Hi,
I’ve been trying to run psi4-ver1.4rc1, with BrianQC, tested with various basis sets, energy/optimize jobs. I’ve tested b3lyp, pbe0, revpbe, with d3 (e.g. b3lyp-d3) or d3bj, or d3mbj. I’ve tried various molecules from big porphyrin analogues to n-butane. All jobs would segfault.
The only dispersion correction that’ll work is if the -D2 (e.g. b3lyp-d). Is there a work-around to get at least -d3 working? (The -nl dispersion correction also errors out with no gradient available error).
To be clear, the calculations crash only when D3 is enabled?
Can you run the dftd3 binary itself without it seg.faulting?
Does it seg.fault without BrianQC turned off?
I just test again with a simple ethane molecule, using def2-tzvp, b3lyp-d3bj (previously tested with d3/d3bj actually, sto-3g/def2-tzvp, b3lyp-d3, b3lyp-d3bj, PBE0-d3, PBE0-d3bj, revPBE-d3, revPBE-d3bj). The calculation will crash with BrianQC turned on. When BrianQC is turned off, the calculation completed successfully. When the calculation crashes, it happens right after BrianQC is initialized.
DFTD3 is installed via psi4-rt. Psi4-1.4rc1 is compiled with all the requirements obtained via conda and using psi4-path-advisor --gcc. OS: CentOS 7.
I just tried to run this with a trial license for BrianQC. My first attempt I got this error:
!------------------------------------------------------------------!
! !
! Program dftd3 is registered with QCEngine, but cannot be found. !
! !
!------------------------------------------------------------------!
That was easily fixed with a
conda install dftd3 -c psi4
After that, all seems to be well. With BrianQC I can modify this test case to use b3lyp-d3 as the method and I get these energies with BrianQC enabled (and default convergence criteria):
=> Energetics <=
Nuclear Repulsion Energy = 9.1966915808960046
One-Electron Energy = -122.3629361763915000
Two-Electron Energy = 45.4862244032741998
DFT Exchange-Correlation Energy = -7.6325667876763426
Empirical Dispersion Energy = -0.0000076300000000
VV10 Nonlocal Energy = 0.0000000000000000
Total Energy = -75.3125946098976300
These are the results without BrianQC enabled:
=> Energetics <=
Nuclear Repulsion Energy = 9.1966915808960046
One-Electron Energy = -122.3629315489535685
Two-Electron Energy = 45.4862194479884181
DFT Exchange-Correlation Energy = -7.6325662102859591
Empirical Dispersion Energy = -0.0000076300000000
VV10 Nonlocal Energy = 0.0000000000000000
Total Energy = -75.3125943603551065
I think the first test is to make sure that your -D3 jobs work without BrainQC enabled, and that the seg fault isn’t just resulting from the QCEngine error that I noted above.
This is without BrianQC turned on.
-----------------------------------------------------------------------
Psi4: An Open-Source Ab Initio Electronic Structure Package
Psi4 1.4rc1 prerelease
Git: Rev {} 2d37164
D. G. A. Smith, L. A. Burns, A. C. Simmonett, R. M. Parrish,
M. C. Schieber, R. Galvelis, P. Kraus, H. Kruse, R. Di Remigio,
A. Alenaizan, A. M. James, S. Lehtola, J. P. Misiewicz, M. Scheurer,
R. A. Shaw, J. B. Schriber, Y. Xie, Z. L. Glick, D. A. Sirianni,
J. S. O'Brien, J. M. Waldrop, A. Kumar, E. G. Hohenstein,
B. P. Pritchard, B. R. Brooks, H. F. Schaefer III, A. Yu. Sokolov,
K. Patkowski, A. E. DePrince III, U. Bozkaya, R. A. King,
F. A. Evangelista, J. M. Turney, T. D. Crawford, C. D. Sherrill,
J. Chem. Phys. 152(18) 184108 (2020). https://doi.org/10.1063/5.0006002
Additional Code Authors
E. T. Seidl, C. L. Janssen, E. F. Valeev, M. L. Leininger,
J. F. Gonthier, R. M. Richard, H. R. McAlexander, M. Saitow, X. Wang,
P. Verma, and M. H. Lechner
Previous Authors, Complete List of Code Contributors,
and Citations for Specific Modules
https://github.com/psi4/psi4/blob/master/codemeta.json
https://github.com/psi4/psi4/graphs/contributors
http://psicode.org/psi4manual/master/introduction.html#citing-psifour
-----------------------------------------------------------------------
Psi4 started on: Tuesday, 11 May 2021 05:08PM
Process ID: 12872
Host: compute-0-4.local
PSIDATADIR: /share/apps/psi4-master/share/psi4
Memory: 500.0 MiB
Threads: 8
==> Input File <==
memory 4gb
molecule ethane {
symmetry c1
0 1
H 1.1851 -0.0039 0.9875
C 0.7516 -0.0225 -0.0209
H 1.1669 0.8330 -0.5693
H 1.1155 -0.9329 -0.5145
C -0.7516 0.0225 0.0209
H -1.1669 -0.8334 0.5687
H -1.1157 0.9326 0.5151
H -1.1850 0.0044 -0.9875
}
Memory set to 3.725 GiB by Python driver.
