Index  | Recent Threads  | Unanswered Threads  | Who's Active  | Guidelines  | Search
 

Quick Go »
No member browsing this thread
Thread Status: Active
Total posts in this thread: 162
Posts: 162   Pages: 17   [ Previous Page | 1 2 3 4 5 6 7 8 9 10 | Next Page ]
[ Jump to Last Post ]
Post new Thread
Author
Previous Thread This topic has been viewed 197489 times and has 161 replies Next Thread
uplinger
Former World Community Grid Tech
Joined: May 23, 2005
Post Count: 3952
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: OpenPandemics GPU Beta Test - Feb 27 2021 [ Issues Thread ]

I'm seeing 100% error rate on intel gpu at the moment. I will be investigating why that is in the scheduler. There is a second issue where some workunits as noted above by threadripper failed 100% on a specific ligand.

For now, the beta is paused.

Thanks,
-Uplinger
[Feb 27, 2021 8:13:29 PM]   Link   Report threatening or abusive post: please login first  Go to top 
koschi
Cruncher
Joined: Dec 16, 2007
Post Count: 5
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: OpenPandemics GPU Beta Test - Feb 27 2021 [ Issues Thread ]

Hi uplinger,

is each step in the GPU WUs comparable to an "old" OPN CPU WU or are these steps smaller?

thanks!
[Feb 27, 2021 8:18:55 PM]   Link   Report threatening or abusive post: please login first  Go to top 
DrMason
Senior Cruncher
Joined: Mar 16, 2007
Post Count: 153
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: OpenPandemics GPU Beta Test - Feb 27 2021 [ Issues Thread ]

Hey uplinger,

I got 4 on my linux mint tricia rig with a GT 1030, and 5 on my windows 10 rig with an RTX 3080. The 5 on the 3080 all completed fine (one of the workunits had 199 jobs, and it completed in about 15 minutes! That's crazy!).

All four on my rig with the GT 1030 errored out. All errors were of the same error: "Error: boinc_get_opencl_ids() failed with error -1". What was interesting to me is that while my rig and maybe another rig or two errored on these units, some of the cohorts are pending validation. I double checked and I have the OpenCL drivers and the CUDA drivers for boinc installed. The previous run of BETA units worked fine on the rig.

Workunits with errors: 21000/40_1 ; 21000/28_2 ; 21004/34_1 ; 21008/63_0
Hope this helps.
----------------------------------------

[Feb 27, 2021 8:28:16 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Grumpy Swede
Master Cruncher
Svíþjóð
Joined: Apr 10, 2020
Post Count: 1886
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: OpenPandemics GPU Beta Test - Feb 27 2021 [ Issues Thread ]

Error out on iGPU (got 4), all 4 "error 10" (ERROR: Failed to find device id). Previous Beta version worked fine on my iGPU HD4600.
But I also got a few for my Nvidia GTX980 (got 10) those crunched OK. I wasn't at the computer when these new Betas were sent out though.

Strangely enough, now even if I Select "YES" for Nvidia (which it has been since last Beta version) on the device profile for that computer, it does no longer ask for Nvidia at all. It asks for Intel though, if I set that to "YES (also have been set to "YES" since last Beta version). It worked before as it should. If setting "YES", to both NVIDIA and Intel GPU, it did previously ask for both.

Something must have happened with that too, since it worked up to 2021-02-27 20:23:49 GMT+1 , where the request says "Requesting new tasks for CPU and NVIDIA GPU and Intel GPU" The next request at 2021-02-27 20:29:04 GMT+1 , and all requests thereafter looks like this "Requesting new tasks for CPU and Intel GPU" . No more requests for Nvidia....

But, it was all fun while it lasted smile
Have a nice Saturday Uplinger. Don't work too much.
----------------------------------------

----------------------------------------
[Edit 9 times, last edit by Grumpy Swede at Feb 27, 2021 9:40:44 PM]
[Feb 27, 2021 8:53:17 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Speedy51
Veteran Cruncher
New Zealand
Joined: Nov 4, 2005
Post Count: 1220
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: OpenPandemics GPU Beta Test - Feb 27 2021 [ Issues Thread ]

I have had 4 results run on my RTX 2070 Windows 10 Pro run times of around 3 and 6 minutes and they are all in PV
----------------------------------------

[Feb 27, 2021 9:31:52 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Dayle Diamond
Senior Cruncher
Joined: Jan 31, 2013
Post Count: 440
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: OpenPandemics GPU Beta Test - Feb 27 2021 [ Issues Thread ]

Eight tasks completed successfully with GTX 1070 Ti and the latest drivers.

