2 Computational Units
2 – 4
2.1.8 Summary
Table 2.1 summarizes some of the arithmetic characteristics of ADSP-2100
family operations. In addition to the numeric types described in this
section, the ADSP-2100 Family C Compiler supports a form of 32-bit
floating-point in which one 16-bit word is the exponent and the other
word is the mantissa. See the ADSP-2100 Family C Tools Manual.
O
PERATION ARITHMETIC FORMATS
ALU Operands Result
Addition Signed or unsigned Interpret flags
Subtraction Signed or unsigned Interpret flags
Logical Operations Binary string same as operands
Division Explicitly signed/unsigned same as operands
ALU Overflow Signed same as operands
ALU Carry Bit 16-bit unsigned same as operands
ALU Saturation Signed same as operands
MAC, Fractional
Multiplication (P) 1.15 Explicitly signed/unsigned 32 bits (2.30)
Multiplication (MR) 1.15 Explicitly signed/unsigned 2.30 shifted to 1.31
Mult / Add 1.15 Explicitly signed/unsigned 2.30 shifted to 1.31
Mult / Subtract 1.15 Explicitly signed/unsigned 2.30 shifted to 1.31
MAC Saturation Signed same as operands
MAC, Integer Mode
Multiplication (P) 1.15 Explicitly signed/unsigned 32 bits (2.30)
Multiplication (MR) 16.0 Explicitly signed/unsigned 32.0 no shift
Mult / Add 16.0 Explicitly signed/unsigned 32.0 no shift
Mult / Subtract 16.0 Explicitly signed/unsigned 32.0 no shift
MAC Saturation Signed same as operands
Shifter
Logical Shift Unsigned / binary string same as operands
Arithmetic Shift Signed same as operands
Exponent Detection Signed same as operands
Table 2.1 Arithmetic Formats
To Next Page
Loading...
Need help?
General
