.version 1.4
	.target sm_20, map_f64_to_f32
	// compiled with C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v4.0\\bin/../open64/lib//be.exe
	// nvopencc 4.0 built on 2011-02-21

	//-----------------------------------------------------------
	// Compiling C:/Users/EYOUNG~1.COM/AppData/Local/Temp/tmpxft_000014c0_00000000-11_matrixMul_kernel.cpp3.i (C:/Users/EYOUNG~1.COM/AppData/Local/Temp/ccBI#.a04000)
	//-----------------------------------------------------------

	//-----------------------------------------------------------
	// Options:
	//-----------------------------------------------------------
	//  Target:ptx, ISA:sm_20, Endian:little, Pointer Size:32
	//  -O3	(Optimization level)
	//  -g0	(Debug level)
	//  -m2	(Report advisories)
	//-----------------------------------------------------------

	.shared .align 4 .b8 __cuda_local_var_87382_38_non_const_As__6[1024];
	.shared .align 4 .b8 __cuda_local_var_87386_38_non_const_Bs__7[1024];

	.entry matrixMul_bs16_32bit (
		.param .u32 __cudaparm_matrixMul_bs16_32bit_C,
		.param .u32 __cudaparm_matrixMul_bs16_32bit_A,
		.param .u32 __cudaparm_matrixMul_bs16_32bit_B,
		.param .s32 __cudaparm_matrixMul_bs16_32bit_wA,
		.param .s32 __cudaparm_matrixMul_bs16_32bit_wB)
	{
	.reg .u32 %r<57>;
	.reg .f32 %f<52>;
	.reg .pred %p<4>;
	.loc	16	107	0
$LDWbegin_matrixMul_bs16_32bit:
	.loc	16	66	0
	cvt.s32.u16 	%r1, %ctaid.x;
	mul24.lo.s32 	%r2, %r1, 16;
	cvt.s32.u16 	%r3, %ctaid.y;
	ld.param.s32 	%r4, [__cudaparm_matrixMul_bs16_32bit_wA];
	mul.lo.s32 	%r5, %r3, %r4;
	mul.lo.s32 	%r6, %r5, 16;
	add.s32 	%r7, %r6, %r4;
	sub.s32 	%r8, %r7, 1;
	cvt.s32.u16 	%r9, %tid.x;
	cvt.s32.u16 	%r10, %tid.y;
	ld.param.s32 	%r11, [__cudaparm_matrixMul_bs16_32bit_wB];
	setp.lt.s32 	%p1, %r8, %r6;
	@%p1 bra 	$Lt_0_3330;
	mov.u32 	%r12, __cuda_local_var_87382_38_non_const_As__6;
	mov.u32 	%r13, __cuda_local_var_87386_38_non_const_Bs__7;
	ld.param.s32 	%r4, [__cudaparm_matrixMul_bs16_32bit_wA];
	add.s32 	%r14, %r4, 15;
	shr.s32 	%r15, %r14, 31;
	mov.s32 	%r16, 15;
	and.b32 	%r17, %r15, %r16;
	add.s32 	%r18, %r17, %r14;
	shr.s32 	%r19, %r18, 4;
	ld.param.s32 	%r11, [__cudaparm_matrixMul_bs16_32bit_wB];
	mul.lo.s32 	%r20, %r10, %r11;
	mul.lo.s32 	%r21, %r10, %r4;
	mul24.lo.u32 	%r22, %r10, 16;
	mul24.lo.u32 	%r23, %r10, 64;
	mul24.lo.u32 	%r24, %r9, 4;
	mul.lo.s32 	%r25, %r11, 16;
	add.s32 	%r26, %r20, %r2;
	add.s32 	%r27, %r21, %r6;
	add.u32 	%r28, %r9, %r22;
	add.u32 	%r29, %r23, %r12;
	add.u32 	%r30, %r24, %r13;
	add.s32 	%r31, %r26, %r9;
	add.s32 	%r32, %r27, %r9;
	mul.lo.u32 	%r33, %r28, 4;
	mul.lo.u32 	%r34, %r31, 4;
	mul.lo.u32 	%r35, %r25, 4;
	mul.lo.u32 	%r36, %r32, 4;
	add.u32 	%r37, %r33, %r12;
	add.u32 	%r38, %r33, %r13;
	add.s32 	%r39, %r21, %r8;
	ld.param.u32 	%r40, [__cudaparm_matrixMul_bs16_32bit_B];
	add.u32 	%r41, %r40, %r34;
	ld.param.u32 	%r42, [__cudaparm_matrixMul_bs16_32bit_A];
	add.u32 	%r43, %r36, %r42;
