[AMBER] Amber16 and CUDA8.0 pmemd.cuda worked, and then stopped working, saying "cudaMemcpyToSymbol: SetSim copy to cSim failed invalid device symbol"

From: Zack Scholl <zns.duke.edu>
Date: Tue, 21 Jun 2016 12:44:14 -0400

Dear all,

I'm using Amber16 with AmberTools16 and CUDA-8.0.

I made a fresh install for Amber16 pmemd.cuda using

./configure -cuda gnu && source amber.sh && make install

After that, pmemd.cuda was installed and it worked. However it did not
install the AmberTools (cpptraj, tleap, etc). So I then ran

./configure gnu && source amber.sh && make install

to get those tools.

However now, whenever I run pmemd.cuda as before, I get an error:

> $AMBERHOME/bin/pmemd.cuda -O -i 01_Min.in -o 01_Min.out -p prmtop -c
inpcrd -r 01_Min.rst -inf 01_Min.mdinfo
cudaMemcpyToSymbol: SetSim copy to cSim failed invalid device symbol

And now I can not get rid of this error.

I've tried fresh installing, reinstalling pmemd.cuda from scratch, without
the tools, and it didn't work this time. I've tried installing the
AmberTools first and then Amber16 with cuda second, but now still gives the
same error.

I am very confused, and I can't think of why it would work before and now
no longer works.

Has anyone else received this type of error, or could give some advice onto
how to proceed? I've seen in the forums that this error had been received,
but it was because of out-of-date drivers which should not be the case here.

Thanks!

By the way, here is the output of my minimization (which worked previously)

> cat 01_Min.out

cat 01_Min.out
          -------------------------------------------------------
          Amber 16 PMEMD 2016
          -------------------------------------------------------

| PMEMD implementation of SANDER, Release 16

| Run on 06/21/2016 at 10:48:43

| Executable path: /usr/local/amber16/bin/pmemd.cuda
| Working directory: /home/zns/homeodomain
| Hostname:

  [-O]verwriting output

File Assignments:
| MDIN:
01_Min.in
| MDOUT:
01_Min.out
| INPCRD:
inpcrd
| PARM:
prmtop
| RESTRT:
01_Min.rst
| REFC:
refc
| MDVEL:
mdvel
| MDEN:
mden
| MDCRD:
mdcrd
| MDINFO:
01_Min.mdinfo
| MDFRC:
mdfrc


 Here is the input file:

Minimize

 &cntrl


imin=1,


ntx=1,


irest=0,


maxcyc=2000,


ncyc=1000,


ntpr=100,


ntwx=0,


cut=8.0,

 /



Note: ig = -1. Setting random seed to 642769 based on wallclock time in
      microseconds.

|--------------------- INFORMATION ----------------------
| GPU (CUDA) Version of PMEMD in use: NVIDIA GPU IN USE.
| Version 16.0.0
|
| 02/25/2016
|
| Implementation by:
| Ross C. Walker (SDSC)
| Scott Le Grand (nVIDIA)
|
| Precision model in use:
| [SPFP] - Single Precision Forces, 64-bit Fixed Point
| Accumulation. (Default)
|
|--------------------------------------------------------

|----------------- CITATION INFORMATION -----------------
|
| When publishing work that utilized the CUDA version
| of AMBER, please cite the following in addition to
| the regular AMBER citations:
|
| - Romelia Salomon-Ferrer; Andreas W. Goetz; Duncan
| Poole; Scott Le Grand; Ross C. Walker "Routine
| microsecond molecular dynamics simulations with
| AMBER - Part II: Particle Mesh Ewald", J. Chem.
| Theory Comput., 2013, 9 (9), pp3878-3888,
| DOI: 10.1021/ct400314y.
|
| - Andreas W. Goetz; Mark J. Williamson; Dong Xu;
| Duncan Poole; Scott Le Grand; Ross C. Walker
| "Routine microsecond molecular dynamics simulations
| with AMBER - Part I: Generalized Born", J. Chem.
| Theory Comput., 2012, 8 (5), pp1542-1555.
|
| - Scott Le Grand; Andreas W. Goetz; Ross C. Walker
| "SPFP: Speed without compromise - a mixed precision
| model for GPU accelerated molecular dynamics
| simulations.", Comp. Phys. Comm., 2013, 184
| pp374-380, DOI: 10.1016/j.cpc.2012.09.022
|
|--------------------------------------------------------

|------------------- GPU DEVICE INFO --------------------
|
| CUDA_VISIBLE_DEVICES: 0
| CUDA Capable Devices Detected: 1
| CUDA Device ID in use: 0
| CUDA Device Name: GeForce GTX 1080
| CUDA Device Global Mem Size: 8113 MB
| CUDA Device Num Multiprocessors: 20
| CUDA Device Core Freq: 1.73 GHz
|
|--------------------------------------------------------


| Conditional Compilation Defines Used:
| PUBFFT
| BINTRAJ
| CUDA
| EMIL

| Largest sphere to fit in unit cell has radius = 27.487


​PS. Sorry if this has been sent out twice. I didn't receive any
confirmation the first time and was not sure if it had been sent (I've
since modified my AMBER mailing-list settings).​
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Jun 21 2016 - 10:00:02 PDT
Custom Search