diff --git a/.gitignore b/.gitignore
index 192394149f8f95364dbcecacf3de72f5dad6e166..c6f0b2d87af21aa24d620699e95a344b84cd14b5 100644
--- a/.gitignore
+++ b/.gitignore
@@ -69,7 +69,6 @@ software/*
 avg
 installed
 app/*
-./*.config
 test.sh
 *.swp
 workloads/*
\ No newline at end of file
diff --git a/images/wechat-group-qr.png b/images/wechat-group-qr.png
index 9cb188fea98b257a55e59433a6ac2c9f64d103bc..5a4fc668ab90caf1785de301a8e398e21b4aa3ce 100644
Binary files a/images/wechat-group-qr.png and b/images/wechat-group-qr.png differ
diff --git a/templates/lammps/2021.5.27/lammps.arm.kokkos.gpu.config b/templates/lammps/2021.5.27/lammps.arm.kokkos.gpu.config
new file mode 100644
index 0000000000000000000000000000000000000000..1941ecd6d36d9406ab9e0d17427a6d7c19a10e6d
--- /dev/null
+++ b/templates/lammps/2021.5.27/lammps.arm.kokkos.gpu.config
@@ -0,0 +1,42 @@
+[SERVER]
+11.11.11.11
+
+[DOWNLOAD]
+lammps/2021.5.27 https://download.lammps.org/tars/lammps-27May2021.tar.gz
+
+[DEPENDENCY]
+module purge
+./jarvis -install nvhpc/11.4 any
+tar -xvf ${JARVIS_DOWNLOAD}/lammps-27May2021.tar.gz
+
+[ENV]
+module use ./software/modulefiles
+module purge
+module load /opt/nvidia/hpc_sdk/modulefiles/nvhpc/21.9
+
+[APP]
+app_name = lammps
+build_dir = ${JARVIS_ROOT}/lammps-27May2021/src
+binary_dir = ${JARVIS_ROOT}/lammps-27May2021/src
+case_dir = ${JARVIS_ROOT}/workloads/lammps/p2p
+
+[BUILD]
+#install voronoi
+make lib-voronoi args="-b"
+sed -i '13c\CCFLAGS = -g -O3 -fopenmp -DNDEBUG -Xcudafe --diag_suppress=unrecognized_pragma' MAKE/OPTIONS/Makefile.kokkos_cuda_mpi
+sed -i '18c\LINKFLAGS =  -g -O3 -fopenmp' MAKE/OPTIONS/Makefile.kokkos_cuda_mpi
+sed -i '26c\KOKKOS_ARCH = Ampere80' MAKE/OPTIONS/Makefile.kokkos_cuda_mpi
+make clean-all
+make clean-machine
+make no-all
+make no-lib
+make yes-manybody yes-molecule yes-replica yes-kspace yes-asphere yes-rigid yes-misc yes-openmp yes-reaxff
+make yes-voronoi
+make yes-kokkos
+make kokkos_cuda_mpi -j
+
+[RUN]
+#run= ucx_info -d | grep 'Component:' 
+run = mpirun --allow-run-as-root -np 2 --bind-to core -x UCX_TLS=sm,cuda_copy
+binary = lmp_kokkos_cuda_mpi -k on g 2 -sf kk -pk kokkos cuda/aware on neigh full comm device binsize 2.8  -in in.lj.txt
+nodes = 1
\ No newline at end of file
diff --git a/templates/namd/2.14/namd.arm.gpu.config b/templates/namd/2.14/namd.arm.gpu.config
new file mode 100644
index 0000000000000000000000000000000000000000..9adfd7b3253e2166951b3dbedc0431cf031204a2
--- /dev/null
+++ b/templates/namd/2.14/namd.arm.gpu.config
@@ -0,0 +1,67 @@
+[SERVER]
+1.1.11.1
+
+[DOWNLOAD]
+https://www.ks.uiuc.edu/Development/Download/download.cgi?PackageName=NAMD
+
+[DEPENDENCY]
+module use ./software/modulefiles
+module purge
+./jarvis -install bisheng/2.1.0 com
+module load bisheng2/2.1.0
+export CC=clang CXX=clang++ FC=flang
+./jarvis -install hmpi/1.2.0 clang
+tar -xzvf $JARVIS_DOWNLOAD/NAMD_2.14_Source.tar.gz
+cd NAMD_2.14_Source
+tar xvf charm-6.10.2.tar
+
+[ENV]
+module use ./software/modulefiles
+module purge
+module load bisheng2/2.1.0
+module load hmpi1/1.2.0
+# add cuda
+export CUDA_INCLUDE_DIRS=/usr/local/cuda/include
+export CUDA_CUDART_LIBRARY=/usr/local/cuda/lib64/libcudart.so
+export CUDA_HOME=/usr/local/cuda
+export PATH=$CUDA_HOME/bin:$PATH
+export LD_LIBRARY_PATH=$CUDA_HOME/lib64:$LD_LIBRARY_PATH
+export C_INCLUDE_PATH=$CUDA_HOME/include:$C_INCLUDE_PATH
+export LIBRARY_PATH=$CUDA_HOME/lib64:$LIBRARY_PATH
+
+[APP]
+app_name = NAMD
+build_dir = $JARVIS_ROOT/NAMD_2.14_Source/
+binary_dir = $JARVIS_ROOT/NAMD_2.14_Source/Linux-ARM64-g++/
+case_dir = ${JARVIS_ROOT}/stmv
+
+[BUILD]
+set -x
+set -e
+cd charm-6.10.2
+#charm_arch='netlrts-linux-arm8-flang-smp-clang'
+#./build charm++ netlrts-linux-arm8 smp -j32 clang flang --with-production
+charm_arch='multicore-arm8-flang-clang'
+./build charm++ multicore-arm8 -j32 clang flang --with-production
+cd $charm_arch/tests/charm++/megatest
+make pgm
+./pgm +p4
+cd ../../../../..
+rm -rf Linux-ARM64-g++
+./config Linux-ARM64-g++ --charm-arch $charm_arch --with-cuda --cxx clang++ --cc clang
+cd Linux-ARM64-g++
+make -j
+
+[CLEAN]
+make clean
+
+[RUN]
+run =
+binary = namd2 +p126 +setcpuaffinity +maffinity +isomalloc_sync +devices 0,1 stmv_nve_cuda.namd
+nodes = 1
+
+[PERF]
+kperf = --spe --uncore
+perf =
+nsys = -y 5s -d 60s
+ncu = --target-processes all --launch-skip 16 --launch-count 1
\ No newline at end of file