	add.s32 	%r44, %r39, %r9;
	mul.lo.u32 	%r45, %r44, 4;
	add.u32 	%r46, %r45, %r42;
	mov.f32 	%f1, 0f00000000;     	// 0
	mov.s32 	%r47, %r19;
$Lt_0_2818:
 //<loop> Loop body line 66, nesting depth: 1, estimated iterations: unknown
	.loc	16	81	0
	ld.global.f32 	%f2, [%r43+0];
	st.shared.f32 	[%r37+0], %f2;
	.loc	16	82	0
	ld.global.f32 	%f3, [%r41+0];
	st.shared.f32 	[%r38+0], %f3;
	.loc	16	85	0
	bar.sync 	0;
	.loc	16	92	0
	ld.shared.f32 	%f4, [%r29+0];
	ld.shared.f32 	%f5, [%r30+0];
	mad.f32 	%f6, %f4, %f5, %f1;
	ld.shared.f32 	%f7, [%r29+4];
	ld.shared.f32 	%f8, [%r30+64];
	mad.f32 	%f9, %f7, %f8, %f6;
	ld.shared.f32 	%f10, [%r29+8];
	ld.shared.f32 	%f11, [%r30+128];
	mad.f32 	%f12, %f10, %f11, %f9;
	ld.shared.f32 	%f13, [%r29+12];
	ld.shared.f32 	%f14, [%r30+192];
	mad.f32 	%f15, %f13, %f14, %f12;
	ld.shared.f32 	%f16, [%r29+16];
	ld.shared.f32 	%f17, [%r30+256];
	mad.f32 	%f18, %f16, %f17, %f15;
	ld.shared.f32 	%f19, [%r29+20];
	ld.shared.f32 	%f20, [%r30+320];
	mad.f32 	%f21, %f19, %f20, %f18;
	ld.shared.f32 	%f22, [%r29+24];
	ld.shared.f32 	%f23, [%r30+384];
	mad.f32 	%f24, %f22, %f23, %f21;
	ld.shared.f32 	%f25, [%r29+28];
	ld.shared.f32 	%f26, [%r30+448];
	mad.f32 	%f27, %f25, %f26, %f24;
	ld.shared.f32 	%f28, [%r29+32];
	ld.shared.f32 	%f29, [%r30+512];
	mad.f32 	%f30, %f28, %f29, %f27;
	ld.shared.f32 	%f31, [%r29+36];
	ld.shared.f32 	%f32, [%r30+576];
	mad.f32 	%f33, %f31, %f32, %f30;
	ld.shared.f32 	%f34, [%r29+40];
	ld.shared.f32 	%f35, [%r30+640];
	mad.f32 	%f36, %f34, %f35, %f33;
	ld.shared.f32 	%f37, [%r29+44];
	ld.shared.f32 	%f38, [%r30+704];
	mad.f32 	%f39, %f37, %f38, %f36;
	ld.shared.f32 	%f40, [%r29+48];
	ld.shared.f32 	%f41, [%r30+768];
	mad.f32 	%f42, %f40, %f41, %f39;
	ld.shared.f32 	%f43, [%r29+52];
	ld.shared.f32 	%f44, [%r30+832];
	mad.f32 	%f45, %f43, %f44, %f42;
	ld.shared.f32 	%f46, [%r29+56];
	ld.shared.f32 	%f47, [%r30+896];
	mad.f32 	%f48, %f46, %f47, %f45;
	ld.shared.f32 	%f49, [%r29+60];
	ld.shared.f32 	%f50, [%r30+960];
	mad.f32 	%f1, %f49, %f50, %f48;
	.loc	16	97	0
	bar.sync 	0;
	.loc	16	66	0
	add.u32 	%r41, %r35, %r41;
	add.u32 	%r43, %r43, 64;
	setp.le.u32 	%p2, %r43, %r46;
	@%p2 bra 	$Lt_0_2818;
	bra.uni 	$Lt_0_2306;
$Lt_0_3330:
	ld.param.s32 	%r11, [__cudaparm_matrixMul_bs16_32bit_wB];
	mul.lo.s32 	%r20, %r10, %r11;
	mov.f32 	%f1, 0f00000000;     	// 0
$Lt_0_2306:
	.loc	16	103	0
	ld.param.u32 	%r48, [__cudaparm_matrixMul_bs16_32bit_C];
	mul.lo.s32 	%r49, %r11, %r3;
	add.s32 	%r50, %r1, %r49;
	mul.lo.s32 	%r51, %r50, 16;
	add.s32 	%r52, %r20, %r51;
	add.s32 	%r53, %r9, %r52;
	mul.lo.u32 	%r54, %r53, 4;
	add.u32 	%r55, %r48, %r54;
	st.global.f32 	[%r55+0], %f1;
	.loc	16	110	0
	exit;
$LDWend_matrixMul_bs16_32bit:
	} // matrixMul_bs16_32bit
	.shared .align 4 .b8 __cuda_local_var_87382_38_non_const_As__4[1024];
	.shared .align 4 .b8 __cuda_local_var_87386_38_non_const_Bs__5[1024];

	.entry matrixMul_bs16_64bit (
