floating point operations at Blackfin

Document created by analog-archivist Employee on Feb 23, 2016Last modified by analog-archivist Employee on Feb 23, 2016
Version 2Show Document
  • View in full screen mode

We would like to know how many cycles it will take the Blackfin processor to
perform a floating point operation (MAC, since, etc) by emulating 32-bit
floating point by way of an optimized library provided by
ADI.

 

Please refer to EE-185: Fast Floating-Point Arithmetic Emulation on the
Blackfin Processor Platform. This EE-Note can be found here:
www.analog.com/ee-notes.

This document shows how to emulate fast floating-point arithmetic on the
Blackfin Processor platform. A two-word format is employed for representing
short and long fast floating-point data types. C-callable assembly
source code is included for the following operations: addition, subtraction,
multiplication and conversion between fixed-point, IEEE-754 floating-point, and
the fast floating-point formats.

Furthermore, attached you'll find the ADSP-BF532 DSP Library CYCLE COUNTS for
VisualDSP++ 3.5, which include cycle counts on floating point operations. For
any questions related to this, please contact the DSP Tools Support Group
directly. They can be reached at processor.tools.support@analog.com.

Outcomes