Jump to content
InsanelyMac Forum

meroy

Members
  • Content count

    44
  • Joined

  • Last visited

About meroy

  • Rank
    InsanelyMac Protégé
  1. Lion is soon out and folks will be able to benchmark GTX 5xx series. The following is taken from my Windows 7 box running a pair of GTX 560's. They came factory OC'd at 900/1800/2004 (4008) 1.012 volts. However, I under-clocked/under-volted them down to 855/1710/2100 (4200) 0.987 volts. Single-Precision: ./nbody -n=61440 -benchmark -- One GPU: 548.804 single-precision GFLOP/s -- Two GPUs via -numdevices=2: 1068.541 single-precision GFLOP/s Double-Precision: ./nbody -fp64 -n=30720 -benchmark -- One GPU: 89.702 double-precision GFLOP/s -- Two GPUs via -numdevices=2: 166.125 double-precision GFLOP/s I wish that NVIDIA will one day make a single-PCB card containing 2 GTX 560's to have a good balance for compute-power and electric power utilization.
  2. One can even benchmark double-precision via the N-Body CUDA demo. GTX 4xx owners will be able to see a much larger increase in performance when comparing to GTX 2xx cards. Here are my results: ./nbody -fp64 -n=30720 -benchmark GTX 295 OC'ed: -- One GPU: 66.966 double-precision GFLOP/s -- Two GPUs via -numdevices=2: 128.672 double-precision GFLOP/s GTX 295 (standard clocks): -- One GPU: 55.010 double-precision GFLOP/s -- Two GPUs via -numdevices=2: 106.478 double-precision GFLOP/s This is where a GTX 4xx can shine over a GTX 2xx variant. Try -n=15360 if -n=30720 reports unspecified launch failure. Not all cards support double-precision.
  3. Nvidia GTX 295 Black Screen

    Steps to go from 10.6.7 to 10.6.8: (for folks already with a working 10.6.7 or earlier and a GTX 2xx card) 1. Ensure your gfx.plist has device_type set correctly. Read my previous post here in the forum. 2. Create the hex via gfxutil and update the EFI string in your boot plist file. 3. Reboot to ensure your 10.6.7 is working 4. Then, follow Tommy's OS update via [url="http://www.insanelymac.com/forum/topic/279450-why-insanelymac-does-not-support-tonymacx86/"]#####[/url] 5. The 10.6.8 OS should boot up fine with the graphics card using the update boot plist Afterwards, you can then set GraphicsEnabler to No and test or just leave it if all is fine. GTX 295 folks will need to set that to No and update to Nvidia's driver for CUDA to work.
  4. Nvidia GTX 295 Black Screen

    My GraphicsEnabler is set to No in my boot.plist file. Here's a suggestion for folks who currently have a working 10.6.7 and a GTX 2xx card. 1. Look at the device_type entry 2. 10.6.8 requires that to be set to NVDA,Parent versus NVDA,GeForce Previously, we had this with 10.6.7 and below. <key>device_type</key> <string>NVDA,GeForce</string> <key>name</key> <string>NVDA,Parent</string> In 10.6.8, it needs to be the following: (name is not necessary). This works with both 10.6.7 and 10.6.8 <key>device_type</key> <string>NVDA,Parent</string> So, just make that change to your current EFI plist file and update the EFI string in your boot plist file prior to updating to 10.6.8. Folks running a GTX 295 card (single-PCB edition) will need Nvidia's Quadro 4000 drivers in order for CUDA to work. The reason is that Apple's native Nvidia driver is not able to properly configure one of the GPUs. However, the system will load and OpenGL is fine. I hope this helps folks out there who are currently running a GTX 2xx card successfully with 10.6.7 or earlier and wanting to update to 10.6.8. Tip for the GTX 295 single-PCB edition: A GTX 295 single-PCB edition requires the device-type entry to be set to NVDA,GeForce for the first GPU (0x0,0x0) and NVDA,Parent for the 2nd GPU (0x2,0x0). A GTX 295 dual-PCB edition (I believe) may need device_type for both GPUs to be set to NVDA,Parent. (I'm not sure... someone with a dual-PCB can verify...) - meroy
  5. Hi all, Just wanted to chime in to say that I finally got sleep working on my system with the GTX 295 card. http://www.insanelymac.com/forum/index.php...t&p=1717189 Note: The culprit was was with the device-type. Each GPU is different. The one having display connections is set to NVDA,Parent and the other to NVDA,GeForce.
  6. Nvidia GTX 295 support

