Message boards :
Number crunching :
Linux CUDA 'Special' App finally available, featuring Low CPU use
Message board moderation
Previous · 1 . . . 17 · 18 · 19 · 20 · 21 · 22 · 23 . . . 83 · Next
Author | Message |
---|---|
TBar Send message Joined: 22 May 99 Posts: 5204 Credit: 840,779,836 RAC: 2,768 ![]() ![]() |
Well, I did suggest you lock NVIDIA X Server Settings to the Launcher the last time. That one App will provide just about any GPU monitor or control you should need. But I cannot find a gpu-manager file and no variation of ./xserver or ./nvidia-xserver runs anything.The best I can tell, you only need to edit the gpu-manager.conf if you have the ATI/AMD driver installed in Ubuntu 14.04.x. If you're just using the nVidia driver you shouldn't mess with it. Mine is right where it should be /etc/init/gpu-manager.conf, and I have been known to run an ATI card in 14.04.x. |
![]() ![]() Send message Joined: 21 Dec 03 Posts: 169 Credit: 41,799,743 RAC: 0 ![]() |
here's my xorg.conf running 3 1070's with one connected to a monitor. Thermal settings work on all 3 of them: Section "ServerLayout" Identifier "Layout0" Screen 0 "Screen0" 0 0 Screen 1 "Screen1" Screen 2 "Screen2" InputDevice "Keyboard0" "CoreKeyboard" InputDevice "Mouse0" "CorePointer" EndSection Section "Files" FontPath "/usr/lib64/X11/fonts/misc/:unscaled" FontPath "/usr/lib64/X11/fonts/100dpi/:unscaled" FontPath "/usr/lib64/X11/fonts/75dpi/:unscaled" FontPath "/usr/lib64/X11/fonts/misc/" FontPath "/usr/lib64/X11/fonts/Type1/" FontPath "/usr/lib64/X11/fonts/Speedo/" FontPath "/usr/lib64/X11/fonts/100dpi/" FontPath "/usr/lib64/X11/fonts/75dpi/" FontPath "/usr/lib64/X11/fonts/cyrillic/" FontPath "/usr/lib64/X11/fonts/TTF/" EndSection Section "InputDevice" # generated from default Identifier "Mouse0" Driver "mouse" Option "Protocol" "auto" Option "Device" "/dev/psaux" Option "Emulate3Buttons" "no" Option "ZAxisMapping" "4 5" EndSection Section "InputDevice" # generated from default Identifier "Keyboard0" Driver "kbd" EndSection Section "Monitor" Identifier "Monitor0" VendorName "Unknown" ModelName "Unknown" HorizSync 28.0 - 33.0 VertRefresh 43.0 - 72.0 Option "DPMS" EndSection Section "Monitor" Identifier "Monitor1" VendorName "Unknown" ModelName "CRT-0" HorizSync 0.0 - 0.0 VertRefresh 0.0 Option "DPMS" EndSection Section "Monitor" Identifier "Monitor2" VendorName "Unknown" ModelName "CRT-0" HorizSync 0.0 - 0.0 VertRefresh 0.0 Option "DPMS" EndSection Section "Device" Identifier "Device0" Driver "nvidia" VendorName "NVIDIA Corporation" BusID "PCI:1:0:0" EndSection Section "Device" Identifier "Device1" Driver "nvidia" VendorName "NVIDIA Corporation" BusID "PCI:2:0:0" EndSection Section "Device" Identifier "Device2" Driver "nvidia" VendorName "NVIDIA Corporation" BusID "PCI:3:0:0" EndSection Section "Screen" Identifier "Screen0" Device "Device0" Monitor "Monitor0" DefaultDepth 24 Option "Coolbits" "4" SubSection "Display" Depth 24 EndSubSection EndSection Section "Screen" Identifier "Screen1" Device "Device1" Monitor "Monitor1" DefaultDepth 24 Option "ConnectedMonitor" "CRT" Option "Coolbits" "4" Option "Stereo" "0" Option "nvidiaXineramaInfoOrder" "CRT-0" Option "metamodes" "nvidia-auto-select +0+0" SubSection "Display" Depth 24 EndSubSection EndSection Section "Screen" Identifier "Screen2" Device "Device2" Monitor "Monitor2" DefaultDepth 24 Option "ConnectedMonitor" "CRT" Option "Coolbits" "4" Option "Stereo" "0" Option "nvidiaXineramaInfoOrder" "CRT-0" Option "metamodes" "nvidia-auto-select +0+0" SubSection "Display" Depth 24 EndSubSection EndSection |
Stephen "Heretic" ![]() ![]() ![]() ![]() Send message Joined: 20 Sep 12 Posts: 5557 Credit: 192,787,363 RAC: 628 ![]() ![]() |
