Dear all,
recently I didn't succeded to simulate my system using NPT conditions and
explicit solvent( 123548 atoms in total )
with AMBER 11 (bugfixes up to 12 applied) on GeForce GTX 470 (*.in file
below) due to memmory allocation error:
cudaMalloc GpuBuffer::Allocate failed out of memory
This seems to me pretty strange from several reasons:
#1
When I put this simulation on Tesla C2050 I have learned
that the memory usage is just 22% (using nvidia-smi command ).
Assuming 3GB memory of C2050, 22% shoud be 660MB but
GeForce GTX 470 has 1280MB available so ?
#2
I also tried to verify amount of memory used with TOP command.
Assuming that usage of RAM memory should be similar to GPU memory
consumption
due to data exchange between GPU and CPU. I have obtained this result:
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
25677 mara 20 0 418m 201m 25m R 100 2.5 2:52.16 pmemd.cuda
So it seems that there is 418MB RAM memory used in total which is roughly
comparable with that 660MB.
#3
NVT simulation of this system was OK also on GeForce GTX 470 !
When I analysed GPU resources usage using nvidia-smi I have obtained
here just 21% memory utilisation !
TOP command showed me here this info:
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
12240 mmaly 20 0 365m 154m 20m R 100 1.9 1:21.63 pmemd.cuda
When I put this NVT simulation just for curiosity to Tesla C2050,
and tried TOP command I got:
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
24921 mara 20 0 406m 193m 25m R 100 2.4 181:57.91 pmemd.cuda
It seems that RAM usage is almost the same in both cases (NPT 418MB, NVT
406MB) on Tesla so
I would suppose the same in case of GPU memory usage, but here
surprisingly in case
of NPT the calculation crashed due to insuficient memory (probably that
GPU as machine has 8GB of RAM )
and in case of NVT nvidia-smi inform me just about 21% memory utilisation.
#4
The last "mystery" is for me also the fact that command nvidia-smi provide
me 20% memory utilisation info regarding to NVT simulation also on Tesla
machine so we have:
21% on GeForce GTX 470 with 1280MB memory and 20% on Tesla C2050 with 3GB
memory regarding to the
identical simulation...
#5
I have no problems to run on GeForce GTX 470 NPT simulation of just a
little smaller
system ( 112166 atoms).
So first of all I would like to know if someone succeeded to run NPT
simulation with explicit solvent
of the system 123548 and more atoms. The other question is regarding to
GPU memory management.
Is it all or at least the most (let say 85% and more) GPU memory available
for allocation of cuda applications
or there is some stronger limitation by default which might be eventually
changed somehow (especially in case of GTX 470)?
And the last my question is related to nvidia-smi record " Memory :". I
was thinking that
it tells us how many percent of GPU memory is actually used but I am not
sure about this interpretation now especially
regarding to #4.
Thanks a lot in advance for any valuable comments !
Best wishes,
Marek
Here is my NPT input file:
///////////////// NPT *.in file
&cntrl
imin=0,irest=1,ntx=5,
nstlim=250000,dt=0.002,
ntc=2,ntf=2,
cut=10.0, ntb=2, ntp=1, taup=1.0,
ntpr=5000, ntwx=5000,
ntt=3, gamma_ln=2.0,
temp0=298,
/
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Feb 18 2011 - 19:00:02 PST
