Posts by Tom*

21) Message boards : Number crunching : What is the cause of these 'validate errors' (Message 63526) Posted 5 May 2015 by Tom* Post: Here are mine http://milkyway.cs.rpi.edu/milkyway/results.php?hostid=573186&offset=0&show_names=0&state=5&appid= 18 Validate errors so far just today on one system
22) Message boards : News : New Modfit Runs (Message 62848) Posted 15 Dec 2014 by Tom* Post: Cliff All your failed tasks are also failed by the other two users doing them. In other Words Bad Tasks. "May Have Bug"
23) Message boards : News : New Separation Modfit Version 1.36 (Message 62588) Posted 18 Oct 2014 by Tom* Post: Several people running ati? were experiencing 6% validate errors on modfit runs. I just started with my HD7950 with same drivers I have been using for several years. These drivers produced no invalids back in April for a team challenge. SETI recently noticed that an API bug in BOINC truncates the STDERR sent to the server. ALL of my Invalids have this example Stderr output <core_client_version>7.2.42</core_client_version> <![CDATA[ <stderr_txt> </stderr_txt> ]]> http://milkyway.cs.rpi.edu/milkyway/results.php?hostid=573186&offset=0&show_names=0&state=5&appid= Windows 7 64 bit boinc 64 bit. FX-8350 My i5-3570 with GTX660 does not show any invalids Thanks Will test my A8-5800 in a couple of days it also has an HD7950
24) Message boards : Number crunching : New Benchmark Thread - times wanted for any hardware, CPU or GPU, old or new! (Message 61806) Posted 30 May 2014 by Tom* Post: With the HD7950 clock set to 900 no variation occurs and the timings are very similar 33.05 seconds average .01 sec variance with No CPU jobs. Its not heat that varies the clock when set to 1112 oc via MSI afterburner as Einstein and SETI do not vary.
25) Message boards : Number crunching : New Benchmark Thread - times wanted for any hardware, CPU or GPU, old or new! (Message 61796) Posted 28 May 2014 by Tom* Post: HD7950 oc at 1100 but drops to 900 during run, jumps between 1100 and 900. AMD A8 at 3.4 Default. Win 7 64 bit ave time is 33.06 seconds but when 4 CPU's are running T4T times go up to 35 and 36 seconds. http://milkyway.cs.rpi.edu/milkyway/results.php?hostid=509400&offset=0&show_names=0&state=0&appid=3 Catalyst old 13.2 beta 7 moded with OpenCL 12.4 due to bad OpenCL compiler released on 13.1 which was latest when I installed the HD7950. They must have fixed the bad compiler but it was still bad in early 13.x versions.
26) Message boards : Number crunching : Server Error: Feeder not running ?? in response to update command (Message 61035) Posted 10 Feb 2014 by Tom* Post: Haven't been able to report completed tasks for awhile. Is there a problem? Thanks Bill
27) Message boards : Number crunching : All Milkyway@Home 1.02 tasks ending in computation error on HD6950. (Message 60441) Posted 20 Nov 2013 by Tom* Post: I PM'd ascension that all his failures are on device 0 on platform 0 which is a turks, device 1 on platform 0 is a cape verde. according to wikipedia turks have no double precision, while cape verde does. so I wonder how his turks worked when by itself? Unless BOINC or Einstein is reporting an incorrect device? ex of failing task Found 2 CL devices Device 'Turks' (Advanced Micro Devices, Inc.:0x1002) (CL_DEVICE_TYPE_GPU) Driver version: 1268.1 (VM) Version: OpenCL 1.2 AMD-APP (1268.1) Compute capability: 0.0 Max compute units: 6 Clock frequency: 800 Mhz Global mem size: 1073741824 Local mem size: 32768 Max const buf size: 65536 Double extension: (none) Device doesn't support double precision Failed to calculate likelihood Is there anyway to just disable device 0 on platform 0? Leaving device 1 to process MilkyWay?
28) Message boards : Number crunching : Updated GPU Requirements (Currently not supporting GPU tasks) (Message 60398) Posted 15 Nov 2013 by Tom* Post: The milkyway application is complaining about not having double precision. I run a HD7950 on Windows and it surely has that capability. although I also get a compute capability of 0.0 (Likely just an Nvidia thingy) Does anyone running Darwin have a driver that works?
29) Message boards : Number crunching : AMD GPU Computation errors (Message 60037) Posted 27 Sep 2013 by Tom* Post: I second Tex's 12.4 OpenCL moded with any catalyst version I have used Darkriders OpenCL 12.4 moded with Catalyst 13.2 Beta 7 from Darkrider since Feb 2013 on a HD7950 for SETI Einstein and Milkyway OpenCL 12.4 seems to work for all my projects. If you are having OpenCL crashes I would definately give it a try, I looked at some crashes recently (all my invalids seem to be due to other peoples crashes) and they do seem to be within the OpenCL DLL's but be absolutely sure you use the recommended Catalyst removal programs if you are downgrading any component.
