SMP client error {Resolved}

Discussion in 'Folding@Home - Join Team Guru3D !' started by ariskar, Oct 21, 2010.

  1. ariskar

    ariskar Master Guru

    Messages:
    206
    Likes Received:
    6
    GPU:
    Nvidia RTX 2080
    I am getting this error, my PC is OC at 3.33 and seems stable (Prime and everest tests) and I keep getting this error for different WU's

    [​IMG]

    I am using the -verbosity 9 -smp 7 parameters same error also on -smp
    Any idea? :bang:

    Thanks in advance!:pc1:
     
    Last edited: Oct 22, 2010
  2. PantherX

    PantherX Folding@Home Team Leader Staff Member

    Messages:
    1,380
    Likes Received:
    0
    GPU:
    Nvidia GTX 1080 Ti
    Please note that F@h is more stressful then Prime95 (not sure of Everest). The only two applications that I know which are useful for stress testing are:
    1) StressCPU 2.0 -> This is a special software that is written to stress the CPU.
    2) IntelBurnTest 2.4 -> When set to 10 iterations with ~2GB RAM is okay. (Maximum Settings are preferred if you really want to push your system to its limits)

    Now I would suggest the following:
    1) Return all settings to stock including the CPU Speed, RAM Timings, etc. and start from scratch.
    2) Loosen (increase) the RAM Timings as an unstable RAM OC can cause issues. You will have to tweak the RAM Timings to get it F@H Stable.
     
  3. Foldinator

    Foldinator New Member

    Messages:
    7
    Likes Received:
    0
    GPU:
    Build in GPU vegas
    IMHO there is no better stress tester theN FAH !
    REASSON WHY I ALWAYS STRESS TEST WITH fah I KNOW ITS A BIT MORE WORK (BABY SITTING) BUT AT THE ENDS ITS WORTH IT...

    Sorry for the crapslock.........
     
  4. ariskar

    ariskar Master Guru

    Messages:
    206
    Likes Received:
    6
    GPU:
    Nvidia RTX 2080
    passed intel burn test, 10 interations, max setting. max temp 70 in core/60 on cpu

    I get same error (bit more rare) @ 3.2 ; 6-7-6-18 1604Mhz ram (XMP for g.skill).

    It passes every burn/stress test anyway
     

  5. Ghost15

    Ghost15 Active Member

    Messages:
    77
    Likes Received:
    0
    GPU:
    EVGA 465 SC 1GB
    ariskar, What are the settings you are using for FAH?

    Some combinations of -smp x cause failures on some non-bigadv workunits
     
  6. PantherX

    PantherX Folding@Home Team Leader Staff Member

    Messages:
    1,380
    Likes Received:
    0
    GPU:
    Nvidia GTX 1080 Ti
    Correct, hence I first suggested the StressCPU as it is based on the Gromacs code used in F@H's FahCores that process work units.
    I prefer not using live WUs, rather I save a copy and use that for stress testing.

    I would suggest that you start with the following RAM settings:
    8-8-8-24 @ 1333 or closer then work your way up from there. It could be possible that you have to tweak the voltage provided to your RAM to get it 100% Stable. The reason why I am stressing on the RAM is because the CoreStatus = C0000005 (-1073741819) is usually assoiated with RAM issues (FahWiki).

    It has already been stated that -smp 7 and -smp has been used. that means that values 7 and 8 have been checked and it didn't make any difference.
     
  7. ariskar

    ariskar Master Guru

    Messages:
    206
    Likes Received:
    6
    GPU:
    Nvidia RTX 2080
    I just competed a A3 WU with 2,000,000 steps with no issues on the same settings while it failed a couple of times before.

    6-7-6-18@1600 are the "stock" values of my Ram modules as recognised by XMP profile as well. They work at stock 1.65~1.66V (auto). They have an additional ram fan, as well as an additional 40mm fan is on the X58 chipset (max 64C in OC from 75C in stock freq.). Therefore I see no heat or stability issues on my PC and that's weird to be caused by Ram or controller instability. So I can see no reason runing the ram under-spec'ed. Anyway I will check also all running processes to see more in detail if there is any conflict.

    Anyway it seems stable for the time being with smp -7 (no -bigadv) @3.2 -> 6-7-6-18@1600.

    Thanks to all of you
     
  8. PantherX

    PantherX Folding@Home Team Leader Staff Member

    Messages:
    1,380
    Likes Received:
    0
    GPU:
    Nvidia GTX 1080 Ti
    Well, sometimes, there might be some bad WUs. If that happens, you have to report them here. However, the bottom line is that the majority of WUs that you were assigned, should fold normally on the system. If not, then it could be:

    A) A bug that is uncovered on your specific hardware configuration. You will have to provide additional details on the Official Forum for further analysis.

    B) "Unstable" OC. In this case, return everything to stock or lower and try again.

    C) Faulty/unsupported/failing Hardware.
     
  9. ariskar

    ariskar Master Guru

    Messages:
    206
    Likes Received:
    6
    GPU:
    Nvidia RTX 2080
    done second 2mil steps WU, all seems ok around 8k ppd A3 6701 project @ 3.2Gzh/6-7-6-18@1600mhz

    I think it should be related to a software or windows issue as the error appeared only when I was using the PC (internet browsing,movie, etc.)
     
    Last edited: Oct 22, 2010
  10. PantherX

    PantherX Folding@Home Team Leader Staff Member

    Messages:
    1,380
    Likes Received:
    0
    GPU:
    Nvidia GTX 1080 Ti
    Well, in that case, I would suggest that you start testing single applications and then move on to combined applications to see what application causes that specific error. AFAIK, there isn't any known application incompatibility other than to pause/exit the F@h Clients when gaming or encoding/decoding videos to get optimum performance in those CPU or GPU intensive applications.

    Could you have a faulty RAM? Because usually more running applications = more RAM used. With SMP2, less RAM is being used so that faulty RAM isn't being exposed. Have you reseated the RAM Modules, changed the positions around and have the latest BIOS Version?
     

  11. ariskar

    ariskar Master Guru

    Messages:
    206
    Likes Received:
    6
    GPU:
    Nvidia RTX 2080
    I just think I found the solution... My BIOS version is the latest (1307), I had the DRAM Voltage set to "Auto". The Ram sticks are rated @ 1.65V, while BIOS insisted to set them at 1.60V (its claimed stock setting with XMP proflie). When stressed (F@H, etc.) not enough current was supplied so 1 of the sticks was temporarily "out of game" due to low current and the system could read only 4 GB out of 6 GB of RAM. I set the DRAM Voltage to 1.66V from bios. Now it works stable on both 3.37Ghz/6-7-6-18@1604mhz and 3.2G.

    It's weird that windows didn't freeze and it passed memtest and IntelBurnTest though.

    Thanks for the post idea PantherX ;)
     
  12. Kaktus

    Kaktus Master Guru

    Messages:
    313
    Likes Received:
    10
    GPU:
    Gigabyte GT 960
    Glad you found the problem!
     
  13. ariskar

    ariskar Master Guru

    Messages:
    206
    Likes Received:
    6
    GPU:
    Nvidia RTX 2080
    thanks Kaktus and the rest of the community for the fast response!
     
  14. PantherX

    PantherX Folding@Home Team Leader Staff Member

    Messages:
    1,380
    Likes Received:
    0
    GPU:
    Nvidia GTX 1080 Ti
    Really glad that it was a voltage issue and that you resolved it quickly.
     

Share This Page