		.param .u32 __cudaparm_matrixMul_bs16_64bit_C,
		.param .u32 __cudaparm_matrixMul_bs16_64bit_A,
		.param .u32 __cudaparm_matrixMul_bs16_64bit_B,
		.param .u32 __cudaparm_matrixMul_bs16_64bit_wA,
		.param .u32 __cudaparm_matrixMul_bs16_64bit_wB)
	{
	.reg .u32 %r<54>;
	.reg .f32 %f<52>;
	.reg .pred %p<4>;
	.loc	16	111	0
$LDWbegin_matrixMul_bs16_64bit:
	.loc	16	66	0
	cvt.u32.u16 	%r1, %ctaid.x;
	mul24.lo.u32 	%r2, %r1, 16;
	cvt.u32.u16 	%r3, %ctaid.y;
	ld.param.u32 	%r4, [__cudaparm_matrixMul_bs16_64bit_wA];
	mul.lo.u32 	%r5, %r3, %r4;
	mul.lo.u32 	%r6, %r5, 16;
	mov.s32 	%r7, %r6;
	add.u32 	%r8, %r6, %r4;
	sub.u32 	%r9, %r8, 1;
	cvt.u32.u16 	%r10, %tid.x;
	cvt.u32.u16 	%r11, %tid.y;
	ld.param.u32 	%r12, [__cudaparm_matrixMul_bs16_64bit_wB];
	setp.lt.u32 	%p1, %r9, %r6;
	@%p1 bra 	$Lt_1_3330;
	mov.u32 	%r13, __cuda_local_var_87382_38_non_const_As__4;
	mov.u32 	%r14, __cuda_local_var_87386_38_non_const_Bs__5;
	ld.param.u32 	%r4, [__cudaparm_matrixMul_bs16_64bit_wA];
	add.u32 	%r15, %r4, 15;
	shr.s32 	%r16, %r15, 31;
	mov.s32 	%r17, 15;
	and.b32 	%r18, %r16, %r17;
	add.s32 	%r19, %r18, %r15;
	shr.s32 	%r20, %r19, 4;
	ld.param.u32 	%r12, [__cudaparm_matrixMul_bs16_64bit_wB];
	mul.lo.u32 	%r21, %r11, %r12;
	mul.lo.u32 	%r22, %r11, %r4;
	mul24.lo.u32 	%r23, %r11, 16;
	mul24.lo.u32 	%r24, %r11, 64;
	mul24.lo.u32 	%r25, %r10, 4;
	mul.lo.u32 	%r26, %r12, 16;
	add.u32 	%r27, %r21, %r2;
	add.u32 	%r28, %r22, %r6;
	add.u32 	%r29, %r23, %r10;
	add.u32 	%r30, %r24, %r13;
	add.u32 	%r31, %r25, %r14;
	add.u32 	%r32, %r27, %r10;
	add.u32 	%r33, %r28, %r10;
	mul.lo.u32 	%r34, %r29, 4;
	mul.lo.u32 	%r35, %r32, 4;
	mul.lo.u32 	%r36, %r26, 4;
	mul.lo.u32 	%r37, %r33, 4;
	add.u32 	%r38, %r34, %r13;
	add.u32 	%r39, %r34, %r14;
	ld.param.u32 	%r40, [__cudaparm_matrixMul_bs16_64bit_B];
	add.u32 	%r41, %r40, %r35;
	ld.param.u32 	%r42, [__cudaparm_matrixMul_bs16_64bit_A];
	add.u32 	%r43, %r42, %r37;
	mov.f32 	%f1, 0f00000000;     	// 0
	mov.s32 	%r44, %r20;
$Lt_1_2818:
 //<loop> Loop body line 66, nesting depth: 1, estimated iterations: unknown
	.loc	16	81	0
	ld.global.f32 	%f2, [%r43+0];
	st.shared.f32 	[%r38+0], %f2;
	.loc	16	82	0
	ld.global.f32 	%f3, [%r41+0];
	st.shared.f32 	[%r39+0], %f3;
	.loc	16	85	0
	bar.sync 	0;
	.loc	16	92	0
	ld.shared.f32 	%f4, [%r30+0];
	ld.shared.f32 	%f5, [%r31+0];
	mad.f32 	%f6, %f4, %f5, %f1;
	ld.shared.f32 	%f7, [%r30+4];
	ld.shared.f32 	%f8, [%r31+64];
	mad.f32 	%f9, %f7, %f8, %f6;
	ld.shared.f32 	%f10, [%r30+8];
	ld.shared.f32 	%f11, [%r31+128];
	mad.f32 	%f12, %f10, %f11, %f9;
	ld.shared.f32 	%f13, [%r30+12];
	ld.shared.f32 	%f14, [%r31+192];
	mad.f32 	%f15, %f13, %f14, %f12;
	ld.shared.f32 	%f16, [%r30+16];
	ld.shared.f32 	%f17, [%r31+256];
	mad.f32 	%f18, %f16, %f17, %f15;
	ld.shared.f32 	%f19, [%r30+20];
	ld.shared.f32 	%f20, [%r31+320];
	mad.f32 	%f21, %f19, %f20, %f18;
	ld.shared.f32 	%f22, [%r30+24];
	ld.shared.f32 	%f23, [%r31+384];
