i'm back in the game and searched around and couldn't find an answer.

My old computer kicked it along time ago and haven't gotten around to getting a decent one. I now have a laptop that has 4 cores for the gpu 2 gpus I guess

anyway seti@home downloads tasks to my boinc and then after 1 or 2 seconds it says computation error on every task for it. I checked event log but all it says is the tasks are finished and uploading them and getting more.

I'm at a loss this is the first time I've faced this and have a new laptop

here some info on my laptop

2016-12-19 4:23:36 PM | | Starting BOINC client version 7.6.33 for windows_x86_64
2016-12-19 4:23:36 PM | | log flags: file_xfer, sched_ops, task
2016-12-19 4:23:36 PM | | Libraries: libcurl/7.47.1 OpenSSL/1.0.2g zlib/1.2.8
2016-12-19 4:23:36 PM | | Data directory: C:\ProgramData\BOINC
2016-12-19 4:23:36 PM | | Running under account easte
2016-12-19 4:23:36 PM | | OpenCL: AMD/ATI GPU 0: AMD Radeon R7 (driver version 2117.13 (VM), device version OpenCL 2.0 AMD-APP (2117.13), 2559MB, 2559MB available, 819 GFLOPS peak)
2016-12-19 4:23:36 PM | | OpenCL: AMD/ATI GPU 1: AMD Radeon (TM) R9 M385X (driver version 2117.13 (VM), device version OpenCL 2.0 AMD-APP (2117.13), 4096MB, 4096MB available, 1434 GFLOPS peak)
2016-12-19 4:23:36 PM | | OpenCL: AMD/ATI GPU 2: AMD Radeon R7 (driver version 2117.13 (VM), device version OpenCL 2.0 AMD-APP (2117.13), 2559MB, 2559MB available, 819 GFLOPS peak)
2016-12-19 4:23:36 PM | | OpenCL: AMD/ATI GPU 3: AMD Radeon (TM) R9 M385X (driver version 2117.13 (VM), device version OpenCL 2.0 AMD-APP (2117.13), 4096MB, 4096MB available, 1434 GFLOPS peak)
2016-12-19 4:23:37 PM | | Host name: LAPTOP-8MJARG9L
2016-12-19 4:23:37 PM | | Processor: 4 AuthenticAMD AMD FX-8800P Radeon R7, 12 Compute Cores 4C+8G [Family 21 Model 96 Stepping 1]
2016-12-19 4:23:37 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 fma cx16 sse4_1 sse4_2 movebe popcnt aes f16c rdrandsyscall nx lm avx avx2 svm sse4a osvw ibs xop skinit wdt lwp fma4 tce tbm page1gb rdtscp fs
2016-12-19 4:23:37 PM | | OS: Microsoft Windows 10: Core x64 Edition, (10.00.14393.00)
2016-12-19 4:23:37 PM | | Memory: 14.96 GB physical, 17.21 GB virtual
2016-12-19 4:23:37 PM | | Disk: 884.60 GB total, 815.46 GB free
2016-12-19 4:23:37 PM | | Local time is UTC -8 hours
2016-12-19 4:23:37 PM | | VirtualBox version: 5.0.18
2016-12-19 4:23:37 PM | | Config: use all coprocessors
2016-12-19 4:23:37 PM | ATLAS@home | URL http://atlasathome.cern.ch/; Computer ID 66878; resource share 100
2016-12-19 4:23:37 PM | Enigma@Home | URL http://www.enigmaathome.net/; Computer ID 203208; resource share 100
2016-12-19 4:23:37 PM | Fight Neglected Diseases | URL https://compute.gridrepublic.org/; Computer ID 46448; resource share 100
2016-12-19 4:23:37 PM | GPUGRID | URL http://www.gpugrid.net/; Computer ID 400118; resource share 100
2016-12-19 4:23:37 PM | SETI@home | URL http://setiathome.berkeley.edu/; Computer ID 8165867; resource share 100
2016-12-19 4:23:37 PM | SETI@home | General prefs: from SETI@home (last modified 07-May-2015 16:27:08)
2016-12-19 4:23:37 PM | SETI@home | Host location: none
2016-12-19 4:23:37 PM | SETI@home | General prefs: using your defaults
2016-12-19 4:23:37 PM | | Reading preferences override file
2016-12-19 4:23:37 PM | | Preferences:
2016-12-19 4:23:37 PM | | max memory usage when active: 7659.59MB
2016-12-19 4:23:37 PM | | max memory usage when idle: 11489.38MB
2016-12-19 4:23:37 PM | | max disk usage: 818.28GB
2016-12-19 4:23:37 PM | | (to change preferences, visit a project web site or select Preferences in the Manager)

