Performance of the European logarithmic microprocessor

Logarithmic arithmetic as an alternative to floating-point: A review

“…We have examined in some detail two case studies in digital signal processing (recursive least-squares, RLS) and numerical methods (Laguerre algorithm) [16]. RLS algorithm results are plotted in Fig.…”

“…The real numbers represented are signed and have magnitudes ranging from 2 -128 to ~2 +128 (i.e., from 2.9  10 -39 to 3.4  10 +38 ). The smallest representable positive value, 40000000 16 , is used as a special code for zero, while C0000000 16 is dedicated to represent NaN.…”

“…Real-time applications in this class are exemplified by RLS-based algorithms, subspace methods required in broadcasting and cellular telephony, Kalman filtering, and Riccati-like equations for advanced real-time control. Graphics systems [16] provide another case in point.…”