    At last, success is achieved with 10.6.8 and a GTX 295 card (Single PCB edition). http://www.insanelymac.com/forum/index.php...t&p=1717189 Note: The culprit was was with the device-type. Each GPU is different. The one having display connections is set to NVDA,Parent and the other to NVDA,GeForce.
  7. Nvidia GTX 295 Black Screen

    After some time, my system is now able to sleep properly. The 2 extensions EvilAppleACPIPlatform.kext and EvilIOPCIFamily.kext (I believe) will break sleep. But I needed those in order for my GTX 295 to work. So, I looked deeper and the problem was with my EFI string. 10.6.7 and earlier were more forgiving, but not 10.6.8. I removed the 2 Evil* extensions and now able to run 10.6.8 with Apple's drivers and also Nvidia's Quadro 4000 drivers as well. The latter is required for CUDA to work when not using the Evil* extensions. The single-PCB edition has a GPU without any physical display ports and that causes CUDA to fail as something doesn't initialize correctly with Apple's drivers. My new plist which works with both Apple and Nvidia's drivers and 10.6.8 without the Evil* extensions: (This may not work for the dual-PCB edition) Note: The culprit was was with the device-type. Each GPU is different. The one having display connections is set to NVDA,Parent and the other to NVDA,GeForce. I'm using a new NVCAP value. I need to test dual displays though. The NVCAP value was obtain from the ROM. <?xml version="1.0" encoding="UTF-8"?> <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd"> <plist version="1.0"> <dict> <key>PciRoot(0x0)/Pci(0x3,0x0)/Pci(0x0,0x0)/Pci(0x0,0x0)/Pci(0x0,0x0)</key> <dict> <key>@0,compatible</key> <string>NVDA,NVMac</string> <key>@0,device_type</key> <string>display</string> <key>@0,name</key> <string>NVDA,Display-A</string> <key>NVCAP</key> <data>BAAAAAAADwAMAAAAAAAABwAAAAA=</data> <key>VRAM,totalsize</key> <data>AAAAOA==</data> <key>device_type</key> <string>NVDA,GeForce</string> <key>model</key> <string>GeForce GTX 295</string> <key>rom-revision</key> <string>62.00.6C.00.05</string> </dict> <key>PciRoot(0x0)/Pci(0x3,0x0)/Pci(0x0,0x0)/Pci(0x2,0x0)/Pci(0x0,0x0)</key> <dict> <key>@0,compatible</key> <string>NVDA,NVMac</string> <key>@0,device_type</key> <string>display</string> <key>@0,name</key> <string>NVDA,Display-A</string> <key>@1,compatible</key> <string>NVDA,NVMac</string> <key>@1,device_type</key> <string>display</string> <key>@1,name</key> <string>NVDA,Display-B</string> <key>NVCAP</key> <data>BAAAAAAADwAMAAAAAAAABwAAAAA=</data> <key>VRAM,totalsize</key> <data>AAAAOA==</data> <key>device_type</key> <string>NVDA,Parent</string> <key>model</key> <string>GeForce GTX 295</string> <key>rom-revision</key> <string>62.00.6C.00.04</string> </dict> </dict> </plist>
