General-purpose systolic arrays

Abstract

hen Sun Microsystems introduced its first workstation, the company could not have imagined how quickly workstations would revolutionize computing. The idea of a community of engineers, scientists, or researchers time-sharing on a single mainframe computer could hardly have become ancient any more quickly. The almost instant wide acceptance of workstations and desktop computers indicates that [hey were quickly recognized as giving the best and most flexible performance for the dollar.Desktop computers proliferated for three reasons. First, very large scale integration (VLSI) and wafer scale integration (WSI), despite some problems, increased the gate density of chips while dramatically lowering their production cost.' Moreover. increased gate density permits a more complicated processor, which in turn promotes parallelism.Second, desktop computers distribute processing power to the user in an easily customized open architecture. Rcal-time applications that require intensive 110 and computation need not consume all the resources of a supercomputer. Also, desktop computers support high-definition screens with color and motion far exceeding those available with any multiple-user. shared-resource mainframe.Third, economical. high-bandwidth networks allow desktop computers to share data. thus retaining the most appealing aspect of centralized computing, resource sharing. Moreover, networks allow the computers to share data with dissimilar computing machines. That is perhaps the most important reason for the acceptance of desktop computers. since all the performance in the world is worth little if the machine is isolated.Today's workstations have redefined the way the computing community distributes processing resources, and tomorrow's machines will continue this trend with higher bandwidth networks and higher computational performance. One way to obtain higher computational performance is t o use special parallel coprocessors t o perform functions such as motion and color support of high-definition screens. Future computationally intensive applications suited for desktop computing machines include real-time text, speech, and image processing. These applications require massive parallelism.'

A Linear Algebra Core Design for Efficient Level-3 BLAS

“…Systolic arrays were popular in the 80s [5]. With increasing memory walls, recent approaches have brought the computation units closer to memory, including hierarchical clustering of shared memory tiles [6] or network-on-chip architectures [7].…”

On the Efficiency of Register File versus Broadcast Interconnect for Collective Communications in Data-Parallel Hardware Accelerators

“…Different optimizations and algorithms for matrix multiplication and more complicated matrix computations are compared and implemented on both 1D [42], [21] and 2D systolic arrays [42], [16], [30]. In [18], the concept of a general systolic array and a taxonomy of systolic array designs is discussed. Systolic arrays pipeline data and have one-sided interfaces.…”

Analyzing concurrency in streaming applications

“…Typically, concurrency optimization is performed for a given system architecture [MMV98]. A good example of this is the research on performance analysis and design optimization for systolic processors [JHS93,Kun98]. Commonly considered optimization criteria are the computation time, latency, throughput and the number of processors.…”