New GPU - I hope I have settings wrong...

Here is my log:

1/16/2013 5:57:50 PM | | Starting BOINC client version 7.0.28 for windows_x86_64
1/16/2013 5:57:50 PM | | log flags: file_xfer, sched_ops, task
1/16/2013 5:57:50 PM | | Libraries: libcurl/7.25.0 OpenSSL/1.0.1 zlib/1.2.6
1/16/2013 5:57:50 PM | | Data directory: C:\ProgramData\BOINC
1/16/2013 5:57:50 PM | | Running under account Nobody316
1/16/2013 5:57:50 PM | | Processor: 8 AuthenticAMD AMD FX(tm)-8350 Eight-Core Processor [Family 21 Model 2 Stepping 0]
1/16/2013 5:57:50 PM | | Processor: 2.00 MB cache
1/16/2013 5:57:50 PM | | Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 htt pni ssse3 cx16 sse4_1 sse4_2 syscall nx lm svm sse4a osvw ibs xop skinit wdt lwp fma4 page1gb rdtscp
1/16/2013 5:57:50 PM | | OS: Microsoft Windows 7: Ultimate x64 Edition, Service Pack 1, (06.01.7601.00)
1/16/2013 5:57:50 PM | | Memory: 15.97 GB physical, 15.97 GB virtual
1/16/2013 5:57:50 PM | | Disk: 511.34 GB total, 361.12 GB free
1/16/2013 5:57:50 PM | | Local time is UTC -8 hours
1/16/2013 5:57:50 PM | | NVIDIA GPU 0: GeForce GTX 650 Ti (driver version 310.90, CUDA version 5.0, compute capability 3.0, 2048MB, 8382372MB available, 1586 GFLOPS peak)
1/16/2013 5:57:50 PM | | NVIDIA GPU 1: GeForce 9400 GT (driver version 310.90, CUDA version 5.0, compute capability 1.1, 512MB, 471MB available, 130 GFLOPS peak)
1/16/2013 5:57:50 PM | | Config: use all coprocessors
1/16/2013 5:57:50 PM | Einstein@Home | URL http://einstein.phys.uwm.edu/; Computer ID 6450237; resource share 200
1/16/2013 5:57:50 PM | SETI@home | URL http://setiathome.berkeley.edu/; Computer ID 6887520; resource share 50
1/16/2013 5:57:50 PM | Einstein@Home | General prefs: from Einstein@Home (last modified 16-Jan-2013 00:35:03)
1/16/2013 5:57:50 PM | Einstein@Home | Host location: none
1/16/2013 5:57:50 PM | Einstein@Home | General prefs: using your defaults

Now if you note there is a difference here is your part of the log:

you have device version OpenCL 1.1 Cuda which I don't I am still researching don't give up we will get this some way... as far as old logs you have to open the dir. it should be something like this :

C:\ProgramData\BOINC in windows 7 and it may be hidden and you might "if not already" have to show hidden files/folders. It should be in there somewhere not for sure as to where yet. Check your preferences on both sites and client. To run both of my cards I had to make a cc_config and put it in the dir. the text of that file is:

1

Also had to use dummy plug for the GT 9400 but both works. Let me know when you have checked the preferences and what you have found/not found.. Will post more when I find something

Edit:
Are you using the new card or the on board card? "What is the monitor plugged in too"

Edit again: lol

If I know and understand right the only card that can run 2 on 1 card is the 690 as it is like a duo core but I could be wrong. I am only running 1 unit per card have not seen a way to run 2at the same time though I have heard it is possible but I think only for the 690 card. Will look more into this as I can. Bare with me as I am on and off alot. lol maybe someone else who has viewed this but has not answered might know or can link to another forum where it is talked about...

Here are some points for you to consider:-

• * A GT630 is a pretty low end GPU. For a few more dollars, a GTX650 would give you a big leap in performance.
  * If you have changed the GPU utilisation factor in your project prefs, to run multiple GPU tasks simultaneously and if you manage to free up enough GPU memory for both tasks to run, don't be surprised to see the run time to essentially double. In other words, you may see very little, if any, performance gain by running 2x.
  * If you do decide to run 2x, you will most likely benefit a little by freeing up a CPU core to better support the feeding of the GPU. You would do this by changing your computing prefs to use only 50% of the processors. Of course, the downside is that all BOINC projects on that host would be able to use only 1 CPU for CPU tasks.
  * Your motherboard is almost certainly PCIe version 1. You would see a significant performance improvement if the card were in a PCIe version 2 slot.
  

So, unfortunately, the upshot of all this is that, without changing hardware, you are probably about your best performance with 2 CPU tasks and 1 GPU tasks. Look on the bright side - at least the GPU task is faster than a CPU task and gives you 500 credits instead of 70 :-).

I have no personal experience but I'm sure I've read somewhere that you could expect better performance from GDDR5. I don't think that 1GB RAM and running 2X will do anything except pretty much double the time it takes to run a task 1x. I think you may well get best performance from the GDDR5 card running tasks singly. Ultimately, the only way to find out for sure is to give it a go and actually measure it :-).

You will obviously use more memory if you run 2x. You should be able to run 2x on the 1GB card just fine but I doubt you could on the 512MB card. In the end I don't think you will gain anything from 2x but you wont know for sure until you try. If you are planning to change to the GDDR5 card, give the current one a run at 2x and take some stats to help you decide before returning it.

You don't need either unless you have some special BOINC client configuration needs - which I don't think you have. In that case you would only need cc_config.xml.