  8. My EVGA GTX 295 Co-op edition is over-clocked and has been running like that since several years now. The settings has allowed it to match a GTX 285 (times 2) in CUDA performance. I'm going to restore the 2 ROMs (one per GPU) back to the factory settings and report back the N-body results for the standard clocks. Using the factory clocks, the N-body results against both GPUs running simultaneously are: (via -numdevices=2) ./nbody -n=61440 -numdevices=2 -benchmark [nbody] starting... Run "nbody -benchmark [-n=<numBodies>]" to measure perfomance. -fullscreen (run n-body simulation in fullscreen mode) -fp64 (use double precision floating point values for simulation) -numdevices=N (use first N CUDA devices for simulation) > Windowed mode > Simulation data stored in system memory > Single precision floating point simulation > 2 Devices used for simulation > Compute 1.3 CUDA device: [GeForce GTX 295] > Compute 1.3 CUDA device: [GeForce GTX 295] 61440 bodies, total time for 10 iterations: 847.679 ms = 44.532 billion interactions per second = 890.638 single-precision GFLOP/s at 20 flops per interaction [nbody] test results... PASSED And the N-body result against one GPU on the GTX 295: ./nbody -n=61440 -benchmark [nbody] starting... Run "nbody -benchmark [-n=<numBodies>]" to measure perfomance. -fullscreen (run n-body simulation in fullscreen mode) -fp64 (use double precision floating point values for simulation) -numdevices=N (use first N CUDA devices for simulation) > Windowed mode > Simulation data stored in video memory > Single precision floating point simulation > 1 Devices used for simulation > Compute 1.3 CUDA device: [GeForce GTX 295] 61440 bodies, total time for 10 iterations: 1687.764 ms = 22.366 billion interactions per second = 447.322 single-precision GFLOP/s at 20 flops per interaction [nbody] test results... PASSED
  9. Hi all. My GTX 295 Co-op edition is up and running in Snow Leopard 10.6.8 using Apple's native drivers in 10.6.8 and the OpenGL framework. There are 2 GPUs on this card, therefore x2 the results. I added a post to a thread on insanelymac to give solution for GTX 295 owners who want to update to 10.6.8: http://www.insanelymac.com/forum/index.php...p;#entry1716804 Basically, there are changes made in 10.6.8 which render the existing EFI string useless unless 2 extensions are installed in /Extra/Extensions. The following is the result of combining both GPUs on the GTX 295 card for the N-Body demo in CUDA 4.0. ./nbody -numdevices=2 -n=61440 -benchmark [nbody] starting... Run "nbody -benchmark [-n=<numBodies>]" to measure perfomance. -fullscreen (run n-body simulation in fullscreen mode) -fp64 (use double precision floating point values for simulation) -numdevices=N (use first N CUDA devices for simulation) > Windowed mode > Simulation data stored in system memory > Single precision floating point simulation > 2 Devices used for simulation > Compute 1.3 CUDA device: [GeForce GTX 295] > Compute 1.3 CUDA device: [GeForce GTX 295] 61440 bodies, total time for 10 iterations: 697.011 ms = 54.158 billion interactions per second = 1083.161 single-precision GFLOP/s at 20 flops per interaction [nbody] test results... PASSED
  10. Nvidia GTX 295 Black Screen