Scratch directory: /state/partition1/psi4scratch/
*** tstart() called on compute-0-4.local
*** at Tue May 11 17:08:09 2021
=> Loading Basis Set <=
Name: DEF2-TZVP
Role: ORBITAL
Keyword: BASIS
atoms 1, 3-4, 6-8 entry H line 15 file /share/apps/psi4-master/share/psi4/basis/def2-tzvp.gbs
atoms 2, 5 entry C line 123 file /share/apps/psi4-master/share/psi4/basis/def2-tzvp.gbs
=> B3LYP-D3: Empirical Dispersion <=
Grimme's -D3 (zero-damping) Dispersion Correction
Grimme S.; Antony J.; Ehrlich S.; Krieg H. (2010), J. Chem. Phys., 132: 154104
s6 = 1.000000
s8 = 1.703000
sr6 = 1.261000
alpha6 = 14.000000
sr8 = 1.000000
---------------------------------------------------------
SCF
by Justin Turney, Rob Parrish, Andy Simmonett
and Daniel G. A. Smith
RKS Reference
8 Threads, 3814 MiB Core
---------------------------------------------------------
==> Geometry <==
Molecular point group: c1
Full point group: C1
Geometry (in Angstrom), charge = 0, multiplicity = 1:
Center X Y Z Mass
------------ ----------------- ----------------- ----------------- -----------------
H 1.185103354167 -0.003893291665 0.987500000000 1.007825032230
C 0.751603354167 -0.022493291665 -0.020900000000 12.000000000000
H 1.166903354167 0.833006708335 -0.569300000000 1.007825032230
H 1.115503354167 -0.932893291665 -0.514500000000 1.007825032230
C -0.751596645833 0.022506708335 0.020900000000 12.000000000000
H -1.166896645833 -0.833393291665 0.568700000000 1.007825032230
H -1.115696645833 0.932606708335 0.515100000000 1.007825032230
H -1.184996645833 0.004406708335 -0.987500000000 1.007825032230
Running in c1 symmetry.
Rotational constants: A = 2.67703 B = 0.67917 C = 0.67917 [cm^-1]
Rotational constants: A = 80255.47782 B = 20361.04124 C = 20360.88887 [MHz]
Nuclear repulsion = 42.383741222683696
Charge = 0
Multiplicity = 1
Electrons = 18
Nalpha = 9
Nbeta = 9
==> Algorithm <==
SCF Algorithm Type is DIRECT.
DIIS enabled.
MOM disabled.
Fractional occupation disabled.
Guess Type is SAD.
Energy threshold = 1.00e-06
Density threshold = 1.00e-06
Integral threshold = 1.00e-12
==> Primary Basis <==
Basis Set: DEF2-TZVP
Blend: DEF2-TZVP
Number of shells: 46
Number of basis functions: 98
Number of Cartesian functions: 108
Spherical Harmonics?: true
Max angular momentum: 3
==> DFT Potential <==
=> Composite Functional: B3LYP-D3ZERO <=
B3LYP-d3zero Hyb-GGA Exchange-Correlation Functional
P. J. Stephens, F. J. Devlin, C. F. Chabalowski, and M. J. Frisch, J. Phys. Chem. 98, 11623 (1994) (10.1021/j100096a001)
Deriv = 1
GGA = TRUE
Meta = FALSE
Exchange Hybrid = TRUE
MP2 Hybrid = FALSE
=> Exchange Functionals <=
0.0800 Slater exchange
0.7200 Becke 88
=> Exact (HF) Exchange <=
0.2000 HF
=> Correlation Functionals <=
0.1900 Vosko, Wilk & Nusair (VWN5_RPA)
0.8100 Lee, Yang & Parr
=> LibXC Density Thresholds <==
XC_HYB_GGA_XC_B3LYP: 1.00E-15
=> Molecular Quadrature <=
Radial Scheme = TREUTLER
Pruning Scheme = NONE
Nuclear Scheme = TREUTLER
BS radius alpha = 1
Pruning alpha = 1
Radial Points = 75
Spherical Points = 302
Total Points = 22815
Total Blocks = 248
Max Points = 254
Max Functions = 97
Weights Tolerance = 1.00E-15
=> Loading Basis Set <=
Name: (DEF2-TZVP AUX)
Role: JKFIT
Keyword: DF_BASIS_SCF
atoms 1, 3-4, 6-8 entry H line 18 file /share/apps/psi4-master/share/psi4/basis/def2-universal-jkfit.gbs
atoms 2, 5 entry C line 198 file /share/apps/psi4-master/share/psi4/basis/def2-universal-jkfit.gbs
Starting with a DF guess…
==> Integral Setup <==
DFHelper Memory: AOs need 0.021 GiB; user supplied 2.733 GiB. Using in-core AOs.
