Analysis of floating point operations in microcontrollers

The purpose of this paper is to identify the advantages of including a floating point hardware / a mathematical co-processor in microcontrollers used for critical floating point operations. Three different microcontrollers are considered: Renesas M16C/62P (CISC without FPU), ATMEGA1280 (RISC without MCU) and Renesas RX62N (CISC with FPU). Gauss-Seidel and Jacobi iterations for a system of 2 and 3 variables using single precision and double precision floating point numbers are implemented. The time taken by each microcontroller is measured using the Timer in the microcontrollers. Based on the time taken, the advantages of the inclusion of the floating point unit are attributed against the cost incurred by the hardware.

[1]  Herman H. Goldstine,et al.  The Jacobi Method for Real Symmetric Matrices , 1959, JACM.

[2]  J. Meijerink,et al.  An iterative solution method for linear systems of which the coefficient matrix is a symmetric -matrix , 1977 .

[3]  Daniel Tabak,et al.  Floating point arithmetic on a RISC , 1988 .

[4]  Peter Kornerup,et al.  Semantics for exact floating point operations , 1991, [1991] Proceedings 10th IEEE Symposium on Computer Arithmetic.

[5]  Larry L. Biro,et al.  Power considerations in the design of the Alpha 21264 microprocessor , 1998, Proceedings 1998 Design and Automation Conference. 35th DAC. (Cat. No.98CH36175).

[6]  Wayne Luk,et al.  Unifying bit-width optimisation for fixed-point and floating-point designs , 2004, 12th Annual IEEE Symposium on Field-Programmable Custom Computing Machines.