Abstract:
A process for transmitting data on a bus, minimizing the switching activity, involves converting the data between a first format (b(t)) and a second format (B(t)) used for transmission of the data. The conversion between said first format (b(t)) and said second format (B(t)) entails the swapping of position of respective bits within a cluster comprising a given number of bits, the swap operation being implementable according to different variants, the maximum number of said variants being equal to the factorial of the aforesaid given number. Each of said variants is identified by a respective pattern (P t ). Among the aforesaid patterns, an optimal pattern is selected which minimizes the switching activity (SA) at the moment of transmission of data on the bus. The data are then transmitted on the bus using the second format (B(t)) generated using said optimal pattern.
Abstract:
Processing method for forming an aggregate signal (y beam ) from a plurality of starting signals (x 1 , ... ,x 5 ), the method comprising the steps of: - acquiring said starting signals (x 1 , ... ,x 5 ) through respective sensors of a homogeneous sensors group (U 1 , ... ,U 5 ); - converting acquired signals in respective digital signals (x 1 , ... ,x 5 ) having data represented with a predetermined bits number; - processing the digital signals (x 1 , ... ,x 5 ) to form aggregate signal (y beam ). The processing step comprises the operations of: - modifying digital signals (x 1 , ... ,x 5 ) changing the data format of each such digital signals from a first format to a second format, each data in the second format having been obtained from a respective data in the first format through an operation of permuting the bits position according to a permutation scheme (p 1 , ... ,p 5 ) associated with said data and to the specific digital signal comprising that data; - forming aggregate signal (y beam ) obtaining said aggregate signal data by means of a bitwise logic operator acting upon said modified digital signal respective data.
Abstract:
Described herein is a method for transmitting data on a bus with minimization of the bus switching activity, comprising the steps of converting the datum to be transmitted from its own original format to a transmission format that minimizes the switching activity on the bus, said conversion consisting in swapping the position of one or more bits of the datum to be transmitted, the swapping being performable according to a plurality of different variants, each of which is identified by a respective sorting pattern, and selecting, between the various sorting patterns, an optimal sorting pattern that minimizes the bus switching activity upon transmission on the bus of the datum generated using said optimal sorting pattern.
Abstract:
A data stream (b(t)) including high ("1") and low ("0") logical states is transmitted over an optical link (16) by means of an optical source (15) adapted to be driven (14) via said the data stream to generate an optical signal for transmission over the optical link (16). The optical signal includes optical pulses generated at the occurrence of high logical ("1") states in said data stream (b(t)). The input data stream (b(t)) is coded (2000) into a coded data stream (B(t)) prior to the transmission over the optical link (16). The coding step minimises the logical high states ("1") in the coded data stream (B(t)), and the optical source (15) is driven by means of the coded data stream (B(t)) wherein the number of logical high states ("1") has been minimised.
Abstract:
In a method for multiplication of floating-point real numbers (f, FN), encoded in a binary way in sign (SGN, SN), exponent (E, EN) and mantissa (M; MN), the multiplication of the mantissa (M; MN) envisages a step of calculation of partial products, which are constituted by a set of addenda (P) corresponding to said mantissa (MN). In order to reduce the size and power consumption of the circuits designed for calculation, there is adopted a method of binary encoding which envisages setting the first bit of the mantissa (MN) to a value 1, in order to obtain a mantissa (MN) having a value comprised between 0.5 and 1. Also proposed are methods for rounding of the product and circuits for the implementation of the multiplication method. Also illustrated are circuits for conversion from and to encoding of floating-point real numbers according to the IEEE754 standard. Preferential application is in portable and/or wireless electronic devices, such as mobile telephones and PDAs, with low power-consumption requirements.
Abstract translation:在以二进制方式编码的符号(SGN,SN),指数(E,EN)和尾数(M; MN)的浮点实数(f,FN)的乘法方法中,尾数( M; MN)设想了由对应于所述尾数(MN)的一组附加(P)构成的部分乘积的计算步骤。 为了减小设计用于计算的电路的尺寸和功耗,采用了一种二进制编码方法,其设想将尾数(MN)的第一比特设置为值1,以获得尾数(MN) 具有包括在0.5和1之间的值。还提出了用于实现乘法的乘积和电路的舍入的方法。 还示出了根据IEEE754标准从浮点实数转换和编码的电路。 优先应用于便携式和/或无线电子设备,例如移动电话和PDA,具有低功耗要求。
Abstract:
A circuit for estimating propagated carries in an adder starting from operands that include actual addition inputs or at least one earlier carry, the circuit performs statistical circuit operations with independent binary traffic for the operands. Preferably, this binary traffic is independent and equiprobable or quasi-equiprobable binary traffic, and the adder is a leading zero anticipatory logic integer adder producing a number having the same number of leading zeroes as the result of the integer addition performed. The carry value may be produced from a logic function (e.g., Karnaugh Map, Quine-McClusky) of the operands, as a logic combination of the operands covering all the 1s in the logic function.