EM searches, BRP Raidiopulsar and FGRP Gamma-Ray Pulsar

Richard Haselgrove
Richard Haselgrove
Joined: 10 Dec 05
Posts: 2143
Credit: 2949640451
RAC: 691283

Version 0.02 of MeerKAT gets

Version 0.02 of MeerKAT gets slightly further - see host 1226365

stderr ends with

[17:29:35][4288][ERROR] Error launching CUDA TSM kernel (error: 1)
[17:29:35][4288][ERROR] Demodulation failed (error: 1015)!
17:29:35 (4288): called boinc_finish(1015)
Ian&Steve C.
Ian&Steve C.
Joined: 19 Jan 20
Posts: 3931
Credit: 46168042642
RAC: 63840826

same error on a

same error on a 550Ti

https://einsteinathome.org/host/12900237/tasks/6/0

 

Quote:

<core_client_version>7.20.2</core_client_version>
<![CDATA[
<message>
The registry is corrupted. The structure of one of the files containing registry data is corrupted, or the system's memory image of the file is corrupted, or the file could not be recovered because the alternate copy or log was absent or corrupted.
 (0x3f7) - exit code 1015 (0x3f7)</message>
<stderr_txt>
Activated exception handling...
[12:30:45][3324][INFO ] Starting data processing...
[12:30:45][3324][INFO ] CUDA global memory status (initial GPU state, including context):
------> Used in total: 226 MB (799 MB free / 1025 MB total) -> Used by this application (assuming a single GPU task): 0 MB
[12:30:45][3324][INFO ] Using CUDA device #0 "GeForce GTX 550 Ti" (192 CUDA cores / 691.20 GFLOPS)
[12:30:45][3324][INFO ] Version of installed CUDA driver: 9010
[12:30:45][3324][INFO ] Version of CUDA driver API used: 3020
[12:30:45][3324][INFO ] Number of generated templates to be used: 1200
[12:30:45][3324][INFO ] Checkpoint file unavailable: M22_1_cfbf00002_segment_2_dms_200_15.cpt (No error).
------> Starting from scratch...
[12:30:45][3324][INFO ] Header contents:
------> Original WAPP file: /atlas/data/TRAPUM_GC/M22/epoch1/M22/2021-02-14-06:43:10/1284/15min_segments/dedispersed_files/cfbf00002/M22_1_cfbf00002_segment_2_dms_200_DM85.75
------> Sample time in microseconds: 76.5607
------> Observation time in seconds: 1284.4762
------> Time stamp (MJD): 59259.287751050571
------> Number of samples/record: 0
------> Center freq in MHz: 857.5673828
------> Channel band in MHz: 3.34375
------> Number of channels/record: 256
------> Nifs: 1
------> RA (J2000): 183623.940001
------> DEC (J2000): -235417.1
------> Galactic l: 0
------> Galactic b: 0
------> Name: M22
------> Lagformat: 0
------> Sum: 1
------> Level: 3
------> AZ at start: 0
------> ZA at start: 0
------> AST at start: 0
------> LST at start: 0
------> Project ID: --
------> Observers: --
------> File size (bytes): 0
------> Data size (bytes): 0
------> Number of samples: 16777216
------> Trial dispersion measure: 85.75 cm^-3 pc
------> Scale factor: 0.933703
[12:30:45][3324][INFO ] Seed for random number generator is -1022884736.
[12:30:52][3324][INFO ] Derived global search parameters:
------> f_A probability = 0.04
------> single bin prob(P_noise > P_thr) = 1.62212e-009
------> thr1 = 20.2395
------> thr2 = 23.4356
------> thr4 = 28.6162
------> thr8 = 37.2366
------> thr16 = 51.9152
[12:30:52][3324][INFO ] CUDA global memory status (GPU setup complete):
------> Used in total: 869 MB (156 MB free / 1025 MB total) -> Used by this application (assuming a single GPU task): 643 MB
[12:30:52][3324][ERROR] Error launching CUDA TSM kernel (error: 1)
[12:30:52][3324][ERROR] Demodulation failed (error: 1015)!
12:30:52 (3324): called boinc_finish(1015)

_________________________________________________________________________

Mr P Hucker
Mr P Hucker
Joined: 12 Aug 06
Posts: 838
Credit: 519123312
RAC: 14007

Has Windows AMD GPU BRP7

Has Windows AMD GPU BRP7 temporarily stopped?  I see 4000 in the queue, are they Cuda/Linux only?  I was getting a nice supply of Windows AMD a few hours ago.  I turned off gamma, because I need beta on to get BRP7, but beta gamma always fail on my cards and always have (nobody ever fixed that?)