    The following maybe of help for folks running 10.6.7 or earlier with a NVIDIA GTS/GTX 2xx graphics card and wanting to update to 10.6.8. Prior to the 10.6.8 update, download two kexts by nawcom http://prasys.info/legacy_kexts_by_nawcom.zip Unzip the file and use Kext Helper to install both EvilAppleACPIPlatform.kext and EvilIOPCIFamily.kext to /Extra/Extensions. This ensured my GTX 295 graphics card will continue to work after the 10.6.8 update. No changes were needed to my boot plist file which has the EFI string for my GTX 295 card. Thanks to nawcom for providing the two kexts. The system sleep is failing though under 10.6.8 for me with this method. However, the display can sleep and wake up without any problems with Apple's native 10.6.8 drivers. - meroy
  11. Nvidia GTX 295 support

    The following maybe of help for folks running 10.6.7 or earlier with a NVIDIA GTS/GTX 2xx graphics card and wanting to update to 10.6.8. Prior to the 10.6.8 update, download two kexts by nawcom http://prasys.info/legacy_kexts_by_nawcom.zip Unzip the file and use Kext Helper to install both EvilAppleACPIPlatform.kext and EvilIOPCIFamily.kext to /Extra/Extensions. This ensured my GTX 295 graphics card will continue to work after the 10.6.8 update. No changes were needed to my boot plist file which has the EFI string for my GTX 295 card. Thanks to nawcom for providing the two kexts. The system sleep is failing though under 10.6.8 for me with this method. However, the display can sleep and wake up without any problems with Apple's native 10.6.8 drivers. - meroy
  12. 10.5.6 and Belkin F7D1101 v.1

    The N300 Belkin Surf & Share Wireless USB Adapter (F7D2101 v1) also works with the same driver.
  13. 10.5.6 and Belkin F7D1101 v.1

    Thanks for this thread. The Belkin F7D1101 Wireless USB adapter is working for me under Snow Leopard 64-bit (10.6.4).
  14. fermi/gtx 480/gtx 470

    -- Snow Leopard 10.6.4 Here are the results for the OpenGL Extension Viewer / Xbench running on a GTX 295 Single-PCB edition at 684/1512/1180 MHz. I'm posting this here so folks can compare with the next driver update suited for GTX 470/480. I too am interested in the new GTX 470/480 Fermi architecture and will come back to this thread when the drivers are updated to fully support Fermi. Only one GPU is used during testing for the results provided below. -- Benchmark 1920x1200x32 1.1 2250 FPS 1.2 3371 FPS 1.3 2837 FPS 1.4 2272 FPS 1.5 2286 FPS 2.0 1580 FPS 2.1 203 FPS -- Single Cube 1920x1200x32 1.1 4625 FPS 1.2 4655 FPS 1.3 4649 FPS 1.4 4579 FPS 1.5 4589 FPS 2.0 4414 FPS 2.1 231 FPS -- Xbench Quartz Graphics Test: 388.27 OpenGL Graphics Test: 315.70 User Interface Test: 601.92 CUDA 3.1 now allows one to specify the GPU devices to use via the CUDA_VISIBLE_DEVICES environment variable. I have mine set to 1. This allows a better balanced system for me. The user interface (OpenGL) will run on device 0 and remain fluid and fast. CUDA apps will run on device 1. This allows both GPUs on the GTX 295 to be utilized by OS X. I like this feature so very much. In addition, I added to /S/L/E/AppleGraphicsPowerManagement.kext/Contents/Info.plist. The entry is added below MacPro4,1. ... <key>MacPro4,1</key> <dict> [color="#0000FF"] <key>Vendor10deDevice05eb</key> <dict> <key>Heuristic</key> <dict> <key>ID</key> <integer>0</integer> <key>IdleInterval</key> <integer>250</integer> <key>SensorOption</key> <integer>1</integer> <key>SensorSampleRate</key> <integer>4</integer> <key>TargetCount</key> <integer>5</integer> <key>Threshold_High</key> <array> <integer>87</integer> <integer>87</integer> <integer>87</integer> <integer>100</integer> </array> <key>Threshold_Low</key> <array> <integer>0</integer> <integer>87</integer> <integer>87</integer> <integer>87</integer> </array> </dict> <key>LogControl</key> <integer>0</integer> <key>control-id</key> <integer>18</integer> </dict>[/color] <key>Vendor10deDevice05e2</key> <dict> ... The results for SmallLuxGPU (OpenCL) are posted here: http://www.insanelymac.com/forum/index.php...p;#entry1499698
  15. Core i7 870 @ 3.68Ghz + EVGA GTX 295 Co-op Edition (Single PCB) @ 684/1512/1180 MHz -- Utilizing both GPUs GPU Only Midrange: 5.3sec GPU Only Highend: 6.5sec GPU Only Ultra: 15.7sec -- Only one GPU is used for the Pixelfilter test FILTER NONE: 855.93M FILTER PREVIEW: 292.16M FILTER GAUSSIAN: 177.43M -- Utilizing both GPUs (Luxball glass Scene) GPU1: Rays/sec 5714K GPU2: Rays/sec 5715K Avg. Rays/sec 11428K
×