	mad.f32 	%f24, %f22, %f23, %f21;
	ld.shared.f32 	%f25, [%r30+28];
	ld.shared.f32 	%f26, [%r31+448];
	mad.f32 	%f27, %f25, %f26, %f24;
	ld.shared.f32 	%f28, [%r30+32];
	ld.shared.f32 	%f29, [%r31+512];
	mad.f32 	%f30, %f28, %f29, %f27;
	ld.shared.f32 	%f31, [%r30+36];
	ld.shared.f32 	%f32, [%r31+576];
	mad.f32 	%f33, %f31, %f32, %f30;
	ld.shared.f32 	%f34, [%r30+40];
	ld.shared.f32 	%f35, [%r31+640];
	mad.f32 	%f36, %f34, %f35, %f33;
	ld.shared.f32 	%f37, [%r30+44];
	ld.shared.f32 	%f38, [%r31+704];
	mad.f32 	%f39, %f37, %f38, %f36;
	ld.shared.f32 	%f40, [%r30+48];
	ld.shared.f32 	%f41, [%r31+768];
	mad.f32 	%f42, %f40, %f41, %f39;
	ld.shared.f32 	%f43, [%r30+52];
	ld.shared.f32 	%f44, [%r31+832];
	mad.f32 	%f45, %f43, %f44, %f42;
	ld.shared.f32 	%f46, [%r30+56];
	ld.shared.f32 	%f47, [%r31+896];
	mad.f32 	%f48, %f46, %f47, %f45;
	ld.shared.f32 	%f49, [%r30+60];
	ld.shared.f32 	%f50, [%r31+960];
	mad.f32 	%f1, %f49, %f50, %f48;
	.loc	16	97	0
	bar.sync 	0;
	.loc	16	66	0
	add.u32 	%r41, %r36, %r41;
	add.u32 	%r7, %r7, 16;
	add.u32 	%r43, %r43, 64;
	setp.ge.u32 	%p2, %r9, %r7;
	@%p2 bra 	$Lt_1_2818;
	bra.uni 	$Lt_1_2306;
$Lt_1_3330:
	ld.param.u32 	%r12, [__cudaparm_matrixMul_bs16_64bit_wB];
	mul.lo.u32 	%r21, %r11, %r12;
	mov.f32 	%f1, 0f00000000;     	// 0
$Lt_1_2306:
	.loc	16	103	0
	ld.param.u32 	%r45, [__cudaparm_matrixMul_bs16_64bit_C];
	mul.lo.u32 	%r46, %r12, %r3;
	add.u32 	%r47, %r1, %r46;
	mul.lo.u32 	%r48, %r47, 16;
	add.u32 	%r49, %r21, %r48;
	add.u32 	%r50, %r10, %r49;
	mul.lo.u32 	%r51, %r50, 4;
	add.u32 	%r52, %r45, %r51;
	st.global.f32 	[%r52+0], %f1;
	.loc	16	114	0
	exit;
$LDWend_matrixMul_bs16_64bit:
	} // matrixMul_bs16_64bit
	.shared .align 4 .b8 __cuda_local_var_87382_38_non_const_As__2[4096];
	.shared .align 4 .b8 __cuda_local_var_87386_38_non_const_Bs__3[4096];

	.entry matrixMul_bs32_32bit (
		.param .u32 __cudaparm_matrixMul_bs32_32bit_C,
		.param .u32 __cudaparm_matrixMul_bs32_32bit_A,
		.param .u32 __cudaparm_matrixMul_bs32_32bit_B,
		.param .s32 __cudaparm_matrixMul_bs32_32bit_wA,
		.param .s32 __cudaparm_matrixMul_bs32_32bit_wB)
	{
	.reg .u32 %r<57>;
	.reg .f32 %f<100>;
	.reg .pred %p<4>;
	.loc	16	115	0
$LDWbegin_matrixMul_bs32_32bit:
	.loc	16	66	0
	cvt.s32.u16 	%r1, %ctaid.x;
	mul24.lo.s32 	%r2, %r1, 32;
	cvt.s32.u16 	%r3, %ctaid.y;
	ld.param.s32 	%r4, [__cudaparm_matrixMul_bs32_32bit_wA];
	mul.lo.s32 	%r5, %r3, %r4;
	mul.lo.s32 	%r6, %r5, 32;
	add.s32 	%r7, %r6, %r4;
	sub.s32 	%r8, %r7, 1;
	cvt.s32.u16 	%r9, %tid.x;
	cvt.s32.u16 	%r10, %tid.y;
	ld.param.s32 	%r11, [__cudaparm_matrixMul_bs32_32bit_wB];
	setp.lt.s32 	%p1, %r8, %r6;
	@%p1 bra 	$Lt_2_3330;
	mov.u32 	%r12, __cuda_local_var_87382_38_non_const_As__2;
	mov.u32 	%r13, __cuda_local_var_87386_38_non_const_Bs__3;
	ld.param.s32 	%r4, [__cudaparm_matrixMul_bs32_32bit_wA];
	add.s32 	%r14, %r4, 31;
	shr.s32 	%r15, %r14, 31;
	mov.s32 	%r16, 31;
	and.b32 	%r17, %r15, %r16;