all cpu tasks are going fine so I'm assuming this a setting or something for GPU? note I did use the use all gpu options as two of those gpu cores were ignored originally which I believe was because of some reporting bug that was fixed long before. as they have differnet amounts of memory. any suggestions? stick to just the two top memory gpus and give up on using all 4 cores?

Looking at the errors it feels as if you got sent the wrong application, one requiring CAL capability from the drivers. Your GPUs don't support CAL anymore on the hardware level, thus those drivers don't load. I think it is strange that according to the task list you got sent: Application version SETI@home v8 v8.19 (opencl_ati5_nocal) windows_intelx86 but when looking at the errors, another application was used: C:\ProgramData\BOINC\projects\setiathome.berkeley.edu\setiathome_8.19_windows_intelx86__opencl_ati5_sah.exe

The ATI5 application requires CAL.

I've asked the admins here to check that the right app is being sent and that this one is internally and externally the correct one.
An immediate workaround for you could be to use the optimized Lunatics application. see this thread for that. Then once you have some extra credit and a RAC > 1, you could report this in this developer thread.

But seeing how you ran through the full complement of tasks for the GPUs for now, and you won't be able to get more work for the next 23 hours, there's no hurry.
Next time you run a new system, test run a couple of tasks and when you see only errors, disable the GPU, go ask for help. Once errors start they hardly ever return to normalcy by themselves.

---

thanks for the info I did obtain more info as I saw an error on the failed work unit for STATUS_ACCESS_VIOLATION


also I reverted back to the option of 0 for "use all gpus" option in the config file and I figured it ignored those two for a very good reason possibly? turns out my cpu is also the radeon r7. when my cpu spikes up its temp goes up as well. and CPU-Z shows my processor as that as well. so I figured it ignored it as to keep it for cpu tasks maybe? and possibly that's why the access violation maybe occurred as it was using the cpu already for other tasks. so I'll wait for the after 23 hours and see if she purrs ahead. in mean time I'm gonna get my kali Linux machine running boinc as well and use the gpu I got in there just got to figure out how I can throttle a bit on there as that gpu goes funky after 90%. cheers and thanks again and ill post back if there is success with the current settings. thanks for the insight.

still same thing after I got more gpu from seti I got to figure out how to get sent the right one

tried lunatics but all I got was missing files in boinc for seti@home so idk if it even supports me main gpu which is a amd radeon r9 m385x. maybe a fix to these issues will come out soon enough. I'm tech savvy but no expert at this

well I'm about done. after trying many things I managed to have 2 work units rip to 20 and one 30% and then error again. and a third started now going to 17 percent and crapped out. I'm using the tthrottle with this I doubt it has an effect on this.

In the world of Nvidia GPUs I would suggest that you did a clean driver installation, using drivers from Nvidia not Microsoft as the latter are often "somewhat restricted" in the computational areas. This may be a wrong tree...

---

Bob Smith
Member of Seti PIPPS (Pluto is a Planet Protest Society)
Somewhere in the (un)known Universe?

yea that's a good idea I have managed to get it to work turns out amd-v was disabled but seem seems to work now that i've enabled it also i'll be upgrading to pro windows 10 tonight to take advantage of hyper visor as well for other projects. now I'm gonna try to enable to other gpu in this laptop then get the kali Linux upstairs going with the gpu in that computer for this as well. have two on the go. its a server doing nothing anyway minds well make some use of it

got to work out the bugs on my end its running all 4 gpus on my computer for seti and a bunch just computationed errored out but now 4 are still going one is at 20% already. got to figure out what exactly is causing it while allowing others to go.