Well, I did suggest you lock NVIDIA X Server Settings to the Launcher the last time. That one App will provide just about any GPU monitor or control you should need. . . That was on the other machine. I had never been able to get the xserver working on this machine (which I only got running under Linux on Friday). On the first machine I could not remember having to do anything beyong the sudo command. I had thought that would run it and then I could lock it to the launch bar. When it didn't I tried searching on x-server, nvidia-xserver and such variants that I could think of but had no hits. I was simply putting in too much (and incorrect) information. Now that I have managed to fire it up it is locked to the launch bar. Now I only have to work out why BOINC manager will NOT lock to the launch bar. When you right click on the icon the option is there but it has no effect :( . . Since I have no intention of installing any AMD/ATI cards I don't think I will pursue the gpu-manager.conf thing. . . So I now have two machines successfully crunching under Linux. Let's see what they can do .... should only take a couple of weeks. . . I will post you a snapshot of what they are doing now. . . Out of curiosity, what is the difference between cool-bits=4 and cool-bits=28? Or where will I find that information? Stephen :) |
Stephen "Heretic" ![]() ![]() ![]() ![]() Send message Joined: 20 Sep 12 Posts: 5557 Credit: 192,787,363 RAC: 628 ![]() ![]() |
. . Hi TBar, . . The GPUs as they are now. \Sun Mar 19 21:27:09 2017 +-----------------------------------------------------------------------------+ | NVIDIA-SMI 375.39 Driver Version: 375.39 | |-------------------------------+----------------------+----------------------+ | GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC | | Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. | |===============================+======================+======================| | 0 GeForce GTX 106... Off | 0000:01:00.0 On | N/A | | 75% 49C P2 57W / 120W | 1929MiB / 6068MiB | 65% Default | +-------------------------------+----------------------+----------------------+ | 1 GeForce GTX 106... Off | 0000:02:00.0 Off | N/A | | 46% 46C P2 80W / 120W | 1730MiB / 6072MiB | 87% Default | +-------------------------------+----------------------+----------------------+ +-----------------------------------------------------------------------------+ | Processes: GPU Memory | | GPU PID Type Process name Usage | |=============================================================================| | 0 1050 G /usr/lib/xorg/Xorg 111MiB | | 0 2097 G compiz 86MiB | | 0 10559 C ...ome_x41p_zi3k+_x86_64-pc-linux-gnu_cuda80 1727MiB | | 1 10566 C ...ome_x41p_zi3k+_x86_64-pc-linux-gnu_cuda80 1727MiB | +-----------------------------------------------------------------------------+ . . So thanks for your patience, hopefully no questionable results from overheated GPUs here ... Stephen :) |
![]() Send message Joined: 6 Jun 02 Posts: 1668 Credit: 623,086,772 RAC: 156 ![]() ![]() |
For information about --cool-bits you could try the following ... http://lmgtfy.com/?q=--cool-bits%3D28 To overcome Heisenbergs: "You can't always get what you want / but if you try sometimes you just might find / you get what you need." -- Rolling Stones |
Stephen "Heretic" ![]() ![]() ![]() ![]() Send message Joined: 20 Sep 12 Posts: 5557 Credit: 192,787,363 RAC: 628 ![]() ![]() |
For information about --cool-bits you could try the following ... . . Thanks Petri, . . I will have a read. Stephen . |
Stephen "Heretic" ![]() ![]() ![]() ![]() Send message Joined: 20 Sep 12 Posts: 5557 Credit: 192,787,363 RAC: 628 ![]() ![]() |
. . The 1060s are now much, much happier. Mon Mar 20 11:07:47 2017 +-----------------------------------------------------------------------------+ | NVIDIA-SMI 375.39 Driver Version: 375.39 | |-------------------------------+----------------------+----------------------+ | GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC | | Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. | |===============================+======================+======================| | 0 GeForce GTX 106... Off | 0000:01:00.0 On | N/A | | 75% 61C P2 92W / 120W | 1923MiB / 6068MiB | 99% Default | +-------------------------------+----------------------+----------------------+ | 1 GeForce GTX 106... Off | 0000:02:00.0 Off | N/A | | 46% 59C P2 88W / 120W | 1730MiB / 6072MiB | 87% Default | +-------------------------------+----------------------+----------------------+ +-----------------------------------------------------------------------------+ | Processes: GPU Memory | | GPU PID Type Process name Usage | |=============================================================================| | 0 1050 G /usr/lib/xorg/Xorg 106MiB | | 0 2097 G compiz 86MiB | | 0 14755 C ...ome_x41p_zi3k+_x86_64-pc-linux-gnu_cuda80 1727MiB | | 1 14722 C ...ome_x41p_zi3k+_x86_64-pc-linux-gnu_cuda80 1727MiB | +-----------------------------------------------------------------------------+ . . Thanks guys. Stephen |
