SPRACQ1 May 2020 TMS320F28374D , TMS320F28374S , TMS320F28375D , TMS320F28375S , TMS320F28376D , TMS320F28376S , TMS320F28377D , TMS320F28377S , TMS320F28379D , TMS320F28379S , TMS320F28384D , TMS320F28384D-Q1 , TMS320F28384S , TMS320F28384S-Q1 , TMS320F28386D , TMS320F28386D-Q1 , TMS320F28386S , TMS320F28386S-Q1 , TMS320F28388D , TMS320F28388S
The C28x processor Fast Integer Division (FINTDIV) unit provides an open and scalable approach to facilitate different data type sizes (16/16, 32/16, 32/32, 64/32, 64/64), signed and unsigned or mixed data type versions (ui32/ui32, i32/ui32, i32/i32). For additional performance, the operations return both the integer and remainder portion of the calculation simultaneously. The division operations are interruptible so as to enable minimum latency for higher priority tasks, a critical requirement for high performance real-time control applications. Unique to this fast integer division unit is support for Truncated, Modulo and Euclidean division formats without any cycle penalty. Each of these formats represents the integer and remainder result in different forms. Below is a brief summary of the various division formats:
Both the Euclidean and Modulo formats are more appropriate for precise control applications because the integer value is linear around the zero point and avoid potential calculation hysteresis. The C28x compiler supports all three division formats for all data types.