==> MemDFJK: Density-Fitted J/K Matrices <==
J tasked: Yes
K tasked: Yes
wK tasked: No
OpenMP threads: 8
Memory [MiB]: 2798
Algorithm: Core
Schwarz Cutoff: 1E-12
Mask sparsity (%): 0.0208
Fitting Condition: 1E-10
=> Auxiliary Basis Set <=
Basis Set: (DEF2-TZVP AUX)
Blend: DEF2-UNIVERSAL-JKFIT
Number of shells: 86
Number of basis functions: 258
Number of Cartesian functions: 298
Spherical Harmonics?: true
Max angular momentum: 4
Cached 100.0% of DFT collocation blocks in 0.061 [GiB].
Minimum eigenvalue in the overlap matrix is 5.2407357025E-04.
Reciprocal condition number of the overlap matrix is 6.5501213601E-05.
Using symmetric orthogonalization.
==> Pre-Iterations <==
SCF Guess: Superposition of Atomic Densities via on-the-fly atomic UHF (no occupation information).
Irrep Nso Nmo
A 98 98
Total 98 98
==> Iterations <==
Total Energy Delta E RMS |[F,P]|
@DF-RKS iter SAD: -79.29190928073774 -7.92919e+01 0.00000e+00
@DF-RKS iter 1: -79.46906741264431 -1.77158e-01 9.08512e-03 DIIS
@DF-RKS iter 2: -79.17412153337423 2.94946e-01 1.11855e-02 DIIS
@DF-RKS iter 3: -79.86338376185476 -6.89262e-01 6.26304e-04 DIIS
@DF-RKS iter 4: -79.86546025189487 -2.07649e-03 1.22636e-04 DIIS
@DF-RKS iter 5: -79.86552888858671 -6.86367e-05 1.44313e-05 DIIS
@DF-RKS iter 6: -79.86552993593324 -1.04735e-06 4.65150e-07 DIIS
@DF-RKS iter 7: -79.86552993708969 -1.15645e-09 9.02709e-08 DIIS
DF guess converged.
==> DirectJK: Integral-Direct J/K Matrices <==
J tasked: Yes
K tasked: Yes
wK tasked: No
Integrals threads: 8
Schwarz Cutoff: 1E-12
@RKS iter 8: -79.86544800908581 -7.98654e+01 1.11136e-05 DIIS
@RKS iter 9: -79.86544841044432 -4.01359e-07 7.08499e-07 DIIS
Energy and wave function converged.
