Message boards :
Number crunching :
Run Multiple WU's on Your GPU
Message board moderation
Previous · 1 · 2 · 3 · 4 · 5 · 6 . . . 11 · Next
Author | Message |
---|---|
Send message Joined: 25 Jan 11 Posts: 271 Credit: 346,072,284 RAC: 0 |
Im running at 1100 Mhz Currently , takes 1min 20 to do 6 WU ( 3 on each card ) while i haven't read through the entire thread in detail, i suspect that some of the folks claiming to get substantial increases in compute efficiency aren't specifically talking about the Milkyway@Home project (despite the fact that we are in the MW@H forums). remember, the app_config.xml file is generic and can be used for a number of different GPU compute-capable projects...so while people should be talking about it here in a "MW@H-only" context, the personal experiences with the app_config.xml file being talked about here may not strictly represent experiences w/ MW@H. the bottom line is that a single MW@H task tends to max out any GPU's utilization - hence the reason most folks only see negligible increases in compute efficiency (if any at all) when running more than one MW@H task simultaneously. if you had looked at your GPU usage in GPU-Z/MSI Afterburner/etc back when you were only running 1 MW@H task at a time, you would have noticed that your GPU usage was probably already pegged at or near 100%. if a single task runs with little or no apparent additional GPU resources, then it should be obvious that 2 simultaneous tasks will take approx. twice as long to crunch, 3 simultaneous tasks will take approx. three times as long to crunch, and so on and so forth. that said, the real increase in PPD that people experience when running multiple MW@H tasks simultaneously comes from "eliminating" the handful of seconds of a MW@H task during which the GPU is idle, as others have already mentioned above. you see, while a majority of a MW@H task is run on the GPU, a small fraction of it actually runs on the CPU (which lasts only a handful of seconds, depending on the GPU). the solution is to run 2 MW@H tasks simultaneously, no more, no less. by doing so and staggering their start times (i.e. not letting 2 tasks start at exactly the same time), the GPU will process 2 tasks almost all of the time, and will process at least one task when the other task is being offloaded to the CPU for a few seconds worth of calculations. thus, the GPU is "never" idle. i put the word never in quotes b/c while true in theory, the GPU doesn't split its resources perfectly and allocate them equally among multiple simultaneously running MW@H tasks in reality, and thus the tasks don't always stay staggered. in other words, pairs of simultaneously running MW@H tasks tend to oscillate between a staggered state and a synchronized state...don't worry though - most of the time they'll run staggered b/c it is statistically very unlikely for that pair of tasks to be in sync more often than they're out of sync. the reason it is pointless to run more than 2 simultaneous MW@H GPU tasks is twofold: 1) it only takes a single task to keep the GPU fully (or almost fully) utilized while the other task is running on the CPU for a short time, and 2) as i mentioned in paragraph two above, MW@H GPU task run times don't decrease by running tasks in parallel (unlike many other DC projects) - the only thing that this does is eliminate GPU idle time. |
Send message Joined: 22 Mar 09 Posts: 6 Credit: 780,375,873 RAC: 18,965 |
Can't get two WU to run on my ATI 6970 using the configuration posted below. I had set CPU's to 12 but found that MW@H always indicated a full CPU usage and cutting back to 11 while running climateprediction.net cut my MW@H times in half. Any suggestions? Doesn't seem like the app_config.xml is really being used even though it's found since the CPU usage always says 0.949 CPU. app_config: <app_config> account xml: <resource_share>100</resource_share> BOINC log: CAL: ATI GPU 0: AMD Radeon HD 6900 series (Cayman) (CAL version 1.4.1848, 2048MB, 2016MB available, 6758 GFLOPS peak) |
Send message Joined: 31 Mar 08 Posts: 22 Credit: 84,159,673 RAC: 0 |
Thank you for this post. I have 3x HD7950 in my crunching box and MW was using only2. I now run 1WU on each GPU, which suits me fine, as overheating is an issue so I don't want to run more than 1 WU ATM. (but at least I know how to fix it) <app_config> <app> <name>milkyway</name> <max_concurrent>3</max_concurrent> <gpu_versions> <gpu_usage>1</gpu_usage> <cpu_usage>.05</cpu_usage> </gpu_versions> </app> </app_config> No need to restart BOINC BTW, just issue a "read config files" from the "advanced menu" |
Send message Joined: 10 May 11 Posts: 3 Credit: 31,650,783 RAC: 0 |
Recently came back to milkyway, testing my cards at two WUs each....kinda warm, 85 to 90C. My rig: AMD FX-8350 at 4100~4300mhz 16gigs of G-Skill Ripjaws RAM at 1600mhz 2-Powercolor R9 280Xs 1030/1500mhz Gigabyte GA970-UD3 MOBO Samsung EVO 850 SSD EVGA Supernova P2 1200W PSU |
Send message Joined: 11 Feb 11 Posts: 57 Credit: 69,475,644 RAC: 0 |
my r9 280x gets really hot too to me it seems as if they have very poor cooling for this series. fan just isnt going fast enough. problem isnt the core temp at my end, but the VRAM temperature seeing 90degrees+ as well -> i need more fans |
Send message Joined: 8 May 09 Posts: 3339 Credit: 524,010,781 RAC: 0 |
my r9 280x gets really hot too Do you use MSIAfterburner? If not it works to control the temps on all brands of gpu's. |
Send message Joined: 11 Feb 11 Posts: 57 Credit: 69,475,644 RAC: 0 |
i use afterburner for overclocking and gpu z and hwinfo64 for temperature monitoring |
Send message Joined: 8 May 09 Posts: 3339 Credit: 524,010,781 RAC: 0 |
i use afterburner for overclocking and gpu z and hwinfo64 for temperature monitoring There is a way to have Afterburner raise and lower the fans speeds based on gpu temps too. |