	add.s32 	%r18, %r17, %r14;
	shr.s32 	%r19, %r18, 5;
	ld.param.s32 	%r11, [__cudaparm_matrixMul_bs32_32bit_wB];
	mul.lo.s32 	%r20, %r10, %r11;
	mul.lo.s32 	%r21, %r10, %r4;
	mul24.lo.u32 	%r22, %r10, 32;
	mul24.lo.u32 	%r23, %r10, 128;
	mul24.lo.u32 	%r24, %r9, 4;
	mul.lo.s32 	%r25, %r11, 32;
	add.s32 	%r26, %r20, %r2;
	add.s32 	%r27, %r21, %r6;
	add.u32 	%r28, %r9, %r22;
	add.u32 	%r29, %r23, %r12;
	add.u32 	%r30, %r24, %r13;
	add.s32 	%r31, %r26, %r9;
	add.s32 	%r32, %r27, %r9;
	mul.lo.u32 	%r33, %r28, 4;
	mul.lo.u32 	%r34, %r31, 4;
	mul.lo.u32 	%r35, %r25, 4;
	mul.lo.u32 	%r36, %r32, 4;
	add.u32 	%r37, %r33, %r12;
	add.u32 	%r38, %r33, %r13;
	add.s32 	%r39, %r21, %r8;
	ld.param.u32 	%r40, [__cudaparm_matrixMul_bs32_32bit_B];
	add.u32 	%r41, %r40, %r34;
	ld.param.u32 	%r42, [__cudaparm_matrixMul_bs32_32bit_A];
	add.u32 	%r43, %r36, %r42;
	add.s32 	%r44, %r39, %r9;
	mul.lo.u32 	%r45, %r44, 4;
	add.u32 	%r46, %r45, %r42;
	mov.f32 	%f1, 0f00000000;     	// 0
	mov.s32 	%r47, %r19;
$Lt_2_2818:
 //<loop> Loop body line 66, nesting depth: 1, estimated iterations: unknown
	.loc	16	81	0
	ld.global.f32 	%f2, [%r43+0];
	st.shared.f32 	[%r37+0], %f2;
	.loc	16	82	0
	ld.global.f32 	%f3, [%r41+0];
	st.shared.f32 	[%r38+0], %f3;
	.loc	16	85	0
	bar.sync 	0;
	.loc	16	92	0
	ld.shared.f32 	%f4, [%r29+0];
	ld.shared.f32 	%f5, [%r30+0];
	mad.f32 	%f6, %f4, %f5, %f1;
	ld.shared.f32 	%f7, [%r29+4];
	ld.shared.f32 	%f8, [%r30+128];
	mad.f32 	%f9, %f7, %f8, %f6;
	ld.shared.f32 	%f10, [%r29+8];
	ld.shared.f32 	%f11, [%r30+256];
	mad.f32 	%f12, %f10, %f11, %f9;
	ld.shared.f32 	%f13, [%r29+12];
	ld.shared.f32 	%f14, [%r30+384];
	mad.f32 	%f15, %f13, %f14, %f12;
	ld.shared.f32 	%f16, [%r29+16];
	ld.shared.f32 	%f17, [%r30+512];
	mad.f32 	%f18, %f16, %f17, %f15;
	ld.shared.f32 	%f19, [%r29+20];
	ld.shared.f32 	%f20, [%r30+640];
	mad.f32 	%f21, %f19, %f20, %f18;
	ld.shared.f32 	%f22, [%r29+24];
	ld.shared.f32 	%f23, [%r30+768];
	mad.f32 	%f24, %f22, %f23, %f21;
	ld.shared.f32 	%f25, [%r29+28];
	ld.shared.f32 	%f26, [%r30+896];
	mad.f32 	%f27, %f25, %f26, %f24;
	ld.shared.f32 	%f28, [%r29+32];
	ld.shared.f32 	%f29, [%r30+1024];
	mad.f32 	%f30, %f28, %f29, %f27;
	ld.shared.f32 	%f31, [%r29+36];
	ld.shared.f32 	%f32, [%r30+1152];
	mad.f32 	%f33, %f31, %f32, %f30;
	ld.shared.f32 	%f34, [%r29+40];
	ld.shared.f32 	%f35, [%r30+1280];
	mad.f32 	%f36, %f34, %f35, %f33;
	ld.shared.f32 	%f37, [%r29+44];
	ld.shared.f32 	%f38, [%r30+1408];
	mad.f32 	%f39, %f37, %f38, %f36;
	ld.shared.f32 	%f40, [%r29+48];
	ld.shared.f32 	%f41, [%r30+1536];
	mad.f32 	%f42, %f40, %f41, %f39;
	ld.shared.f32 	%f43, [%r29+52];
	ld.shared.f32 	%f44, [%r30+1664];
	mad.f32 	%f45, %f43, %f44, %f42;
	ld.shared.f32 	%f46, [%r29+56];
	ld.shared.f32 	%f47, [%r30+1792];
	mad.f32 	%f48, %f46, %f47, %f45;
	ld.shared.f32 	%f49, [%r29+60];
	ld.shared.f32 	%f50, [%r30+1920];
	mad.f32 	%f51, %f49, %f50, %f48;
	ld.shared.f32 	%f52, [%r29+64];