Stephen "Heretic" ![]() ![]() ![]() ![]() Send message Joined: 20 Sep 12 Posts: 5557 Credit: 192,787,363 RAC: 628 ![]() ![]() |
. . Also, for what it is worth, I seem to have a very acceptable level of inconclusive tasks as well, . . On the 1050ti only 21 out of 761 (2.8%) http://setiathome.berkeley.edu/results.php?hostid=8222433&offset=0&show_names=0&state=3&appid= . . On the 2 x 1060s it has 35 out of 1694 (2.0%) http://setiathome.berkeley.edu/results.php?hostid=8226523&offset=0&show_names=0&state=3&appid= . . That seems pretty good to me. Stephen |
Grant (SSSF) Send message Joined: 19 Aug 99 Posts: 13885 Credit: 208,696,464 RAC: 304 ![]() ![]() |
. . Also, for what it is worth, I seem to have a very acceptable level of inconclusive tasks as well, Actually, they're not that good. GTX 1050 247 pending, 36 inconclusive gives 14.57% GTX 1060s 576 pending, 40 inconclusive gives 6.9% GTX 950 238 pending, 1 inconclusive gives 0.42% (unusually and exceptionally low). Just had a look at my inconclusives, and they're running down around 2.2%, unusually low for them too. Probably due to the low percentage of GBT work? Grant Darwin NT |
Stephen "Heretic" ![]() ![]() ![]() ![]() Send message Joined: 20 Sep 12 Posts: 5557 Credit: 192,787,363 RAC: 628 ![]() ![]() |
[ur]http://us.download.nvidia.com/XFree86/Linux-x86_64/364.19/README/xconfigoptions.html[/url] Option "Coolbits" "integer" Enables various unsupported features, such as support for GPU clock manipulation in the NV-CONTROL X extension. This option accepts a bit mask of features to enable. WARNING: this may cause system damage and void warranties. This utility can run your computer system out of the manufacturer's design specifications, including, but not limited to: higher system voltages, above normal temperatures, excessive frequencies, and changes to BIOS that may corrupt the BIOS. Your computer's operating system may hang and result in data loss or corrupted images. Depending on the manufacturer of your computer system, the computer system, hardware and software warranties may be voided, and you may not receive any further manufacturer support. NVIDIA does not provide customer service support for the Coolbits option. It is for these reasons that absolutely no warranty or guarantee is either express or implied. Before enabling and using, you should determine the suitability of the utility for your intended use, and you shall assume all responsibility in connection therewith. When "2" (Bit 1) is set in the "Coolbits" option value, the NVIDIA driver will attempt to initialize SLI when using GPUs with different amounts of video memory. When "4" (Bit 2) is set in the "Coolbits" option value, the nvidia-settings Thermal Monitor page will allow configuration of GPU fan speed, on graphics boards with programmable fan capability. When "8" (Bit 3) is set in the "Coolbits" option value, the PowerMizer page in the nvidia-settings control panel will display a table that allows setting per-clock domain and per-performance level offsets to apply to clock values. This is allowed on certain GeForce GPUs. Not all clock domains or performance levels may be modified. When "16" (Bit 4) is set in the "Coolbits" option value, the nvidia-settings command line interface allows setting GPU overvoltage. This is allowed on certain GeForce GPUs. When this option is set for an X screen, it will be applied to all X screens running on the same GPU. The default for this option is 0 (unsupported features are disabled). . . So this tells me that the cool-bits value of 32 referred to in some forum messages is rubbish, it is not supported. Nor is any odd value. 00000 . . Value 0 - disables all coolbits options 00010 . . Value 2 - enables SLI configuration for GPUs with disparate amounts of memory. 00100 . . Value 4 - enables fan speed control on the Thermal Monitor page. 01000 . . Value 8 - enables GPU clock control on the Powermizer page. 10000 . . Value 16 - enables GPU over voltage control by commandline. . . So 28 is enabling Clock and Voltage control for over-clocking the GPU as well as fan control for cooling. . . For those wanting cooling only 4 is the best value. . . Interesting that if you have multiple GPUs running in SLI (with different memory sizes) you need to have a value of 2,6,10,18 or 30 ... . . Sorry for the redundancy but I thought some people might like to know about the voiding warranty caveat. Stephen . [/url] |