So far it seems NVIDIA cards are having an easier time with the beta.
[Feb 27, 2021 10:12:33 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Bryn Mawr
Senior Cruncher
Joined: Dec 26, 2018
Post Count: 310
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: OpenPandemics GPU Beta Test - Feb 27 2021 [ Issues Thread ]

Hey uplinger,

I got 4 on my linux mint tricia rig with a GT 1030, and 5 on my windows 10 rig with an RTX 3080. The 5 on the 3080 all completed fine (one of the workunits had 199 jobs, and it completed in about 15 minutes! That's crazy!).

All four on my rig with the GT 1030 errored out. All errors were of the same error: "Error: boinc_get_opencl_ids() failed with error -1". What was interesting to me is that while my rig and maybe another rig or two errored on these units, some of the cohorts are pending validation. I double checked and I have the OpenCL drivers and the CUDA drivers for boinc installed. The previous run of BETA units worked fine on the rig.

Workunits with errors: 21000/40_1 ; 21000/28_2 ; 21004/34_1 ; 21008/63_0
Hope this helps.



Exactly the same situation for me on my GT710, error Error: boinc_get_opencl_ids() failed with error -1 whilst the wingman is pv and the last beta all worked.
[Feb 27, 2021 11:09:31 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Grumpy Swede
Master Cruncher
Svíþjóð
Joined: Apr 10, 2020
Post Count: 1886
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: OpenPandemics GPU Beta Test - Feb 27 2021 [ Issues Thread ]

I'll reset the project (after my last CPU WU is done) in preparation for the next Beta run. Just in case there's something abnormal left, after this "failed" test session.
----------------------------------------

----------------------------------------
[Edit 1 times, last edit by Grumpy Swede at Feb 27, 2021 11:29:17 PM]
[Feb 27, 2021 11:17:41 PM]   Link   Report threatening or abusive post: please login first  Go to top 
widdershins
Veteran Cruncher
Scotland
Joined: Apr 30, 2007
Post Count: 673
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: OpenPandemics GPU Beta Test - Feb 27 2021 [ Issues Thread ]

I've had another couple of units which have thrown errors of the type "- Unhandled Exception Record -
Reason: Access Violation (0xc0000005) at address 0x000000013FC7DB30 read attempt to address 0x00000000"

Card is a GTX 730 on W7x64

Below is the reported capabilities for OpenCL, and I've run a test suite which suggests OpenCL is working. Any ideas why my Beta Units are failing?

===================================[ OpenCL Capabilities ]
- Num OpenCL platforms: 1
- CL_PLATFORM_NAME: NVIDIA CUDA
- CL_PLATFORM_VENDOR: NVIDIA Corporation
- CL_PLATFORM_VERSION: OpenCL 1.2 CUDA 10.0.132
- CL_PLATFORM_PROFILE: FULL_PROFILE
- Num devices: 1