	ld.shared.f32 	%f53, [%r30+2048];
	mad.f32 	%f54, %f52, %f53, %f51;
	ld.shared.f32 	%f55, [%r29+68];
	ld.shared.f32 	%f56, [%r30+2176];
	mad.f32 	%f57, %f55, %f56, %f54;
	ld.shared.f32 	%f58, [%r29+72];
	ld.shared.f32 	%f59, [%r30+2304];
	mad.f32 	%f60, %f58, %f59, %f57;
	ld.shared.f32 	%f61, [%r29+76];
	ld.shared.f32 	%f62, [%r30+2432];
	mad.f32 	%f63, %f61, %f62, %f60;
	ld.shared.f32 	%f64, [%r29+80];
	ld.shared.f32 	%f65, [%r30+2560];
	mad.f32 	%f66, %f64, %f65, %f63;
	ld.shared.f32 	%f67, [%r29+84];
	ld.shared.f32 	%f68, [%r30+2688];
	mad.f32 	%f69, %f67, %f68, %f66;
	ld.shared.f32 	%f70, [%r29+88];
	ld.shared.f32 	%f71, [%r30+2816];
	mad.f32 	%f72, %f70, %f71, %f69;
	ld.shared.f32 	%f73, [%r29+92];
	ld.shared.f32 	%f74, [%r30+2944];
	mad.f32 	%f75, %f73, %f74, %f72;
	ld.shared.f32 	%f76, [%r29+96];
	ld.shared.f32 	%f77, [%r30+3072];
	mad.f32 	%f78, %f76, %f77, %f75;
	ld.shared.f32 	%f79, [%r29+100];
	ld.shared.f32 	%f80, [%r30+3200];
	mad.f32 	%f81, %f79, %f80, %f78;
	ld.shared.f32 	%f82, [%r29+104];
	ld.shared.f32 	%f83, [%r30+3328];
	mad.f32 	%f84, %f82, %f83, %f81;
	ld.shared.f32 	%f85, [%r29+108];
	ld.shared.f32 	%f86, [%r30+3456];
	mad.f32 	%f87, %f85, %f86, %f84;
	ld.shared.f32 	%f88, [%r29+112];
	ld.shared.f32 	%f89, [%r30+3584];
	mad.f32 	%f90, %f88, %f89, %f87;
	ld.shared.f32 	%f91, [%r29+116];
	ld.shared.f32 	%f92, [%r30+3712];
	mad.f32 	%f93, %f91, %f92, %f90;
	ld.shared.f32 	%f94, [%r29+120];
	ld.shared.f32 	%f95, [%r30+3840];
	mad.f32 	%f96, %f94, %f95, %f93;
	ld.shared.f32 	%f97, [%r29+124];
	ld.shared.f32 	%f98, [%r30+3968];
	mad.f32 	%f1, %f97, %f98, %f96;
	.loc	16	97	0
	bar.sync 	0;
	.loc	16	66	0
	add.u32 	%r41, %r35, %r41;
	add.u32 	%r43, %r43, 128;
	setp.le.u32 	%p2, %r43, %r46;
	@%p2 bra 	$Lt_2_2818;
	bra.uni 	$Lt_2_2306;
$Lt_2_3330:
	ld.param.s32 	%r11, [__cudaparm_matrixMul_bs32_32bit_wB];
	mul.lo.s32 	%r20, %r10, %r11;
	mov.f32 	%f1, 0f00000000;     	// 0
$Lt_2_2306:
	.loc	16	103	0
	ld.param.u32 	%r48, [__cudaparm_matrixMul_bs32_32bit_C];
	mul.lo.s32 	%r49, %r11, %r3;
	add.s32 	%r50, %r1, %r49;
	mul.lo.s32 	%r51, %r50, 32;
	add.s32 	%r52, %r20, %r51;
	add.s32 	%r53, %r9, %r52;
	mul.lo.u32 	%r54, %r53, 4;
	add.u32 	%r55, %r48, %r54;
	st.global.f32 	[%r55+0], %f1;
	.loc	16	118	0
	exit;
$LDWend_matrixMul_bs32_32bit:
	} // matrixMul_bs32_32bit
	.shared .align 4 .b8 __cuda_local_var_87382_38_non_const_As__0[4096];
	.shared .align 4 .b8 __cuda_local_var_87386_38_non_const_Bs__1[4096];

	.entry matrixMul_bs32_64bit (
		.param .u32 __cudaparm_matrixMul_bs32_64bit_C,
		.param .u32 __cudaparm_matrixMul_bs32_64bit_A,
		.param .u32 __cudaparm_matrixMul_bs32_64bit_B,
		.param .u32 __cudaparm_matrixMul_bs32_64bit_wA,
		.param .u32 __cudaparm_matrixMul_bs32_64bit_wB)
	{
	.reg .u32 %r<54>;
	.reg .f32 %f<100>;
	.reg .pred %p<4>;
	.loc	16	119	0
$LDWbegin_matrixMul_bs32_64bit:
	.loc	16	66	0
	cvt.u32.u16 	%r1, %ctaid.x;
	mul24.lo.u32 	%r2, %r1, 32;
