3.3.12.2.1 DSP Multiply Instructions

The DSP instructions that use the multiplier are summarized in Table 3-4.

Table 3-4. DSP Instructions that Use the Multiplier
DSP Instruction⁽¹⁾	Description	Algebraic Equivalent
MAC	Multiply and Add to Accumulator or Square and Add to Accumulator	a = a + b c* or a = a + b²
MSC	Multiply and Subtract from Accumulator	a = a – b c*
MPY	Multiply	a = b c*
MPY.N	Multiply and Negate Result	a = -b c*
SQR	Square to Accumulator	a = b ^ 2
SQRAC	Square and Accumulate	a = a + (b²)
ED	Partial Euclidean Distance	a = (b – c)²
EDAC	Add Partial Euclidean Distance to the Accumulator	a = a + (b – c)²
Note: DSP instructions using the multiplier can operate in Fractional (1.15/1.31) or Integer modes.

The DSP Multiplier Unsigned/Signed Control (US[1:0]) bits (CORCON[13:12]) determine whether the DSP multiply instructions are signed (default), unsigned or mixed-sign. The US[1:0] bits do not influence the MCU multiply instructions, which have specific instructions for a signed or unsigned operation. If the USx bits are set to ‘01’, the input operands for instructions shown in Table 3-4 are considered as unsigned values, which are always zero-extended into the 33th bit of the multiplier value. If the USx bits are set to ‘00’, the operands are sign-extended.

If the USx bits (CORCON[13:12]) are set to ‘10’, the operands for the instructions listed in Table 3-4 are considered as unsigned values. The result is zero-extended prior to any operation with the accumulator (which will always effectively be signed).