Alécio Pedro Delazari Binotto

An Effective Dynamic Scheduling Runtime and Tuning System for Heterogeneous Multi and Many-Core Desktop Platforms

A personal computer can be considered as a one-node heterogeneous cluster that simultaneously processes several application tasks. It can be composed by, for example, asymmetric CPU and GPUs. This way, a high-performance heterogeneous platform is built on a desktop for data intensive engineering calculations. In our perspective, a workload distribution over the Processing Units (PUs) plays a key role in such systems. This issue presents challenges since the cost of a task at a PU is non-deterministic and can be affected by parameters not known a priori. This paper presents a context-aware runtime and tuning system based on a compromise between reducing the execution time of engineering applications - due to appropriate dynamic scheduling - and the cost of computing such scheduling applied on a platform composed of CPU and GPUs. Results obtained in experimental case studies are encouraging and a performance gain of 21.77% was achieved in comparison to the static assignment of all tasks to the GPU.

Iterative SLE Solvers over a CPU-GPU Platform

GPUs (Graphics Processing Units) have become one of the main co-processors that contributed to desktops towards high performance computing. Together with multi-core CPUs, a powerful heterogeneous execution platform is built for massive calculations. To improve application performance and explore this heterogeneity, a distribution of workload in a balanced way over the PUs (Processing Units) plays an important role for the system. However, this problem faces challenges since the cost of a task at a PU is non-deterministic and can be influenced by several parameters not known a priori, like the problem size domain. We present a comparison of iterative SLE (Systems of Linear Equations) solvers, used in many scientific and engineering applications, over a heterogeneous CPU-GPUs platform and characterize scenarios where the solvers obtain better performances. A new technique to improve memory access on matrix-vector multiplication used by SLEs on GPUs is described and compared to standard implementations for CPU and GPUs. Such timing profiling is analyzed and break-even points based on the problem sizes are identified for this implementation, pointing whether our technique is faster to use GPU instead of CPU. Preliminary results show the importance of this study applied to a real-time CFD (Computational Fluid Dynamics) application with geometry modification.

Towards dynamic reconfigurable load-balancing for hybrid desktop platforms

High-performance platforms are required by applications that use massive calculations. Actually, desktop accelerators (like the GPUs) form a powerful heterogeneous platform in conjunction with multi-core CPUs. To improve application performance on these hybrid platforms, load-balancing plays an important role to distribute workload. However, such scheduling problem faces challenges since the cost of a task at a Processing Unit (PU) is non-deterministic and depends on parameters that cannot be known a priori, like input data, online creation of tasks, scenario changing, etc. Therefore, self-adaptive computing is a potential paradigm as it can provide flexibility to explore computational resources and improve performance on different execution scenarios. This paper presents an ongoing PhD research focused on a dynamic and reconfigurable scheduling strategy based on timing profiling for desktop accelerators. Preliminary results analyze the performance of solvers for SLEs (Systems of Linear Equations) over a hybrid CPU and multi-GPU platform applied to a CFD (Computational Fluid Dynamics) application. The decision of choosing the best solver as well as its scheduling must be performed dynamically considering online parameters in order to achieve a better application performance.

Effective Dynamic Scheduling on Heterogeneous Multi/Manycore Desktop Platforms

GPUs (Graphics Processing Units) have become one of the main co-processors that contributed to desktops towards high performance computing. Together with multicore CPUs and other co-processors, a powerful heterogeneous execution platform is built on a desktop for data intensive calculations. In our perspective, we see the modern desktop as a heterogeneous cluster that can deal with several applications’tasks at the same time. To improve application performance and explore such heterogeneity, a distribution of workload over the asymmetric PUs (Processing Units) plays an important role for the system. However, this problem faces challenges since the cost of a task at a PU is non-deterministic and can be influenced by several parameters not known a priori, like the problem size domain. We present a context-aware architecture that maximizes application performance on such platforms. This approach combines a model for a first scheduling based on an offline performance benchmark with a runtime model that keeps track of tasks’ real performance. We carried a demonstration using a CPU-GPU platform for computing iterative SLEs (Systems of Linear Equations) solvers using the number of unknowns as the main parameter for assignment decision. We achieved a gain of 38.3% in comparison to the static assignment of all tasks to the GPU (which is done by current programming models, such as Open CL and CUDA for Nvidia).