	cvt.u32.u16 	%r3, %ctaid.y;
	ld.param.u32 	%r4, [__cudaparm_matrixMul_bs32_64bit_wA];
	mul.lo.u32 	%r5, %r3, %r4;
	mul.lo.u32 	%r6, %r5, 32;
	mov.s32 	%r7, %r6;
	add.u32 	%r8, %r6, %r4;
	sub.u32 	%r9, %r8, 1;
	cvt.u32.u16 	%r10, %tid.x;
	cvt.u32.u16 	%r11, %tid.y;
	ld.param.u32 	%r12, [__cudaparm_matrixMul_bs32_64bit_wB];
	setp.lt.u32 	%p1, %r9, %r6;
	@%p1 bra 	$Lt_3_3330;
	mov.u32 	%r13, __cuda_local_var_87382_38_non_const_As__0;
	mov.u32 	%r14, __cuda_local_var_87386_38_non_const_Bs__1;
	ld.param.u32 	%r4, [__cudaparm_matrixMul_bs32_64bit_wA];
	add.u32 	%r15, %r4, 31;
	shr.s32 	%r16, %r15, 31;
	mov.s32 	%r17, 31;
	and.b32 	%r18, %r16, %r17;
	add.s32 	%r19, %r18, %r15;
	shr.s32 	%r20, %r19, 5;
	ld.param.u32 	%r12, [__cudaparm_matrixMul_bs32_64bit_wB];
	mul.lo.u32 	%r21, %r11, %r12;
	mul.lo.u32 	%r22, %r11, %r4;
	mul24.lo.u32 	%r23, %r11, 32;
	mul24.lo.u32 	%r24, %r11, 128;
	mul24.lo.u32 	%r25, %r10, 4;
	mul.lo.u32 	%r26, %r12, 32;
	add.u32 	%r27, %r21, %r2;
	add.u32 	%r28, %r22, %r6;
	add.u32 	%r29, %r23, %r10;
	add.u32 	%r30, %r24, %r13;
	add.u32 	%r31, %r25, %r14;
	add.u32 	%r32, %r27, %r10;
	add.u32 	%r33, %r28, %r10;
	mul.lo.u32 	%r34, %r29, 4;
	mul.lo.u32 	%r35, %r32, 4;
	mul.lo.u32 	%r36, %r26, 4;
	mul.lo.u32 	%r37, %r33, 4;
	add.u32 	%r38, %r34, %r13;
	add.u32 	%r39, %r34, %r14;
	ld.param.u32 	%r40, [__cudaparm_matrixMul_bs32_64bit_B];
	add.u32 	%r41, %r40, %r35;
	ld.param.u32 	%r42, [__cudaparm_matrixMul_bs32_64bit_A];
	add.u32 	%r43, %r42, %r37;
	mov.f32 	%f1, 0f00000000;     	// 0
	mov.s32 	%r44, %r20;
$Lt_3_2818:
 //<loop> Loop body line 66, nesting depth: 1, estimated iterations: unknown
	.loc	16	81	0
	ld.global.f32 	%f2, [%r43+0];
	st.shared.f32 	[%r38+0], %f2;
	.loc	16	82	0
	ld.global.f32 	%f3, [%r41+0];
	st.shared.f32 	[%r39+0], %f3;
	.loc	16	85	0
	bar.sync 	0;
	.loc	16	92	0
	ld.shared.f32 	%f4, [%r30+0];
	ld.shared.f32 	%f5, [%r31+0];
	mad.f32 	%f6, %f4, %f5, %f1;
	ld.shared.f32 	%f7, [%r30+4];
	ld.shared.f32 	%f8, [%r31+128];
	mad.f32 	%f9, %f7, %f8, %f6;
	ld.shared.f32 	%f10, [%r30+8];
	ld.shared.f32 	%f11, [%r31+256];
	mad.f32 	%f12, %f10, %f11, %f9;
	ld.shared.f32 	%f13, [%r30+12];
	ld.shared.f32 	%f14, [%r31+384];
	mad.f32 	%f15, %f13, %f14, %f12;
	ld.shared.f32 	%f16, [%r30+16];
	ld.shared.f32 	%f17, [%r31+512];
	mad.f32 	%f18, %f16, %f17, %f15;
	ld.shared.f32 	%f19, [%r30+20];
	ld.shared.f32 	%f20, [%r31+640];
	mad.f32 	%f21, %f19, %f20, %f18;
	ld.shared.f32 	%f22, [%r30+24];
	ld.shared.f32 	%f23, [%r31+768];
	mad.f32 	%f24, %f22, %f23, %f21;
	ld.shared.f32 	%f25, [%r30+28];
	ld.shared.f32 	%f26, [%r31+896];
	mad.f32 	%f27, %f25, %f26, %f24;
	ld.shared.f32 	%f28, [%r30+32];
	ld.shared.f32 	%f29, [%r31+1024];
	mad.f32 	%f30, %f28, %f29, %f27;
	ld.shared.f32 	%f31, [%r30+36];
	ld.shared.f32 	%f32, [%r31+1152];
	mad.f32 	%f33, %f31, %f32, %f30;
	ld.shared.f32 	%f34, [%r30+40];
	ld.shared.f32 	%f35, [%r31+1280];
