From bit level systolic arrays to HDTV processor chips

The initial part of this paper reviews the early challenges (c 1980) in achieving real-time silicon implementations of DSP computations. In particular, it discusses research on application specific architectures, including bit level systolic circuits that led to important advances in achieving the DSP performance levels then required. These were many orders of magnitude greater than those achievable using programmable (including early DSP) processors, and were demonstrated through the design of commercial digital correlator and digital filter chips. As is discussed, an important challenge was the application of these concepts to recursive computations as occur, for example, in Infinite Impulse Response (IIR) filters. An important breakthrough was to show how fine grained pipelining can be used if arithmetic is performed most significant bit (msb) first. This can be achieved using redundant number systems, including carry-save arithmetic. This research and its practical benefits were again demonstrated through a number of novel IIR filter chip designs which at the time, exhibited performance much greater than previous solutions. The architectural insights gained coupled with the regular nature of many DSP and video processing computations also provided the foundation for new methods for the rapid design and synthesis of complex DSP System-on-Chip (SoC), Intellectual Property (IP) cores. This included the creation of a wide portfolio of commercial SoC video compression cores (MPEG2, MPEG4, H.264) for very high performance applications ranging from cell phones to High Definition TV (HDTV). The work provided the foundation for systematic methodologies, tools and design flows including high-level design optimizations based on "algorithmic engineering" and also led to the creation of the Abhainn tool environment for the design of complex heterogeneous DSP platforms comprising processors and multiple FPGAs. The paper concludes with a discussion of the problems faced by designers in developing complex DSP systems using current SoC technology. © 2007 Springer Science+Business Media, LLC.

IMPLEMENTATION OF SIGNAL PROCESSING FUNCTIONS USING 1-BIT SYSTOLIC ARRAYS.

The use of systolic arrays of 1-bit cells to implement a range of important signal processing functions is demonstrated. Two examples, a pipelined multiplier and a pipelined bit-slice transform circuit, are given. This approach has many important implications for silicon technology, and these are outlined.

YIELD ENHANCEMENT OF BIT LEVEL SYSTOLIC ARRAY CHIPS USING FAULT TOLERANT TECHNIQUES.

Methods by which bit level systolic array chips can be made fault tolerant are discussed briefly. Using a simple analysis based on both Poisson and Bose-Einstein statistics authors demonstrate that such techniques can be used to obtain significant yield enhancement. Alternatively, the dimensions of an array can be increased considerably for the same initial (nonfault tolerant) chip yield.

OPTIMISED BIT LEVEL SYSTOLIC ARRAY FOR CONVOLUTION.

A bit level systolic array for computing the convolution operation is described. The circuit in question is highly regular and ideally suited to VLSI chip design. It is also optimized in the sense that all the cells contribute to the computation on each clock cycle. This makes the array almost four times more efficient than one which was previously described.

USE OF UNIDIRECTIONAL DATA FLOW IN BIT-LEVEL SYSTOLIC ARRAY CHIPS.

We show how the architecture of two recently reported bit-level systolic array circuits - a single-bit coefficient correlator and a multibit convolver - may be modified to incorporate unidirectional data flow. This feature has advantages in terms of chip cascadability, fault tolerance and possible wafer-scale integration.

UTILISATION OF BIT LEVEL SYSTOLIC ARRAYS IN WORD LEVEL SYSTEMS.

Bit level systolic array structures for computing sums of products are studied in detail. It is shown that these can be sub-divided into two classes and that, within each class, architectures can be described in terms of a set of constraint equations. It is further demonstrated that high performance system level functions with attractive VLSI properties can be constructed by matching data flow geometries in bit level and word level architectures.

BIT-LEVEL SYSTOLIC ARRAY CIRCUIT FOR MATRIX VECTOR MULTIPLICATION.

A bit-level systolic array for computing matrix x vector products is described. The operation is carried out on bit parallel input data words and the basic circuit takes the form of a 1-bit slice. Several bit-slice components must be connected together to form the final result, and authors outline two different ways in which this can be done. The basic array also has considerable potential as a stand-alone device, and its use in computing the Walsh-Hadamard transform and discrete Fourier transform operations is briefly discussed.

MAPPING SYSTEM LEVEL FUNCTIONS ON TO BIT LEVEL SYSTOLIC ARRAYS.

Bit-level systolic-array structures for computing sums of products are studied in detail. It is shown that these can be subdivided into two classes and that within each class architectures can be described in terms of a set of constraint equations. It is further demonstrated that high-performance system-level functions with attractive VLSI properties can be constructed by matching data-flow geometries in bit-level and word-level architectures.

BIT-LEVEL SYSTOLIC ARRAYS.

A systolic array is an array of individual processing cells each of which has some local memory and is connected only to its nearest neighbours in the form of a regular lattice. On each cycle of a simple clock every cell receives data from its neighbouring cells and performs a specific processing operation on it. The resulting data is stored within the cell and passed on to neighbouring cells on the next clock cycle. This paper gives an overview of work to date and illustrates the application of bit-level systolic arrays by means of two examples: (1) a pipelined bit-slice circuit for computing matrix x vector transforms; and (2) a bit serial structure for multi-bit convolution.

RELATIONSHIP BETWEEN WORD AND BIT LEVEL SYSTOLIC ARRAYS AS APPLIED TO MATRIX multiplied by MATRIX MULTIPLICATIONS.

The mapping of matrix multiplied by matrix multiplication onto both word and bit level systolic arrays has been investigated. It has been found that well defined word and bit level data flow constraints must be satisfied within such circuits. An efficient and highly regular bit level array has been generated by exploiting the basic compatibilities in data flow symmetries at each level of the problem. A description of the circuit which emerges is given and some details relating to its practical implementation are discussed.

Use of data dependence graphs in the design of bit-level systolic arrays

The use of bit-level systolic array circuits as building blocks in the construction of larger word-level systolic systems is investigated. It is shown that the overall structure and detailed timing of such systems may be derived quite simply using the dependence graph and cut-set procedure developed by S. Y. Kung (1988). This provides an attractive and intuitive approach to the bit-level design of many VLSI signal processing components. The technique can be applied to ripple-through and partly pipelined circuits as well as fully systolic designs. It therefore provides a means of examining the relative tradeoff between levels of pipelining, chip area, power consumption, and throughput rate within a given VLSI design.

VQ tree search system based on bit level systolic arrays

A bit-level systolic array system for performing a binary tree vector quantization (VQ) codebook search is described. This is based on a highly regular VLSI building block circuit. The system in question exhibits a very high data rate suitable for a range of real-time applications. A technique is described which reduces the storage requirements of such a system by 50%, with a corresponding decrease in hardware complexity.

MULTIBIT CONVOLUTION USING A BIT LEVEL SYSTOLIC ARRAY.

A novel design for multibit convolver circuits is described. The circuits take the form of systolic arrays of simple one-bit processor and memory cells, with the result that they can operate at very high data rates and should be easy to implement using VLSI technology. An efficient method for handling two's complement data within the array is described and the relative advantages of this convolver design compared with more conventional circuits is discussed.