==> Post-Iterations <==
Electrons on quadrature grid:
Ntotal = 18.0001069966 ; deviation = 1.070e-04
Orbital Energies [Eh]
---------------------
Doubly Occupied:
1A -10.159640 2A -10.159395 3A -0.757632
4A -0.614901 5A -0.437755 6A -0.437746
7A -0.375611 8A -0.342731 9A -0.342712
Virtual:
10A 0.038508 11A 0.077395 12A 0.105638
13A 0.105646 14A 0.121854 15A 0.121876
16A 0.192134 17A 0.193028 18A 0.233922
19A 0.233973 20A 0.241825 21A 0.241894
22A 0.396359 23A 0.420270 24A 0.428064
25A 0.436250 26A 0.436294 27A 0.495233
28A 0.495246 29A 0.549411 30A 0.570287
31A 0.647808 32A 0.647853 33A 0.826928
34A 0.826973 35A 0.919466 36A 1.035545
37A 1.035583 38A 1.042489 39A 1.042521
40A 1.098506 41A 1.472169 42A 1.472176
43A 1.479022 44A 1.517560 45A 1.534760
46A 1.544067 47A 1.564068 48A 1.564092
49A 1.727606 50A 1.742869 51A 1.788305
52A 1.788336 53A 1.910395 54A 1.910440
55A 1.979296 56A 2.122795 57A 2.277178
58A 2.318099 59A 2.318144 60A 2.333996
61A 2.334004 62A 2.393773 63A 2.393830
64A 2.523515 65A 2.523560 66A 2.566097
67A 2.566148 68A 2.635732 69A 2.635820
70A 2.695616 71A 2.855350 72A 2.919342
73A 2.997503 74A 2.997580 75A 3.083777
76A 3.083850 77A 3.157823 78A 3.160124
79A 3.265372 80A 3.339096 81A 3.339147
82A 3.469205 83A 3.497090 84A 3.497142
85A 3.742764 86A 3.742842 87A 3.910032
88A 3.910084 89A 3.917968 90A 4.171211
91A 4.330357 92A 4.330460 93A 4.417098
94A 4.417166 95A 4.549909 96A 4.727892
97A 22.227159 98A 22.308604
Final Occupation by Irrep:
A
DOCC [ 9 ]
@RKS Final Energy: -79.86544841044432
=> Energetics <=
Nuclear Repulsion Energy = 42.3837412226836960
One-Electron Energy = -189.6079461370884189
Two-Electron Energy = 77.9692548923227378
DFT Exchange-Correlation Energy = -10.6087106083623510
Empirical Dispersion Energy = -0.0017877800000000
VV10 Nonlocal Energy = 0.0000000000000000
Total Energy = -79.8654484104443156
Computation Completed
Properties will be evaluated at 0.000000, 0.000000, 0.000000 [a0]
Properties computed using the SCF density matrix
Nuclear Dipole Moment: [e a0]
X: -0.0001 Y: -0.0001 Z: -0.0000
Electronic Dipole Moment: [e a0]
X: 0.0001 Y: 0.0001 Z: -0.0000
Dipole Moment: [e a0]
X: 0.0000 Y: -0.0000 Z: -0.0000 Total: 0.0001
Dipole Moment: [D]
X: 0.0000 Y: -0.0001 Z: -0.0001 Total: 0.0001
*** tstop() called on compute-0-4.local at Tue May 11 17:08:11 2021
Module time:
user time = 9.71 seconds = 0.16 minutes
system time = 0.31 seconds = 0.01 minutes
total time = 2 seconds = 0.03 minutes
Total time:
user time = 9.71 seconds = 0.16 minutes
system time = 0.31 seconds = 0.01 minutes
total time = 2 seconds = 0.03 minutes
FCHKWriter: !WARNING! method ‘DFT’’ renamed to label ‘DFT’.
FCHKWriter: Writing t13.fchk with label ’ SCF Density’.
Psi4 stopped on: Tuesday, 11 May 2021 05:08PM
Psi4 wall time for execution: 0:00:02.53
*** Psi4 exiting successfully. Buy a developer a beer!
++++++++++++++++++++++++++++++++++++++++++++++++++++++++
This is with BrianQC turned on.
