jianmu
/
lammps
forked from lijiext/lammps
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

git-svn-id: svn://svn.icms.temple.edu/lammps-ro/trunk@12685 f3b2605a-c512-4ea7-a41b-209d697bcdaa

This commit is contained in:
sjplimp 2014-11-11 15:08:18 +00:00
parent 59f631c129
commit 909d622b35
2 changed files with 16 additions and 14 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

15
doc/package.html
View File

@ -275,12 +275,13 @@ the default.
<P>The <I>split</I> keyword can be used for load balancing force calculations <P>The <I>split</I> keyword can be used for load balancing force calculations
between CPU and GPU cores in GPU-enabled pair styles. If 0 < <I>split</I> < between CPU and GPU cores in GPU-enabled pair styles. If 0 < <I>split</I> <
1.0, a fixed fraction of particles is offloaded to the GPU while force 1.0, a fixed fraction of particles is offloaded to the GPU while force
calculation for the other particles occurs simulataneously on the calculation for the other particles occurs simulataneously on the CPU.
CPU. If <I>split</I> < 0.0, the optimal fraction (based on CPU and GPU If <I>split</I> < 0.0, the optimal fraction (based on CPU and GPU timings)
timings) is calculated every 25 timesteps. If <I>split</I> = 1.0, all is calculated every 25 timesteps, i.e. dynamic load-balancing across
force calculations for GPU accelerated pair styles are performed on the CPU and GPU is performed. If <I>split</I> = 1.0, all force
the GPU. In this case, other <A HREF = "pair_hybrid.html">hybrid</A> pair calculations for GPU accelerated pair styles are performed on the GPU.
interactions, <A HREF = "bond_style.html">bond</A>, <A HREF = "angle_style.html">angle</A>, In this case, other <A HREF = "pair_hybrid.html">hybrid</A> pair interactions,
<A HREF = "bond_style.html">bond</A>, <A HREF = "angle_style.html">angle</A>,
<A HREF = "dihedral_style.html">dihedral</A>, <A HREF = "improper_style.html">improper</A>, and <A HREF = "dihedral_style.html">dihedral</A>, <A HREF = "improper_style.html">improper</A>, and
<A HREF = "kspace_style.html">long-range</A> calculations can be performed on the <A HREF = "kspace_style.html">long-range</A> calculations can be performed on the
CPU while the GPU is performing force calculations for the GPU-enabled CPU while the GPU is performing force calculations for the GPU-enabled
@ -299,7 +300,7 @@ package gpu 2 split -1 # input script command
utilized. Each GPU device would be shared by 4 CPU cores. The CPU utilized. Each GPU device would be shared by 4 CPU cores. The CPU
cores would perform force calculations for some fraction of the cores would perform force calculations for some fraction of the
particles at the same time the GPUs performed force calculation for particles at the same time the GPUs performed force calculation for
the other particles. the other particles.
</P> </P>
<P>The <I>gpuID</I> keyword allows selection of which GPUs on each node will <P>The <I>gpuID</I> keyword allows selection of which GPUs on each node will
be used for a simulation. The <I>first</I> and <I>last</I> values specify the be used for a simulation. The <I>first</I> and <I>last</I> values specify the

15
doc/package.txt
View File

@ -269,12 +269,13 @@ the default.
The {split} keyword can be used for load balancing force calculations The {split} keyword can be used for load balancing force calculations
between CPU and GPU cores in GPU-enabled pair styles. If 0 < {split} < between CPU and GPU cores in GPU-enabled pair styles. If 0 < {split} <
1.0, a fixed fraction of particles is offloaded to the GPU while force 1.0, a fixed fraction of particles is offloaded to the GPU while force
calculation for the other particles occurs simulataneously on the calculation for the other particles occurs simulataneously on the CPU.
CPU. If {split} < 0.0, the optimal fraction (based on CPU and GPU If {split} < 0.0, the optimal fraction (based on CPU and GPU timings)
timings) is calculated every 25 timesteps. If {split} = 1.0, all is calculated every 25 timesteps, i.e. dynamic load-balancing across
force calculations for GPU accelerated pair styles are performed on the CPU and GPU is performed. If {split} = 1.0, all force
the GPU. In this case, other "hybrid"_pair_hybrid.html pair calculations for GPU accelerated pair styles are performed on the GPU.
interactions, "bond"_bond_style.html, "angle"_angle_style.html, In this case, other "hybrid"_pair_hybrid.html pair interactions,
"bond"_bond_style.html, "angle"_angle_style.html,
"dihedral"_dihedral_style.html, "improper"_improper_style.html, and "dihedral"_dihedral_style.html, "improper"_improper_style.html, and
"long-range"_kspace_style.html calculations can be performed on the "long-range"_kspace_style.html calculations can be performed on the
CPU while the GPU is performing force calculations for the GPU-enabled CPU while the GPU is performing force calculations for the GPU-enabled
@ -293,7 +294,7 @@ In this case, all CPU cores and GPU devices on the nodes would be
utilized. Each GPU device would be shared by 4 CPU cores. The CPU utilized. Each GPU device would be shared by 4 CPU cores. The CPU
cores would perform force calculations for some fraction of the cores would perform force calculations for some fraction of the
particles at the same time the GPUs performed force calculation for particles at the same time the GPUs performed force calculation for
the other particles. the other particles.
The {gpuID} keyword allows selection of which GPUs on each node will The {gpuID} keyword allows selection of which GPUs on each node will
be used for a simulation. The {first} and {last} values specify the be used for a simulation. The {first} and {last} values specify the