![]() ![]() ![]() ![]() Send message Joined: 1 Dec 99 Posts: 2786 Credit: 685,657,289 RAC: 835 ![]() ![]() |
There are 32 valid coolbits values and all are valid. I'll let you think that over for awhile :) |
TBar Send message Joined: 22 May 99 Posts: 5204 Credit: 840,779,836 RAC: 2,768 ![]() ![]() |
. . Sorry for the redundancy but I thought some people might like to know about the voiding warranty caveat.You didn't seem very concerned about that here; Where is a Linux version of Afterburner ..... wwwwaaahhhhhhh!Or do you somehow think nVidia will allow some Third party App to use nVidia's own tweaks without the same penalties? AfterBurner is based on RivaTuner. RivaTuner has been around for quite some time. I remember back when it was standard practice for a Repair facility to Immediately Void the warranty if you were stupid enough to turn in your machine with RivaTuner on it. You asked for a tweak tool such as Afterburner, and someone provided it. I use Coolbits 12 myself, and after a few attempts with the OCing I gave up on it. I still use the Fan setting though. |
![]() ![]() ![]() ![]() Send message Joined: 1 Dec 99 Posts: 2786 Credit: 685,657,289 RAC: 835 ![]() ![]() |
No one warned me of hitting enter too soon with an sudo chmod command, but live and learn when you have 3 windows open for reading/ testing and misplace where you are. BTW sudo chmod -R /etc is BAD, and you lose rights to sudo! Luckily I had a root window open to fix it, and the sys log pointed me to other issues to fix, and my USB boot disk was nice to compare to what needed fixing. I'm sure I will find more, but it seems fine now. |
rob smith ![]() ![]() ![]() Send message Joined: 7 Mar 03 Posts: 22667 Credit: 416,307,556 RAC: 380 ![]() ![]() |
and "sudo rm -R -f *.*" from the root of the filing system is even worse.... Bob Smith Member of Seti PIPPS (Pluto is a Planet Protest Society) Somewhere in the (un)known Universe? |
Stephen "Heretic" ![]() ![]() ![]() ![]() Send message Joined: 20 Sep 12 Posts: 5557 Credit: 192,787,363 RAC: 628 ![]() ![]() |
There are 32 valid coolbits values and all are valid. I'll let you think that over for awhile :) . . Hmmm you would need to explain that idea. Since there is no function assigned to the 0 bit with a value of 1, then any odd value will have the 0 bit set to one and is invalid. Whether the server software itself can simply overlook that and treat it as the next lower even value I don't know and don't want to try. But all even values from 0 to 30 have significance. Stephen ? |
Stephen "Heretic" ![]() ![]() ![]() ![]() Send message Joined: 20 Sep 12 Posts: 5557 Credit: 192,787,363 RAC: 628 ![]() ![]() |
. . Sorry for the redundancy but I thought some people might like to know about the voiding warranty caveat.You didn't seem very concerned about that here; . . No actually, I was after a monitoring tool like Afterburner. Have you ever used the monitor functions it provides? They are brilliant. Without it I feel blind trying to deal with running the GPU. I am not into overclocking. Fascinating though that manufacturers provided overclocking tools/functionality but threaten to invalidate your warranty if you use them? . . Hence why I was interested in knowing what the different values for cool-bits mean. The fan controls were my most critical concern and that is another thing for which I rely on Afterburner, it has a great fan profile controller. Not a single fixed speed fan setting, like cool-bits but a nice pan-temperature profile to let it run quiet when cool and stepping up to deal with the temps from heavy use, keeping the running temps right where I want them. But I guess anyone can get the wrong end of the stick when they are coming from a different perspective. . . I had wanted to know why some ppl suggested a value of 28 when most others were just recommending 4. Well now I have found out. For future reference I only need a value of 4 but since it is running I will let it go until if or when I need to do something more about configuration. Stephen . |