tid: 0x0 [2021.05.11-17:21:14.314] [INFO] Detected compatible BrianQC interface version 5
tid: 0x0 [2021.05.11-17:21:14.314] [INFO] Reading config from /share/apps/brianqc/1.1.0/config.json
tid: 0x0 [2021.05.11-17:21:14.314] [INFO] Setting Brian root dir to /share/apps/brianqc/1.1.0
tid: 0x0 [2021.05.11-17:21:14.315] [INFO] Found kernel DB in config: /share/apps/brianqc/1.1.0/integrators/17103982592-1586948883_kernel_g_v1_1_0_0a5a7a8830_CUDA_GeForce_GTX_1080_Ti.db
tid: 0x0 [2021.05.11-17:21:14.316] [INFO] Found DFT DB in config: /share/apps/brianqc/1.1.0/integrators/15589376-1589985794_dft_kernel_v1_1_0_9fb824961_CUDA_GeForce_GTX_1080_Ti.db
tid: 0x0 [2021.05.11-17:21:14.316] [INFO] Found MM DB in config: /share/apps/brianqc/1.1.0/integrators/897024-1587559005_mm_kernel_v1_1_0_e1ee343bd1_CUDA_GeForce_GTX_1080_Ti.db
tid: 0x0 [2021.05.11-17:21:14.316] [INFO] Start BrianQC module initialization
tid: 0x0 [2021.05.11-17:21:14.319] [SYSTEM] BrianQC version: v1_1_0_9b6b114d45 licensed to: farid
tid: 0x0 [2021.05.11-17:21:14.323] [SYSTEM] Visit us at: http://brianqc.com
tid: 0x0 [2021.05.11-17:21:14.334] [SYSTEM] BrianQC copyright Streamnovation Ltd. 2013-2019
tid: 0x0 [2021.05.11-17:21:14.343] [SYSTEM] Contributors: Gy. Cserey, G. Kis, I. Ladjanszki, A. Rak, G. J. Tornai, A. Vas, B. Czetenyi
tid: 0x0 [2021.05.11-17:21:14.551] [INFO] Detected device: CPU
tid: 0x0 [2021.05.11-17:21:14.551] [INFO] Detected device: CUDA GeForce GTX 1080 Ti#0000:84:00.0
tid: 0x0 [2021.05.11-17:21:14.575] [INFO] Allocating 3 threads for GPU: CUDA GeForce GTX 1080 Ti#0000:84:00.0 on node 0
tid: 0x0 [2021.05.11-17:21:16.544] [INFO] Scratch dir is disabled, storing everything in memory
tid: 0x0 [2021.05.11-17:21:16.544] [INFO] End BrianQC module initialization
tid: 0x0 [2021.05.11-17:22:10.099] [INFO] Detected compatible BrianQC interface version 5
tid: 0x0 [2021.05.11-17:22:10.099] [INFO] Reading config from /share/apps/brianqc/1.1.0/config.json
tid: 0x0 [2021.05.11-17:22:10.099] [INFO] Setting Brian root dir to /share/apps/brianqc/1.1.0
tid: 0x0 [2021.05.11-17:22:10.100] [INFO] Found kernel DB in config: /share/apps/brianqc/1.1.0/integrators/17103982592-1586948883_kernel_g_v1_1_0_0a5a7a8830_CUDA_GeForce_GTX_1080_Ti.db
tid: 0x0 [2021.05.11-17:22:10.100] [INFO] Found DFT DB in config: /share/apps/brianqc/1.1.0/integrators/15589376-1589985794_dft_kernel_v1_1_0_9fb824961_CUDA_GeForce_GTX_1080_Ti.db
tid: 0x0 [2021.05.11-17:22:10.100] [INFO] Found MM DB in config: /share/apps/brianqc/1.1.0/integrators/897024-1587559005_mm_kernel_v1_1_0_e1ee343bd1_CUDA_GeForce_GTX_1080_Ti.db
tid: 0x0 [2021.05.11-17:22:10.101] [INFO] Start BrianQC module initialization
tid: 0x0 [2021.05.11-17:22:10.104] [SYSTEM] BrianQC version: v1_1_0_9b6b114d45 licensed to: farid
tid: 0x0 [2021.05.11-17:22:10.108] [SYSTEM] Visit us at: http://brianqc.com
tid: 0x0 [2021.05.11-17:22:10.119] [SYSTEM] BrianQC copyright Streamnovation Ltd. 2013-2019
tid: 0x0 [2021.05.11-17:22:10.128] [SYSTEM] Contributors: Gy. Cserey, G. Kis, I. Ladjanszki, A. Rak, G. J. Tornai, A. Vas, B. Czetenyi
tid: 0x0 [2021.05.11-17:22:10.338] [INFO] Detected device: CPU
tid: 0x0 [2021.05.11-17:22:10.338] [INFO] Detected device: CUDA GeForce GTX 1080 Ti#0000:84:00.0
tid: 0x0 [2021.05.11-17:22:10.359] [INFO] Allocating 3 threads for GPU: CUDA GeForce GTX 1080 Ti#0000:84:00.0 on node 0
tid: 0x0 [2021.05.11-17:22:12.368] [INFO] Scratch dir is disabled, storing everything in memory
tid: 0x0 [2021.05.11-17:22:12.368] [INFO] End BrianQC module initialization
This is the content of the queue file:
tid: 0x0 [2021.05.11-17:09:54.849] [INFO] Detected compatible BrianQC interface version 5
tid: 0x0 [2021.05.11-17:09:54.849] [INFO] Reading config from /share/apps/brianqc/1.1.0/config.json