did notice it seems to be the 2 bit guppy ones that are failing but the one 22###### one is running just fine. wonder what the difference between the two are

good news i got it fixed for some reason my system kept telling me i had the most up 2 date drivers and various places said i did as well but after i removal and then a search it detected better newer drivers which i thought strange.. but least they the right ones

2016-12-21 5:00:40 PM | | Starting BOINC client version 7.2.47 for windows_x86_64
2016-12-21 5:00:40 PM | | log flags: file_xfer, sched_ops, task
2016-12-21 5:00:40 PM | | Libraries: libcurl/7.25.0 OpenSSL/1.0.1 zlib/1.2.6
2016-12-21 5:00:40 PM | | Data directory: C:\ProgramData\BOINC
2016-12-21 5:00:40 PM | | Running under account easte
2016-12-21 5:00:40 PM | | OpenCL: AMD/ATI GPU 0: Carrizo (driver version 2236.10, device version OpenCL 1.2 AMD-APP (2236.10), 2559MB, 2559MB available, 512 GFLOPS peak)
2016-12-21 5:00:40 PM | | OpenCL: AMD/ATI GPU 1: Bonaire (driver version 2236.10, device version OpenCL 1.2 AMD-APP (2236.10), 3072MB, 3072MB available, 896 GFLOPS peak)
2016-12-21 5:00:40 PM | | OpenCL CPU: AMD FX-8800P Radeon R7, 12 Compute Cores 4C+8G (OpenCL driver vendor: Advanced Micro Devices, Inc., driver version 2236.10 (sse2,avx,fma4), device version OpenCL 1.2 AMD-APP (2236.10))
2016-12-21 5:00:40 PM | | Host name: LAPTOP-8MJARG9L
2016-12-21 5:00:40 PM | | Processor: 4 AuthenticAMD AMD FX-8800P Radeon R7, 12 Compute Cores 4C+8G [Family 21 Model 96 Stepping 1]
2016-12-21 5:00:40 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 fma cx16 sse4_1 sse4_2 movebe popcnt aes syscall nx lm svm sse4a osvw ibs xop skinit wdt lwp fma4 tce tbm page1gb rdtscp
2016-12-21 5:00:40 PM | | OS: Microsoft Windows 8.1: Core x64 Edition, (06.03.9600.00)
2016-12-21 5:00:40 PM | | Memory: 14.96 GB physical, 17.21 GB virtual
2016-12-21 5:00:40 PM | | Disk: 884.60 GB total, 771.11 GB free
2016-12-21 5:00:40 PM | | Local time is UTC -8 hours

instead of it originally saying gpu 0, 1, 2, and 3 it just has gpu 0 and gpu 1

its running seti tasked just fine now also found out my system throttled automatically when it got too hot found out the design had a filter in the intake vent so i removed it as i could barely blow through the clean filter as is. dropped my cpu temp down 25 degree C during 90% runtime on it which was great and with gpus running im getting around 60 degree C. if it goes to 66 or over my apu throttles the system down to 800 mhz to cool down. so by removing the screen even my gaming is rarely any longer affected.

Now according to BOINC gpu 0 peak GFLOPS is 512 and gpu 1 peak GFLOPS is 896.

im a little confused that isnt what the peak gflops for those specific cards.
gpu 0 details on hardware information says it should be 819.2 GFLOPs not 512.
gpu 1 details on hardware information says it should be 1.43 TFLOPS not 896.

I got this information from SiSoftware Sandra from the info it detected. and ive also looked online about the gpus and stated the same. so does it really matter what boinc detects? itll still use the full GFLOPS if it can? thanks for all the help so far

BOINC reads the value of the peak GFLOPs directly from the drivers. It is the theoretical peak performance value that the GPU manufacturer gave the GPU.

---

ah maybe its just the drivers issues as the peak gflops detected by hardware programs as well as looking up the device and info i did while making the purcahse of the laptop show way higher then that. maybe power settings somewhere cud be limiting thanks again. seti is working 100% now