diff --git a/src/nmplv/Arithmetics/nm/VEC_MulV_AddC__nm64s.asm b/src/nmplv/Arithmetics/nm/VEC_MulV_AddC__nm64s.asm
index 08c1d6be..e745a446 100644
--- a/src/nmplv/Arithmetics/nm/VEC_MulV_AddC__nm64s.asm
+++ b/src/nmplv/Arithmetics/nm/VEC_MulV_AddC__nm64s.asm
@@ -23,7 +23,6 @@
 
 begin ".text_nmplv"
 
-.if .nm64revision == 5;
 
 
 /////////////////////////////////////////////////////////////////////////////////////////
@@ -72,6 +71,5 @@ global _nmppsMul_AddC_64s:label;
 return;
 .wait;
 
-.endif;
 end ".text_nmplv";
 
diff --git a/src/nmplv/Arithmetics/pc/nmpps-Sub.c b/src/nmplv/Arithmetics/pc/nmpps-Sub.c
index 854cc051..7fef5ce4 100644
--- a/src/nmplv/Arithmetics/pc/nmpps-Sub.c
+++ b/src/nmplv/Arithmetics/pc/nmpps-Sub.c
@@ -22,9 +22,18 @@
 #include "nmtype.h"
 /////////////////////////////////////////////////////////////////////////////////////////
 // Subtraction of two char buffers
-// 2826 ticks per 10240 elements with C++ call (0.276 clocks per element)
-// The order of intermediate data transmissions in function is:
-//		Buf1[],Buf2			=> pDstVec[]				- Subtraction	(in-place is supported)
+void nmppsSub_8s(
+		const nm8s*			pSrcVec1,               // input buffer         :long Local [VecSize/4]
+		const nm8s*			pSrcVec2,               // input buffer         :long Local [VecSize/4]
+		nm8s*				pDstVec,                        // output buffer        :long Global[VecSize/4]
+		int					nSize                   // size of input buffer in 8-bit elements. nSize=[0,8,16,24..]
+               )
+{
+	int i;
+	for (i=0; i<nSize; i++)
+		pDstVec[i] = pSrcVec1[i] - pSrcVec2[i];
+		
+}
 
 //////////////////////////////////////////////////////////////////////////////////////
 // Subtraction of two short vectors
diff --git a/src/nmplv/Select/nm/VEC_CmpEq0__nm16u15b.asm b/src/nmplv/Select/nm/VEC_CmpEq0__nm16u15b.asm
index 7cfeb057..8e7de547 100644
--- a/src/nmplv/Select/nm/VEC_CmpEq0__nm16u15b.asm
+++ b/src/nmplv/Select/nm/VEC_CmpEq0__nm16u15b.asm
@@ -106,6 +106,7 @@ global _nmppsCmpEq0_16u15b:label;
 .repeat 6;
 	rep 32 data = [ar0++gr0], ftw with vsum, activate data, afifo;
 	//WTW_REG_DEPRICATED(gr5);
+	wtw;
 	ar0-=1022;
 .endrepeat;
 	rep 32 wfifo = [ar4++], ftw;
@@ -158,4 +159,4 @@ global _nmppsCmpEq0_16u15b:label;
 end ".text_nmplv";
 //********************************************************************/
 //		END OF FILE BoolConvert16to1.asm                                 */
-//********************************************************************/
\ No newline at end of file
+//********************************************************************/