	mad.f32 	%f36, %f34, %f35, %f33;
	ld.shared.f32 	%f37, [%r30+44];
	ld.shared.f32 	%f38, [%r31+1408];
	mad.f32 	%f39, %f37, %f38, %f36;
	ld.shared.f32 	%f40, [%r30+48];
	ld.shared.f32 	%f41, [%r31+1536];
	mad.f32 	%f42, %f40, %f41, %f39;
	ld.shared.f32 	%f43, [%r30+52];
	ld.shared.f32 	%f44, [%r31+1664];
	mad.f32 	%f45, %f43, %f44, %f42;
	ld.shared.f32 	%f46, [%r30+56];
	ld.shared.f32 	%f47, [%r31+1792];
	mad.f32 	%f48, %f46, %f47, %f45;
	ld.shared.f32 	%f49, [%r30+60];
	ld.shared.f32 	%f50, [%r31+1920];
	mad.f32 	%f51, %f49, %f50, %f48;
	ld.shared.f32 	%f52, [%r30+64];
	ld.shared.f32 	%f53, [%r31+2048];
	mad.f32 	%f54, %f52, %f53, %f51;
	ld.shared.f32 	%f55, [%r30+68];
	ld.shared.f32 	%f56, [%r31+2176];
	mad.f32 	%f57, %f55, %f56, %f54;
	ld.shared.f32 	%f58, [%r30+72];
	ld.shared.f32 	%f59, [%r31+2304];
	mad.f32 	%f60, %f58, %f59, %f57;
	ld.shared.f32 	%f61, [%r30+76];
	ld.shared.f32 	%f62, [%r31+2432];
	mad.f32 	%f63, %f61, %f62, %f60;
	ld.shared.f32 	%f64, [%r30+80];
	ld.shared.f32 	%f65, [%r31+2560];
	mad.f32 	%f66, %f64, %f65, %f63;
	ld.shared.f32 	%f67, [%r30+84];
	ld.shared.f32 	%f68, [%r31+2688];
	mad.f32 	%f69, %f67, %f68, %f66;
	ld.shared.f32 	%f70, [%r30+88];
	ld.shared.f32 	%f71, [%r31+2816];
	mad.f32 	%f72, %f70, %f71, %f69;
	ld.shared.f32 	%f73, [%r30+92];
	ld.shared.f32 	%f74, [%r31+2944];
	mad.f32 	%f75, %f73, %f74, %f72;
	ld.shared.f32 	%f76, [%r30+96];
	ld.shared.f32 	%f77, [%r31+3072];
	mad.f32 	%f78, %f76, %f77, %f75;
	ld.shared.f32 	%f79, [%r30+100];
	ld.shared.f32 	%f80, [%r31+3200];
	mad.f32 	%f81, %f79, %f80, %f78;
	ld.shared.f32 	%f82, [%r30+104];
	ld.shared.f32 	%f83, [%r31+3328];
	mad.f32 	%f84, %f82, %f83, %f81;
	ld.shared.f32 	%f85, [%r30+108];
	ld.shared.f32 	%f86, [%r31+3456];
	mad.f32 	%f87, %f85, %f86, %f84;
	ld.shared.f32 	%f88, [%r30+112];
	ld.shared.f32 	%f89, [%r31+3584];
	mad.f32 	%f90, %f88, %f89, %f87;
	ld.shared.f32 	%f91, [%r30+116];
	ld.shared.f32 	%f92, [%r31+3712];
	mad.f32 	%f93, %f91, %f92, %f90;
	ld.shared.f32 	%f94, [%r30+120];
	ld.shared.f32 	%f95, [%r31+3840];
	mad.f32 	%f96, %f94, %f95, %f93;
	ld.shared.f32 	%f97, [%r30+124];
	ld.shared.f32 	%f98, [%r31+3968];
	mad.f32 	%f1, %f97, %f98, %f96;
	.loc	16	97	0
	bar.sync 	0;
	.loc	16	66	0
	add.u32 	%r41, %r36, %r41;
	add.u32 	%r7, %r7, 32;
	add.u32 	%r43, %r43, 128;
	setp.ge.u32 	%p2, %r9, %r7;
	@%p2 bra 	$Lt_3_2818;
	bra.uni 	$Lt_3_2306;
$Lt_3_3330:
	ld.param.u32 	%r12, [__cudaparm_matrixMul_bs32_64bit_wB];
	mul.lo.u32 	%r21, %r11, %r12;
	mov.f32 	%f1, 0f00000000;     	// 0
$Lt_3_2306:
	.loc	16	103	0
	ld.param.u32 	%r45, [__cudaparm_matrixMul_bs32_64bit_C];
	mul.lo.u32 	%r46, %r12, %r3;
	add.u32 	%r47, %r1, %r46;
	mul.lo.u32 	%r48, %r47, 32;
	add.u32 	%r49, %r21, %r48;
	add.u32 	%r50, %r10, %r49;
	mul.lo.u32 	%r51, %r50, 4;
	add.u32 	%r52, %r45, %r51;
	st.global.f32 	[%r52+0], %f1;
	.loc	16	122	0
	exit;
$LDWend_matrixMul_bs32_64bit:
	} // matrixMul_bs32_64bit