jason, do you want to divulge the predicted timeline for new apps that harness the power of the Kepler and Maxwell hardware? How far out are they.... 6 months, one year??

Cheers, Keith

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

Jason, do you want to divulge the predicted timeline for new apps that harness the power of the Kepler and Maxwell hardware? How far out are they.... 6 months, one year??

Cheers, Keith

Hard to predict with current family & personal issues, but with Cuda 7.0 being released a couple of days ago, that's one less technical roadblock. (testers were experiencing unexplained reliability issues with Cuda 6.0 and 6.5, and no appreciable performance gains with the current application (x41zc) architecture so they were out)

Also I've been migrating build system to Gradle ( see http://en.wikipedia.org/wiki/Gradle ), it complicates the timeline a bit. That's an extra development burden up front expected to ease cross platform release in the long run (so worthwhile)

Aside from the infrastructure changes, the reengineering parts involved place alpha test x42 builds within the 3 month timeframe. That's after the already confirmed architectural changes needed to reduce the chattiness, up the load with fewer instances, and scale better from the smallest Cuda device through to the TiTan-X.

So short version ~3 months to x42 alpha, which is more or less a completely reengineered design based on everything we found. Aside from improved Cuda scaling, it's expected to have support for OpenCL devices and in a later revision AP (though those come later, and not based on current code/techniques)

[Edit:] Note that Windows 10 release, and adapting to accomodate WDDM2.0/dirextX12 techniques & best practices may or may not extend the timeline. That'll probably be a bit cleaer after I get to play with the tech preview this weekend ( USB is made, machine & new SSD are waiting)

---

"Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions.

P.S. They is a whole complete wiki for config.xml at the BOINC site for the syntax of the files.

Client configuration
Application configuration

How many work units per card are you running?

I am running 3 per card.

---

So short version ~3 months to x42 alpha, which is more or less a completely reengineered design based on everything we found. Aside from improved Cuda scaling, it's expected to have support for OpenCL devices and in a later revision AP (though those come later, and not based on current code/techniques)

[Edit:] Note that Windows 10 release, and adapting to accomodate WDDM2.0/dirextX12 techniques & best practices may or may not extend the timeline. That'll probably be a bit cleaer after I get to play with the tech preview this weekend ( USB is made, machine & new SSD are waiting)

Thanks for the update Jason. Good luck with the Windows10 Tech Preview. So, likely a X42 general release sometime around the end of the year barring hitting any roadblocks with the new Windows graphics engine. Looking forward to it.

Cheers, Keith

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

first thing I would raise the priority in your mbcuda.cfg

find that in your setiathome folder and open with notepad

go down to the last few line and change them to the following

remove the ;; before the last 3 lines

processpriority = normal
pfblockspersm = 16
pfperiodsperlaunch = 400


(where it says processpriority you can either make it normal, abovenormal or high) but depending on what you use the computer for beside crunching these will affect usage. If it's only a cruncher then you can try abovenormal or high

I also see that there is a 670 in the same machine as the 970.

One of the other will have to say if the 670 might be slowing down the 970. I know when I had a 750 in with some 780s it tended to slow the 780s somewhat. Others might know more than I on that.

Happy crunching...

Zalster

thanks... ordered my GTX 980 today to replace my SLIed 570s. will go back to SLI when i have saved enough money to get the second 980.

---