Many computational problems consider memory throughput a performance bottleneck, especially in the domain of parallel computing. Software needs to be attuned to hardware features like cache architectures or concurrent memory banks to reach a decent level of performance efficiency. This can be achieved by selecting the right memory layouts for data structures or changing the order of data structure traversal. In this work, we present an abstraction for traversing a set of regular data structures (e.g., multidimensional arrays) that allows the design of traversal-agnostic algorithms. Such algorithms can easily optimize for memory performance and employ semi-automated parallelization or autotuning without altering their internal code. We also add an abstraction for autotuning that allows defining tuning parameters in one place and removes boilerplate code. The proposed solution was implemented as an extension of the Noarr library that simplifies a layout-agnostic design of regular data structures. It is implemented entirely using C++ template meta-programming without any nonstandard dependencies, so it is fully compatible with existing compilers, including CUDA NVCC or Intel DPC++. We evaluate the performance and expressiveness of our approach on the Polybench-C benchmarks.

Self-adaptive systems comprise a complex domain of computing systems that are intensively studied but sparsely employed in real applications. Furthermore, recent trends in computer science are steering towards machine learning which has yet to fully penetrate this domain. We would like to present Simdex --- a realistic simulator of the self-adaptive backend that dispatches computing jobs among multiple workers. It is based on ReCodEx, a system for semi-automated evaluation of coding assignments that have been used for the past 5 years at our School of Computer Science. The simulator replays the workload logs recorded from ReCodEx over that period which provides a quite thorough evaluation and near-to-real feedback for the simulated scenarios. Furthermore, the design of the simulator is highly modular and allows the implementation of different self-adaptive controllers, including ones based on machine learning, as we demonstrate in our examples.

K-means is one of the most frequently used algorithms for unsupervised clustering data analysis. Individual steps of the k-means algorithm include nearest neighbor finding, efficient distance computation, and cluster-wise reduction, which may be generalized to many other purposes in data analysis, visualization, and machine learning. Efficiency of the available implementations of k-means computation steps therefore directly affect many other applications. In this work, we examine the performance limits in the context of modern massively parallel GPU accelerators. Despite the existence of many published papers on this topic, we have found that crucial performance aspects of the GPU implementations remain unaddressed, including the optimizations for memory bandwidth, cache limits, and workload dispatching on problem instances of varying cluster count, dataset size, and dimensionality. We present a detailed analysis of individual computation steps and propose several optimizations that improve the overall performance on contemporary GPU architectures. Our open-source prototype exhibits significant speedup over the current state-of-the-art implementations in virtually all practical scenarios.

At present time, the programmers may choose from a number of streaming languages. They cover various aspects of the development process of streaming applications; however, specification of complex or runtime-dependent parts of the applications still remains a great challenge. We have analysed a large amount of requirements raised by the development of multiple data streaming parallel applications and proposed a novel language called Bobolang. It contains syntactic and semantic features which allow the programmer to naturally solve most of the problems, which we met in the design of streaming applications. The language is used to specify the structure of the whole application as well as the inner structure of each operator. Thanks to the properties of the language, Bobolang can create an optimized evaluation plan which is capable of making the best use of the available hardware resources. The language has been employed in several practical problems and it has proven itself to be a very powerful tool for the development of data-intensive parallel applications.

The Signature Quadratic Form Distance on feature signatures represents a flexible distance-based similarity model for effective content-based multimedia retrieval. Although metric indexing approaches are able to speed up query processing by two orders of magnitude, their applicability to large-scale multimedia databases containing billions of images is still a challenging issue. In this paper, we propose the utilization of GPUs for efficient query processing with the Signature Quadratic Form Distance. We show how to process multiple distance computations in parallel and demonstrate efficient query processing by comparing many-core GPU with multi-core CPU implementations.