diff --git a/lib/gpu/Makefile.linux b/lib/gpu/Makefile.linux
index 3fa1dcdce6..7ceecf8a60 100644
--- a/lib/gpu/Makefile.linux
+++ b/lib/gpu/Makefile.linux
@@ -55,9 +55,9 @@ CUDA_PRECISION = -D_SINGLE_DOUBLE
 
 CUDA_INCLUDE = -I$(CUDA_HOME)/include
 CUDA_LIB = -L$(CUDA_HOME)/lib64 -L$(CUDA_HOME)/lib64/stubs
-CUDA_OPTS = -DUNIX -O3 --use_fast_math $(LMP_INC) -Xcompiler -fPIC
+CUDA_OPTS = -DUNIX -O3 --use_fast_math $(LMP_INC) -Xcompiler -fPIC -std=c++11
 
-CUDR_CPP = mpicxx -DMPI_GERYON -DUCL_NO_EXIT -DMPICH_IGNORE_CXX_SEEK -DOMPI_SKIP_MPICXX=1 -fPIC
+CUDR_CPP = mpicxx -DMPI_GERYON -DUCL_NO_EXIT -DMPICH_IGNORE_CXX_SEEK -DOMPI_SKIP_MPICXX=1 -fPIC -std=c++11
 CUDR_OPTS = -O2 $(LMP_INC) # -xHost -no-prec-div -ansi-alias
 
 BIN_DIR = ./
diff --git a/lib/gpu/lal_gauss.cu b/lib/gpu/lal_gauss.cu
index 6a50aa6725..2192fb39ca 100644
--- a/lib/gpu/lal_gauss.cu
+++ b/lib/gpu/lal_gauss.cu
@@ -60,11 +60,9 @@ __kernel void k_gauss(const __global numtyp4 *restrict x_,
     numtyp4 ix; fetch4(ix,i,pos_tex); //x_[i];
     int itype=ix.w;
 
-    numtyp factor_lj;
     for ( ; nbor<nbor_end; nbor+=n_stride) {
 
       int j=dev_packed[nbor];
-      factor_lj = sp_lj[sbmask(j)];
       j &= NEIGHMASK;
 
       numtyp4 jx; fetch4(jx,j,pos_tex); //x_[j];
@@ -148,11 +146,9 @@ __kernel void k_gauss_fast(const __global numtyp4 *restrict x_,
     int iw=ix.w;
     int itype=fast_mul((int)MAX_SHARED_TYPES,iw);
 
-    numtyp factor_lj;
     for ( ; nbor<nbor_end; nbor+=n_stride) {
 
       int j=dev_packed[nbor];
-      factor_lj = sp_lj[sbmask(j)];
       j &= NEIGHMASK;
 
       numtyp4 jx; fetch4(jx,j,pos_tex); //x_[j];