- CL_DEVICE_NAME: GeForce GT 730
- CL_DEVICE_VENDOR: NVIDIA Corporation
- CL_DRIVER_VERSION: 416.34
- CL_DEVICE_PROFILE: FULL_PROFILE
- CL_DEVICE_VERSION: OpenCL 1.2 CUDA
- CL_DEVICE_TYPE: GPU
- CL_DEVICE_VENDOR_ID: 0x10DE
- CL_DEVICE_MAX_COMPUTE_UNITS: 2
- CL_DEVICE_MAX_CLOCK_FREQUENCY: 901MHz
- CL_NV_DEVICE_COMPUTE_CAPABILITY_MAJOR: 3
- CL_NV_DEVICE_COMPUTE_CAPABILITY_MINOR: 5
- CL_NV_DEVICE_REGISTERS_PER_BLOCK: 65536
- CL_NV_DEVICE_WARP_SIZE: 32
- CL_NV_DEVICE_GPU_OVERLAP: 1
- CL_NV_DEVICE_KERNEL_EXEC_TIMEOUT: 1
- CL_NV_DEVICE_INTEGRATED_MEMORY: 0
- CL_DEVICE_ADDRESS_BITS: 32
- CL_DEVICE_MAX_MEM_ALLOC_SIZE: 262144KB
- CL_DEVICE_GLOBAL_MEM_SIZE: 1024MB
- CL_DEVICE_MAX_PARAMETER_SIZE: 4352
- CL_DEVICE_GLOBAL_MEM_CACHELINE_SIZE: 128 Bytes
- CL_DEVICE_GLOBAL_MEM_CACHE_SIZE: 32KB
- CL_DEVICE_ERROR_CORRECTION_SUPPORT: NO
- CL_DEVICE_LOCAL_MEM_TYPE: Local (scratchpad)
- CL_DEVICE_LOCAL_MEM_SIZE: 48KB
- CL_DEVICE_MAX_CONSTANT_BUFFER_SIZE: 64KB
- CL_DEVICE_MAX_WORK_ITEM_DIMENSIONS: 3
- CL_DEVICE_MAX_WORK_ITEM_SIZES: [1024 ; 1024 ; 64]
- CL_DEVICE_MAX_WORK_GROUP_SIZE: 1024
- CL_EXEC_NATIVE_KERNEL: 21340932
- CL_DEVICE_IMAGE_SUPPORT: YES
- CL_DEVICE_MAX_READ_IMAGE_ARGS: 256
- CL_DEVICE_MAX_WRITE_IMAGE_ARGS: 16
- CL_DEVICE_IMAGE2D_MAX_WIDTH: 16384
- CL_DEVICE_IMAGE2D_MAX_HEIGHT: 16384
- CL_DEVICE_IMAGE3D_MAX_WIDTH: 4096
- CL_DEVICE_IMAGE3D_MAX_HEIGHT: 4096
- CL_DEVICE_IMAGE3D_MAX_DEPTH: 4096
- CL_DEVICE_MAX_SAMPLERS: 32
- CL_DEVICE_PREFERRED_VECTOR_WIDTH_CHAR: 1
- CL_DEVICE_PREFERRED_VECTOR_WIDTH_SHORT: 1
- CL_DEVICE_PREFERRED_VECTOR_WIDTH_INT: 1
- CL_DEVICE_PREFERRED_VECTOR_WIDTH_LONG: 1
- CL_DEVICE_PREFERRED_VECTOR_WIDTH_FLOAT: 1
- CL_DEVICE_PREFERRED_VECTOR_WIDTH_DOUBLE: 1
- CL_DEVICE_EXTENSIONS: 16
- Extensions:
- cl_khr_global_int32_base_atomics
- cl_khr_global_int32_extended_atomics
- cl_khr_local_int32_base_atomics
- cl_khr_local_int32_extended_atomics
- cl_khr_fp64
- cl_khr_byte_addressable_store
- cl_khr_icd
- cl_khr_gl_sharing
- cl_nv_compiler_options
- cl_nv_device_attribute_query
- cl_nv_pragma_unroll
- cl_nv_d3d9_sharing
- cl_nv_d3d10_sharing
- cl_khr_d3d10_sharing
- cl_nv_d3d11_sharing
- cl_nv_copy_opts

Edited to add WU ID's
BETA_OPN1_0020080_00290 (first beta run)
BETA_ OPNG_ 0021002_ 00020_ 1 (second beta run)
----------------------------------------
[Edit 1 times, last edit by widdershins at Feb 27, 2021 11:55:39 PM]
[Feb 27, 2021 11:52:08 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Jake1402
Senior Cruncher
USA
Joined: Dec 30, 2005
Post Count: 180
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: OpenPandemics GPU Beta Test - Feb 27 2021 [ Issues Thread ]

I received 11 on my 2 Windows 10 machines with AMD Radeon R9 270X...all are in the PV cage. I did not receive any on my 2 Linux Mint 20.1 machines with AMD RX 560's.
----------------------------------------
Join the Chicago-IL-USA team!
2 AMD FX 8320/AMD R9 270X/Win 10
2 AMD FX 8320/AMD RX 560/Linux Mint 20.3
Intel Pentium G240/Win 10
----------------------------------------
[Edit 2 times, last edit by Jake1402 at Feb 28, 2021 6:50:55 PM]
[Feb 28, 2021 12:18:57 AM]   Link   Report threatening or abusive post: please login first  Go to top 
Posts: 162   Pages: 17   [ Previous Page | 1 2 3 4 5 6 7 8 9 10 | Next Page ]
[ Jump to Last Post ]
Post new Thread