If this page takes an hour to load, reduce posts per page to 20 in your settings, then the tinpot 486 Einstein uses can handle it.

Richard Haselgrove
Richard Haselgrove
Joined: 10 Dec 05
Posts: 2143
Credit: 2949640451
RAC: 691283

Peter Hucker of the Scottish

Peter Hucker of the Scottish Boinc Team wrote:

Has Windows AMD GPU BRP7 temporarily stopped?  I see 4000 in the queue, are they Cuda/Linux only?  I was getting a nice supply of Windows AMD a few hours ago.  I turned off gamma, because I need beta on to get BRP7, but beta gamma always fail on my cards and always have (nobody ever fixed that?)

The Einstein project has a policy of "only one unreliable application result per workunit". ('unreliable' including self-compiled applications and project Beta applications).

It's likely that the unsent tasks are waiting for a 'reliable' app to be declared, so that comparison tasks can be issued and validation completed.

Ian&Steve C.
Ian&Steve C.
Joined: 19 Jan 20
Posts: 3931
Credit: 46168042642
RAC: 63840826

or just check the

or just check the applications page and see that the project has pulled the Windows-AMD and Linux-Nvidia apps. so they wont be sent. 

either for updates, or they determined that those apps are good enough and don't require further testing.

 

https://einsteinathome.org/apps.php

_________________________________________________________________________

Guðni Már Gilbert
Guðni Már Gilbert
Joined: 30 Jun 20
Posts: 12
Credit: 437172656
RAC: 332968

I am getting an error as

I am getting an error as well:

 

cecht
cecht
Joined: 7 Mar 18
Posts: 1530
Credit: 2880649158
RAC: 2065474

I had to abort the second

I had to abort the second attempt at running an M22 task on my RX 5600 XT host (https://einsteinathome.org/host/12879640), because it was shaping up like yesterday's aborted attempt that hadn't finished after 19 hr.

After an hour, today's task was 50% complete. The GPU load, however, was at a constant 99% with constant 0% memory load, which seems unusual. My RX 570 cards have completed many M22 tasks with no problems, though still waiting on validation.

Is anyone else having issues with a Navi 10 card?

Ideas are not fixed, nor should they be; we live in model-dependent reality.

archae86
archae86
Joined: 6 Dec 05
Posts: 3157
Credit: 7211794931
RAC: 946970

cecht wrote:Is anyone else

cecht wrote:
Is anyone else having issues with a Navi 10 card?

Yes to Navi.  Not trouble, and not Linux, but I have seemingly successful completion and returns awaiting validation from three systems, with

RX 5700

RX 6800

RX 6800XT

Despite heavy throttling (for reduced power consumption and room heating) the completion times mostly range from 30 to 47 seconds.

 

Ian&Steve C.
Ian&Steve C.
Joined: 19 Jan 20
Posts: 3931
Credit: 46168042642
RAC: 63840826

cecht wrote: I had to abort

cecht wrote:

I had to abort the second attempt at running an M22 task on my RX 5600 XT host (https://einsteinathome.org/host/12879640), because it was shaping up like yesterday's aborted attempt that hadn't finished after 19 hr.

After an hour, today's task was 50% complete. The GPU load, however, was at a constant 99% with constant 0% memory load, which seems unusual. My RX 570 cards have completed many M22 tasks with no problems, though still waiting on validation.

Is anyone else having issues with a Navi 10 card?

I wouldn’t be surprised if it was some kind of driver issue. Navi 10 always kind of got the short end of the stick in that regard. Also possible to be some compilation flag or argument that needs to be added. 

_________________________________________________________________________

Mr P Hucker
Mr P Hucker
Joined: 12 Aug 06
Posts: 838
Credit: 519123312
RAC: 14007

archae86 wrote: Despite

archae86 wrote:

Despite heavy throttling (for reduced power consumption and room heating) the completion times mostly range from 30 to 47 seconds.

Can't you just run less cards?

Or do what I do and open the windows.

If this page takes an hour to load, reduce posts per page to 20 in your settings, then the tinpot 486 Einstein uses can handle it.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.