x86 Instruction Set Reference

PAVGB/PAVGW - Average Packed Integers

Opcode Mnemonic Description
0F E0 /r PAVGB mm1, mm2/m64 Average packed unsigned byte integers from mm2/m64 and mm1 with rounding.
66 0F E0, /r PAVGB xmm1, xmm2/m128 Average packed unsigned byte integers from xmm2/m128 and xmm1 with rounding.
0F E3 /r PAVGW mm1, mm2/m64 Average packed unsigned word integers from mm2/m64 and mm1 with rounding.
66 0F E3 /r PAVGW xmm1, xmm2/m128 Average packed unsigned word integers from xmm2/m128 and xmm1 with rounding.
Description

Performs an SIMD average of the packed unsigned integers from the source operand (second operand) and the destination operand (first operand), and stores the results in the destination operand. For each corresponding pair of data elements in the first and second operands, the elements are added together, a 1 is added to the temporary sum, and that result is shifted right one bit position. The source operand can be an MMX technology register or a 64-bit memory location or it can be an XMM register or a 128-bit memory location. The destination operand can be an MMX technology register or an XMM register.

The PAVGB instruction operates on packed unsigned bytes and the PAVGW instruction operates on packed unsigned words.
Operation 
switch(Instruction) {
	case PAVGB:
		if(OperandSize == 64) {
			//PAVGB instruction with 64-bit operands:
			//temporary sum before shifting is 9 bits
			Source[0..7] = (Source[0..7] + Destination[0..7] + 1) >> 1;
			Source[8..15] = (Source[8..15] + Destination[8..15] + 1) >> 1;
			Source[16..23] = (Source[16..23] + Destination[16..23] + 1) >> 1;
			Source[24..31] = (Source[24..31] + Destination[24..31] + 1) >> 1;
			Source[32..39] = (Source[32..39] + Destination[32..39] + 1) >> 1;
			Source[40..47] = (Source[40..47] + Destination[40..47] + 1) >> 1;
			Source[48..55] = (Source[48..55] + Destination[48..55] + 1) >> 1;
			Source[56..63] = (Source[56..63] + Destination[56..63] + 1) >> 1;
		}
		else {
			//PAVGB instruction with 128-bit operands:
			//temporary sum before shifting is 9 bits
			Source[0..7] = (Source[0..7] + Destination[0..7] + 1) >> 1;
			Source[8..15] = (Source[8..15] + Destination[8..15] + 1) >> 1;
			Source[16..23] = (Source[16..23] + Destination[16..23] + 1) >> 1;
			Source[24..31] = (Source[24..31] + Destination[24..31] + 1) >> 1;
			Source[32..39] = (Source[32..39] + Destination[32..39] + 1) >> 1;
			Source[40..47] = (Source[40..47] + Destination[40..47] + 1) >> 1;
			Source[48..55] = (Source[48..55] + Destination[48..55] + 1) >> 1;
			Source[56..63] = (Source[56..63] + Destination[56..63] + 1) >> 1;
			Source[64..71] = (Source[64..71] + Destination[64..71] + 1) >> 1;
			Source[72..79] = (Source[72..79] + Destination[72..79] + 1) >> 1;
			Source[80..87] = (Source[80..87] + Destination[80..87] + 1) >> 1;
			Source[88..95] = (Source[88..95] + Destination[88..95] + 1) >> 1;
			Source[96..103] = (Source[96..103] + Destination[96..103] + 1) >> 1;
			Source[104..111] = (Source[104..111] + Destination[104..111] + 1) >> 1;
			Source[112..119] = (Source[112..119] + Destination[112..119] + 1) >> 1;
			Source[120..127] = (Source[120..127] + Destination[120..127] + 1) >> 1;
		}
		break;
	case PAVGW:
		if(OperandSize == 64) {
			//PAVGW instruction with 64-bit operands:
			//temporary sum before shifting is 17 bits
			Source[0..15] = (Source[0..15] + Destination[0..15] + 1) >> 1;
			Source[16..31] = (Source[16..31] + Destination[16..31] + 1) >> 1;
			Source[32..47] = (Source[32..47] + Destination[32..47] + 1) >> 1;
			Source[48..63] = (Source[48..63] + Destination[48..63] + 1) >> 1;
		}
		else {
			//PAVGW instruction with 128-bit operands:
			//temporary sum before shifting is 17 bits
			Source[0..15] = (Source[0..15] + Destination[0..15] + 1) >> 1;
			Source[16..31] = (Source[16..31] + Destination[16..31] + 1) >> 1;
			Source[32..47] = (Source[32..47] + Destination[32..47] + 1) >> 1;
			Source[48..63] = (Source[48..63] + Destination[48..63] + 1) >> 1;
			Source[64..79] = (Source[64..79] + Destination[64..79] + 1) >> 1;
			Source[80..95] = (Source[80..95] + Destination[80..95] + 1) >> 1;
			Source[96..111] = (Source[96..111] + Destination[96..111] + 1) >> 1;
			Source[112..127] = (Source[112..127] + Destination[112..127] + 1) >> 1;
		}
		break;
}
Flags affected

None.

Protected Mode Exceptions
#GP(0)If a memory operand effective address is outside the CS, DS, ES, FS, or GS segment limit. (128-bit operations only) If a memory operand is not aligned on a 16-byte boundary, regardless of segment.
#GP(0)If a memory operand effective address is outside the CS, DS, ES, FS, or GS segment limit. (128-bit operations only) If a memory operand is not aligned on a 16-byte boundary, regardless of segment.
#SS(0)If a memory operand effective address is outside the SS segment limit.
#UDIf EM in CR0 is set. (128-bit operations only) If OSFXSR in CR4 is 0. (128-bit operations only) If CPUID feature flag SSE2 is 0.
#NMIf TS in CR0 is set.
#MF(64-bit operations only) If there is a pending x87 FPU exception.
#PF(fault-code)If a page fault occurs.
Real-Address Mode Exceptions
#GP(0)(128-bit operations only) If a memory operand is not aligned on a 16-byte boundary, regardless of segment. If any part of the operand lies outside of the effective address space from 0 to FFFFH.
#GP(0)(128-bit operations only) If a memory operand is not aligned on a 16-byte boundary, regardless of segment. If any part of the operand lies outside of the effective address space from 0 to FFFFH.
#UDIf EM in CR0 is set. (128-bit operations only) If OSFXSR in CR4 is 0. (128-bit operations only) If CPUID feature flag SSE2 is 0.
#NMIf TS in CR0 is set.
Virtual-8086 Mode Exceptions
Same exceptions as in Real Address Mode
#PF(fault-code)For a page fault.
#PF(fault-code)For a page fault.
Numeric Exceptions
None.
Instruction Latency Throughput Execution Unit
CPUID0F3n/0F2n/069n0F3n/0F2n/069n0F2n
PAVGB/PAVGW mm, mm2/2/-1/1/-MMX_ALU
PAVGB/PAVGW xmm, xmm2/2/-2/2/-MMX_ALU