30) Message boards : Number crunching : 7950 GPU core low frequency while crunching WUs (Message 58017) Posted 25 Apr 2013 by Tom* Post: I am running Cat 13.2 beta 7 on my HD7950 modded with 12.4 OpenCL. I also see the Core freq drop to 900 from 1152 when running MilkyWay, its a solid 900 during the run, however I also see the Core Freq drop during SETI AP runs for split seconds before going back to 1152. Since my MilkyWay runs are always 40 or 51 seconds I am not worried about the Freqing Core clock:-) GPU load for MilkyWay is 99% while for a single run of SETI AP it bounces between 65 and 45 % when GPU Usage is over 71% thats when the core Freq drops to 900. So issue is likely in the Catalyst drivers
31) Message boards : Number crunching : OpenCL on Intel (Message 57888) Posted 11 Apr 2013 by Tom* Post: I think you are correct, No FP64 on the GPU in OpenCL as of yet. 2013 SDK Double Precision Floating Point Supported Devices IntelÂ® CPU Yes IntelÂ® HD Graphics (GPU) No The CPU device implements cl_khr_fp64 to support for double precision floating-point. Double precision floating-point is a requirement for some scientific computing algorithms/applications. It is just not clear if they do all of the cl_khr_fp64 on the CPU or if they use the FP32 portion of the GPU in conjunction with the CPU to emulate FP64 not sure parallelization to any degree works in either case.
32) Message boards : Number crunching : OpenCL on Intel (Message 57885) Posted 11 Apr 2013 by Tom* Post: Does the following statement from Tom's Hardware mean no Milkyway support?? or is the OpenCL FP32 only a software driver/API thingy where FP64 might come later? "HD Graphics 4000, on the other hand, supports FP32/FP64 under DirectCompute and FP32 in OpenCL. Intel currently lacks Khronos certification for ARB_gpu_shader_fp64, so itâ€™s not enabled. " EDIT - Just had to read a little further API from Intel Needed "itâ€™ll be interesting to see how HD Graphics 4000 compares to derivative discrete GPUs from Nvidia (particularly if Intel adds its own OpenCL FP64 extension). " Link http://www.tomshardware.com/reviews/ivy-bridge-benchmark-core-i7-3770k,3181-6.html
33) Message boards : Number crunching : Updating GPU drivers (Message 57809) Posted 5 Apr 2013 by Tom* Post: Sorry but I do not think GT330M has double precision, Milkyway GPU requires double precision. NVIDIA says the GT330M has a compute capability of 1.2 and Wikipedia says Double precision floats (CUDA compute capability 1.3 and above) I also saw an entry in a gpugrid forum saying their gt330m does not have double precision. No driver update will give you this, however their are plenty of BOINC projects that will work fine . Bill
34) Message boards : News : New N-Body Release (Message 57388) Posted 28 Feb 2013 by Tom* Post: The validator will decide that, when you retun your result. Does the Validator accept donations?:-) EDIT - I shouldn't have said anything The percentage complete jumped back to 18.210% at Elapsed 44:30:22:-( Deadline is 3/9/2013
35) Message boards : News : New N-Body Release (Message 57385) Posted 28 Feb 2013 by Tom* Post: I also have a 1.06 N-Body workunit that has run for 43.5 hours and is 37.8% complete. Since there are 5 other failed tasks on this workunit, should I let it run hoping to get another user to complete this workunit? Someone mentioned that they do not need 1.06 Workunits anymore?? If so is the proper procedure here to abort 1.06 WU in favor of 1.07? Thanks
36) Message boards : Number crunching : ATI generate some invalids and nVidia not (Message 57334) Posted 24 Feb 2013 by Tom* Post: No Invalids since that first one The errors come in groups of three possibly something is resetting your 79xx ATI Since I run Nvidia I cannot help with driver versions and which are best or worst for MilkyWay someone else who has been thru catalyst versions should jump in now . Are you running more than one GPU job at a time? Good luck it seems to be running good otherwise
37) Message boards : Number crunching : ATI generate some invalids and nVidia not (Message 57332) Posted 23 Feb 2013 by Tom* Post: Homeboy Homepage --> Returning participants â€¢Your account - view stats, modify preferences Tasks
38) Message boards : Number crunching : ATI generate some invalids and nVidia not (Message 57329) Posted 23 Feb 2013 by Tom* Post: State: All (131) \| In progress (25) \| Pending (64) \| Valid (37) \| Invalid (1) \| Error (4) Looks good to me so far All new and returning users for this project need to proove themselves At least thats how it works for me. While the pending all show validation inconclusive that is 100% normal for this project. After MilkyWay validates some more (not sure how many more) and you still keep the real Invalids down to 0 or 1 Milkyway is one of the few projects (that I participate in) that have a minimum quorum of 1 (not sure how they do this)? Have patience and watch your real INVALIDS as far as Einstein If you check your Programs and Features and you have Intels OpenCL installed try as a test to remove it as it sometimes interferes with OpenCL on other Devices. 7.0.52 is supposed to detect this but not sure if it fixes this.

Previous 20