rob smith ![]() ![]() ![]() Send message Joined: 7 Mar 03 Posts: 22667 Credit: 416,307,556 RAC: 380 ![]() ![]() |
Using coolbits = 4 (or coolbits = 28) there is an "afterburner style" fan control available, it is buried within the nVidia control panel (sorry I can't remember what it is called). I tend just push fans to the max, as the GPUs are running 24/7 and close to 100%, with an ambient temperature close to 30C at this time of year, and rising towards 40C (or beyond) in the summer. The reason for some people suggesting 28, and others 4 is quite simple - 28 gives you a lot of overclocking controls including fan control, while 4 only gives you fan control. Bob Smith Member of Seti PIPPS (Pluto is a Planet Protest Society) Somewhere in the (un)known Universe? |
TBar Send message Joined: 22 May 99 Posts: 5204 Credit: 840,779,836 RAC: 2,768 ![]() ![]() |
. . Sorry for the redundancy but I thought some people might like to know about the voiding warranty caveat.You didn't seem very concerned about that here; Maybe you should just go back to windows since you and Linux will probably never get along. If you are just after monitoring tools then the Stock NVIDIA X Server Settings is all you need. Afterburner also provides OCing tools just as Coolbits 28, if you don't want to use particular functions of 28, then Don't use them. Just as with Afterburner. You were given the Coolbits number that best matches the functions in Afterburner, which is what you asked for. Those overclocking tools/functionality is provided by nVidia because people asked for them, mainly people such as MSI, EVGA, etc. Imagine the third party vendors trying to build their modified nVidia cards without those tools. I doubt nVidia ever intended those tweaks to make it to the public, but since they did there isn't much they can do about it, except warn that misuse can bring penalties. They could get really nasty with the vendors, but then hacking tools such as RivaTuner would become even more popular. I find nothing unusual about the situation. Those fan profiles are only useful for varying loads, and not helpful for work such as SETI work where the GPU remains maxed out continuously, having to listen to fans constantly ramp up and down is very annoying to most people. Personally, I'd rather have constant noise. All I can say is if you need to sit and stare at a particular GPU monitoring tool all day, maybe you should go there. I mean, why stay where you're not happy? ;-) |
Stephen "Heretic" ![]() ![]() ![]() ![]() Send message Joined: 20 Sep 12 Posts: 5557 Credit: 192,787,363 RAC: 628 ![]() ![]() |
Maybe you should just go back to windows since you and Linux will probably never get along. . . If your app ever appears as a Windows app in a Lunatics release I may do just that. Perhaps that would make you happier than you seem to be right now. If you are just after monitoring tools then the Stock NVIDIA X Server Settings is all you need. Afterburner also provides OCing tools just as Coolbits 28, if you don't want to use particular functions of 28, then Don't use them. . . Yes I know that functionality exists in Afterburner, and I don't use it there either. Mainly I needed fan control for cooling which I now have thanks to those who provided the info on using Xserver. I was curious about the difference between the setting values that people were giving, so I found out, and having done so I posted the info for any other Linux newbies like myself who might find it useful. Just as with Afterburner. You were given the Coolbits number that best matches the functions in Afterburner, which is what you asked for. Those overclocking tools/functionality is provided by nVidia because people asked for them, mainly people such as MSI, EVGA, etc. Imagine the third party vendors trying to build their modified nVidia cards without those tools. I doubt nVidia ever intended