tid: 0x0 [2021.05.11-17:09:54.850] [INFO] Setting Brian root dir to /share/apps/brianqc/1.1.0
tid: 0x0 [2021.05.11-17:09:54.850] [INFO] Found kernel DB in config: /share/apps/brianqc/1.1.0/integrators/17103982592-1586948883_kernel_g_v1_1_0_0a5a7a8830_CUDA_GeForce_GTX_1080_Ti.db
tid: 0x0 [2021.05.11-17:09:54.851] [INFO] Found DFT DB in config: /share/apps/brianqc/1.1.0/integrators/15589376-1589985794_dft_kernel_v1_1_0_9fb824961_CUDA_GeForce_GTX_1080_Ti.db
tid: 0x0 [2021.05.11-17:09:54.851] [INFO] Found MM DB in config: /share/apps/brianqc/1.1.0/integrators/897024-1587559005_mm_kernel_v1_1_0_e1ee343bd1_CUDA_GeForce_GTX_1080_Ti.db
tid: 0x0 [2021.05.11-17:09:54.851] [INFO] Start BrianQC module initialization
tid: 0x0 [2021.05.11-17:09:54.855] [SYSTEM] BrianQC version: v1_1_0_9b6b114d45 licensed to: farid
tid: 0x0 [2021.05.11-17:09:54.858] [SYSTEM] Visit us at: http://brianqc.com
tid: 0x0 [2021.05.11-17:09:54.869] [SYSTEM] BrianQC copyright Streamnovation Ltd. 2013-2019
tid: 0x0 [2021.05.11-17:09:54.879] [SYSTEM] Contributors: Gy. Cserey, G. Kis, I. Ladjanszki, A. Rak, G. J. Tornai, A. Vas, B. Czetenyi
tid: 0x0 [2021.05.11-17:09:55.100] [INFO] Detected device: CPU
tid: 0x0 [2021.05.11-17:09:55.100] [INFO] Detected device: CUDA GeForce GTX 1080 Ti#0000:05:00.0
tid: 0x0 [2021.05.11-17:09:55.125] [INFO] Allocating 3 threads for GPU: CUDA GeForce GTX 1080 Ti#0000:05:00.0 on node 0
tid: 0x0 [2021.05.11-17:09:57.151] [INFO] Scratch dir is disabled, storing everything in memory
tid: 0x0 [2021.05.11-17:09:57.151] [INFO] End BrianQC module initialization
/opt/gridengine/default/spool/compute-0-4/job_scripts/1161: line 17: 13209 Segmentation fault /share/apps/psi4-master/bin/psi4 -n $NPROC $INPUT.inp $INPUT.out
Any idea what is wrong with the input?
This is the input with BrianQC turned on:
memory 4gb
molecule ethane {
symmetry c1
0 1
H 1.1851 -0.0039 0.9875
C 0.7516 -0.0225 -0.0209
H 1.1669 0.8330 -0.5693
H 1.1155 -0.9329 -0.5145
C -0.7516 0.0225 0.0209
H -1.1669 -0.8334 0.5687
H -1.1157 0.9326 0.5151
H -1.1850 0.0044 -0.9875
}
set = {
basis def2-tzvp
reference rhf
dft_grid_name sg1
scf_type direct
step_type nr
geom_maxiter 200
brianqc_enable true
}
E, wfn = energy(‘scf’, dft_functional=‘b3lyp-d3’, return_wfn=True)
fchk(wfn,‘t13.fchk’)
My trial license has expired now, so I can’t debug any further at this point unfortunately. I was able to run the above input without any problems on a V100 card. I tried different numbers of threads because it appears that your input is requesting more than one, but none of my tests could reproduce the error. Do you have access to any cards other than a 1080? That might be worth trying to see if it is specific to a particular type of hardware.
Must something weird with my system. I’ve tried a lot of things, installed various libs, installed fresh new environments, tried all python3 versions, and rebooted the system, but nothing works. I think I’ll just stick with -D2 for now, and comeback when v1.4 final is released or when I have more time to look at this. Currently I don’t have any other cards. Thanks for looking at this issue.
You should also contact BrianQC for advice.
OK, I’m working with BrianQC to try figure out what is going on.
Given that I’m unable to reproduce it on this end, I think the only way I can offer help is to walk you through the steps for tracing the error in a debugger. It’s an involved process, but if you’re willing to get your hands dirty and can’t find a simple resolution through the BrianQC folks, please let me know.
I wouldn’t mind to get my hands dirty as long as I can get this to work. It does seem like the segfault is coming during the BrianQC initialization, so I’m still hoping the BrianQC people will figure this out. Thanks.