GPGPUs have recently emerged as powerful vehicles for general-purpose high-performance computing. Although a new Compute Unified Device Architecture (CUDA) programming model from NVIDIA offers improved programmability for general computing, programming GPGPUs is still complex and error-prone. This paper presents a compiler framework for automatic source-to-source translation of standard OpenMP applications into CUDA-based GPGPU applications. The goal of this translation is to further improve programmability and make existing OpenMP applications amenable to execution on GPGPUs. In this paper, we have identified several key transformation techniques, which enable efficient GPU global memory access, to achieve high performance. Experimental results from two important kernels (JACOBI and SPMUL) and two NAS OpenMP Parallel Benchmarks (EP and CG) show that the described translator and compile-time optimizations work well on both regular and irregular applications, leading to performance improvements of up to 50X over the unoptimized translation (up to 328X over serial).

Irregular applications pose challenges in optimizing communication, due to the difficulty of analyzing irregular data accesses accurately and efficiently. This challenge is especially big when translating irregular shared-memory applications to message-passing form for clusters. The lack of effective irregular data analysis in the translation system results in unnecessary or redundant communication, which limits application scalability. In this paper, we present a Lean Distributed Shared Memory (LDSM) system, which features a fast and accurate irregular data access (IDA) analysis. The analysis uses a region-based diff method and makes use of a runtime library that is optimized for irregular applications. We describe three optimizations that improve the LDSM system performance. A parallel array reduction transformation reduces overheads in the analysis. A packed communication optimization and a differential communication optimization effectively eliminate unnecessary and redundant messages. We evaluate the performance of the optimized LDSM system on a set of representative irregular benchmarks. The optimized LDSM executes irregular applications on average 45% faster than the hand-tuned MPI applications.

OpenMP has emerged as an important model and language extension for shared-memory parallel programming. On shared-memory platforms, OpenMP offers an intuitive, incremental approach to parallel programming. In this paper, we present techniques that extend the ease of shared-memory parallel programming in OpenMP to distributed-memory platforms as well. First, we describe a combined compile-time/runtime system that uses an underlying software distributed shared memory system and exploits repetitive data access behavior in both regular and irregular program sections. We present a compiler algorithm to detect such repetitive data references and an API to an underlying software distributed shared memory system to orchestrate the learning and proactive reuse of communication patterns. Second, we introduce a direct translation of standard OpenMP into MPI message-passing programs for execution on distributed memory systems. We present key concepts and describe techniques to analyze and efficiently handle both regular and irregular accesses to shared data. Finally, we evaluate the performance achieved by our approaches on representative OpenMP applications.

Stochastic computation uses pulse streams to represent numbers. In this paper, we have studied the novel method to implement the number system which uses the ratio of the number of one (high) pulses and the number of zero (low) pulses in a pulse stream. With this number system, if we let P be the probability that the pulse is one in a pulse stream, then the number Y we want to deal with is defined as Y=P/(1-P). We have studied the method to implement the basic operations such as an addition, a multiplication and a division using this number system and considered the error characteristics of such operations in stochastic computation. Also we introduce an averaging circuit to reduce the error which is inherent in stochastic computation. According to the study, by using this number system we can implement artificial neural networks and represent the simulation results for the digit recognition problem.>

In this paper, we propose a new approach of solving the stereopsis problem with a discrete-time cellular neural network (DTCNN) where each node has connections only with its local neighbors. Because the matching process of stereo correspondence depends on its geometrically local characteristics, the DTCNN is suitable for the stereo correspondence. Moreover, it can be easily implemented in VLSI. Therefore, we employed a two-layer DTCNN with dual templates, which are determined with the back propagation learning rule. Based on evaluation of the proposed approach for several random dot stereograms, its performance is better than that of the Marr-Poggio algorithm.>