those tweaks to make it to the public, but since they did there isn't much they can do about it, except warn that misuse can bring penalties. They could get really nasty with the vendors, but then hacking tools such as RivaTuner would become even more popular. I find nothing unusual about the situation. . . I am not privvy to what happens within Nvidia Corp or the manufacturing companies but I feel confident that they have access to much lower level routines to effect fan and clock control than a user level interface such as xserver. But I found the documentation ironic and that amused me, if you are unable to appreciate the irony in that I am sorry. You have been very helpful and it was never my intention to offend you. Those fan profiles are only useful for varying loads, and not helpful for work such as SETI work where the GPU remains maxed out continuously, having to listen to fans constantly ramp up and down is very annoying to most people. Personally, I'd rather have constant noise. . . While I generally have my rigs crunching 24/7 there are times when I do housekeeping on them and the fans do not need to be kept at high speed. The beauty of the continuous speed control over the whole range of temps is that the changes in fan speed for either set of conditions is fairly small but can keep the temps constant despite the varying loads within each set of conditions. I find it very useful and satisfying in its operation, and reassuring that when some condition such as the room temp goes up or it strikes a particularly chewy guppi and gets hot under the collar the fan control can and will compensate. There is very little audible change in this process so it not in any way grating for me. It may not be the same for all users but I suspect the companies that spend the money making such tools are betting it is for many. All I can say is if you need to sit and stare at a particular GPU monitoring tool all day, maybe you should go there. I mean, why stay where you're not happy? ;-) . . Since I am not in front of the computers all day (though probably more than is healthy :) ) that isn't an issue. But I do like to check when returning to the room that things are running well, and a quick glimpse at the monitor screen can be instantly re-assuring, or an instant alert if things are not running as they should. Call me OCD if you like but I appreciate getting feedback on what is happening on these beasties. Also when setting things up or modifying them such tools provide fast feedback on whether you have it right or not. I do not often look at the tacho in my car, but when I drive a car without one I soon notice it's absence. Imagine a car without a speedo and you might get the feeling LOL. Stephen :) |
TBar Send message Joined: 22 May 99 Posts: 5204 Credit: 840,779,836 RAC: 2,768 ![]() ![]() |
. . I am not privvy to what happens within Nvidia Corp or the manufacturing companies but I feel confident that they have access to much lower level routines to effect fan and clock control than a user level interface such as xserver. But I found the documentation ironic and that amused me, if you are unable to appreciate the irony in that I am sorry. You have been very helpful and it was never my intention to offend you.Just curious. Since the Coolbits options are not included in the Public release of nVidia Settings, it would appear they are not intended for the general public. So, who do you think they where intended for? Now that the Vendors are releasing Software that uses the built-in nVidia tweaks, the choice nVidia has is to either make them available to those that want them, or see those people use someone else's software not under Nvidia's control. It's somewhat similar to an Automobile that can do well over 100mph. You will almost never be able to reach over 100mph as in most cases you will be penalized for trying. You can play around with Coolbits all you wish, but if you go to extremes and damage the hardware, there will be penalties. Nothing unusual there. I just leave the BOINC Manager on my desktop, sometimes with NVIDIA-SMI displayed. It works for most people. |
©2025 University of California
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.