Send message Joined: 11 Feb 11 Posts: 57 Credit: 69,475,644 RAC: 0 |
yea i know that. i need 100% whenever i have milkyway running anyway so i just have 2 profiles, 1 with 100% fan speed for milkyway and the 2nd for other applications (games or the like) with 60-70% but they only go with a cap of 3000 rpm, the ones from my 6850 went up to 5800rpm at 100% |
Send message Joined: 24 Jan 11 Posts: 715 Credit: 556,325,854 RAC: 55,323 |
I run SIV (System Information Viewer) on my systems to control the temps of my video cards. I run a mix of projects that tax the thermal conditions of my video cards to different degrees. With SIV, I just set a thermal setpoint of 65 C. for all cards and SIV controls the fan speeds accordingly to maintain the setpoint. When running Einstein, the cards can throttle way down their fan speeds and when running MilkyWay, they ramp up to max to maintain the setpoint. Just set and forget. Pretty simple. |
Send message Joined: 8 May 09 Posts: 3339 Credit: 524,010,781 RAC: 0 |
I run SIV (System Information Viewer) on my systems to control the temps of my video cards. I run a mix of projects that tax the thermal conditions of my video cards to different degrees. With SIV, I just set a thermal setpoint of 65 C. for all cards and SIV controls the fan speeds accordingly to maintain the setpoint. When running Einstein, the cards can throttle way down their fan speeds and when running MilkyWay, they ramp up to max to maintain the setpoint. Just set and forget. Pretty simple. I will look at it thanks, it even sounds like it works on both AMD and Nvidia cards. I found it very informative but like the simplicity of Afterburner better. Also do NOT download it from CNET, it installed software BESIDES the software I asked to download!! |
Send message Joined: 8 May 09 Posts: 3339 Credit: 524,010,781 RAC: 0 |
I run SIV (System Information Viewer) on my systems to control the temps of my video cards. I run a mix of projects that tax the thermal conditions of my video cards to different degrees. With SIV, I just set a thermal setpoint of 65 C. for all cards and SIV controls the fan speeds accordingly to maintain the setpoint. When running Einstein, the cards can throttle way down their fan speeds and when running MilkyWay, they ramp up to max to maintain the setpoint. Just set and forget. Pretty simple. [Edit] Damned CNET ALSO added Yahoo Search to my home pages even though I told it NOT TO DO IT!!! That's the LAST time I use them to download anything!!!![/edit] |
Send message Joined: 24 Jan 11 Posts: 715 Credit: 556,325,854 RAC: 55,323 |
I always download the latest SIV beta update directly from within the program and get the latest standard release directly from the developers website. I help him out with his beta testing since he is always looking to support the latest hardware. Yes, SIV supports both ATI and Nvidia cards. Here is the direct link to the website for the program. System Information Viewer |
Send message Joined: 8 Apr 13 Posts: 89 Credit: 517,085,245 RAC: 0 |
Oh, don't dl anything from CNet, it has become one of the worst dl sites ever. SIV has tons of features, but I find it sometimes too difficult to use.. BTW, I'm the author of HWiNFO64 ;-) |
Send message Joined: 28 Nov 14 Posts: 51 Credit: 86,696,721 RAC: 0 |
Hi Folks, Just had a bit of a runaround with the latest version of BOINC[7.6.9]:-/ I was running 2 WU on one of my NV GPU's using the app_config.xml file posted earlier. However that was causing other non MW@H WU to run slower so I decided to revert to 1 WU on the gpu.. However when I removed the aforementioned app_config.xml file from the project directory BOINC kept on running 2 WU on the gpu:-( After some hunting I discovered the both the client state files had an entry of 0.5 for the GPU:-/ Had to edit both files back to 1.000000 WU on the GPU.. I don't know why BOINC put those entries in the client state files or why when the app_config.xml was removed, it didn't remove those entries in client state files, but I surmise its a bug of some sort.. But should anyone else want to use the app_config.xml file to run multiple WU's and decide to revert back to a prior configuration my guess is they will run into the same problem I had.. Regards, Regards, Cliff. -- Been there Done That, still no Damn T-Shirt |
Send message Joined: 8 May 09 Posts: 3339 Credit: 524,010,781 RAC: 0 |
Hi Folks, I think here at MW changes only take effect when you get NEW workunits, not on the existing ones. Unless you do what you did of course. |
Send message Joined: 4 Sep 12 Posts: 219 Credit: 456,474 RAC: 0 |
It does actually say at the very bottom of the Application configuration documentation If you remove app_config.xml, or one of its entries, you must reset the project in order to restore the proper values. Obviously, you wouldn't want to do that while you had any work cached. |
Send message Joined: 27 Nov 12 Posts: 8 Credit: 126,516,924 RAC: 0 |
What's wrong with rereading the configuration? Works for me. Boinc Manager > Options > Read configuration. I suppose that's the item in English, my Manager spells them in Dutch. |
Send message Joined: 4 Sep 12 Posts: 219 Credit: 456,474 RAC: 0 |
'Read configuration' works fine for replacing one value with another, or adding a value where none existed before. It doesn't work always for removing a value once it has embedded itself in the system, but in most cases you can just leave it there. It's the way they coded it. |
Send message Joined: 28 Nov 14 Posts: 51 Credit: 86,696,721 RAC: 0 |
Hi Mikey, yeah, makes sense, but I just wish someone in the programming dept at BOINC would makes those sorts of changes immediate.. Save a lot of faffing about to sort out a problem, particularly if you have a large[ish] cache of WU. Regards Regards, Cliff. -- Been there Done That, still no Damn T-Shirt |
©2024 Astroinformatics Group