Dynamic Reconfiguration of Tasks Applied to an UAV System Using Aspect Orientation

Many modern applications require high-performance platforms to deal with a variety of algorithms requiring massive calculations. Moreover, low-cost powerful hardware (e.g., GPU, PPU) and CPUs with multiple cores have become abundant, and can be combined in heterogeneous architectures. To cope with this, reconfigurable computing is a potential paradigm as it can provide flexibility to explore the computational resources on hybrid and multi-core desktop architectures. The workload can optimally be (re)distributed over heterogeneous cores along the lifecycle of an application, aiming for best performance. As the first step towards a run-time reconfigurable load-balancing framework, application requirements and crosscutting concerns related to timing play an important role for task allocation decisions. In this paper, we present the use of aspect-oriented paradigms to address non-functional application timing constraints in the design phase. The DERAF aspectspsila framework is extended to support reconfiguration requirements; and a strategy for load-balancing is described. In addition, we present preliminary evaluation using an Unmanned Aerial Vehicle (UAV) based surveillance system as case study.

Real-time task reconfiguration support applied to an UAV-based surveillance system

Modern surveillance systems, such as those based on the use of unmanned aerial vehicles, required powerful high-performance platforms to deal with many different algorithms that make use of massive calculations. At the same time, low-cost and high-performance specific hardware (e.g., GPU, PPU) are rising and the CPUs turned to multiple cores, characterizing together an interesting and powerful heterogeneous execution platform. Therefore, reconfigurable computing is a potential paradigm for those scenarios as it can provide flexibility to explore the computational resources on heterogeneous cluster attached to a high-performance computer system platform. As the first step towards a run-time reconfigurable workload balancing framework targeting that kind of platform, application time requirements and its crosscutting behavior play an important role for task allocation decisions. This paper presents a strategy to reallocate specific tasks in a surveillance system composed by a fleet of unmanned aerial vehicles using aspect-oriented paradigms in order to address non-functional application timing constraints in the design phase. An aspect support from a framework called DERAF is used to support reconfiguration requirements and provide the resource information needed by the reconfigurable load-balancing strategy. Finally, for the case study, a special attention on radar image processing will be given.

BraTrack: a low-cost marker-based optical stereo tracking system

Optical motion tracking systems are widely employed in virtual- and mixed reality applications enabling users to directly interact with 3D space. Important efforts have been made to make such technologies available to a larger group of users by significantly reducing their deployment costs. In this poster we introduce the result of a German-Brazilian cooperation: The first commercial low-cost marker-based optical tracking system developed in South-America. Despite using cheaper technologies and being low-cost which usually imply worse performance and accuracy, we developed a system with sufficient accuracy for professional use. There have been significant research efforts to develop cheaper tracking devices to provide interaction capabilities in virutal- and mixed reality. The Augmented Reality Tool Kit (ARToolkit) [1] is a good example of a very popular low-cost academic optical tracker working with any type of webcam. As a non-academic example of affordable tracking systems, Nintendos latest video game console (Nintendo Wii) comes with the Wiimote interaction device containing a built-in motion tracker that allows for immersive interaction metaphors in home environments. However, most such available low-cost solutions lack in suitability for professional applications due to their usually insufficient accuracy, tracking range or latency time. The starting point of this project was PTrack [2], an academic marker-based single-camera tracker developed at Fraunhofer IGD. PTrack development has shown that an accurate optical tracking system can be built from components that are affordable for a much wider group of users, from the academic AR/VR community to companies’ developers. The performance and accuracy of PTrack were assessed through a comprehensive set of experiments which have proved the suitability of the system for professional applications. One of the main topics raised by PTrack as further research was an extension of the system into a multiple stereo camera configuration, which led to the present work.

Dynamic Self-Rescheduling of Tasks over a Heterogeneous Platform

Modern applications require powerful high-performance platforms to deal with many different algorithms that make use of massive calculations. At the same time, low-cost and high-performance specific hardware (e.g., GPU, PPU) are rising and the CPUs turned to multiple cores, characterizing together an interesting and powerful heterogeneous execution platform. Therefore, self-adaptive computing is a potential paradigm for those scenarios as it can provide flexibility to explore the computational resources on heterogeneous cluster attached to a high-performance computer system platform. As the first step towards a run-time reschedule load-balancing framework targeting that kind of platform, application time requirements and its crosscutting behavior play an important role for task allocation decisions. This paper presents a strategy for self-reallocation of specific tasks, including dynamic created ones, using aspect-oriented paradigms to address non-functional application timing constraints in the design phase. Additionally, as a case study, a special attention on radar image processing will be given in the context of a surveillance system based on unmanned aerial vehicles (UAV).

Cloud-based Remote Visualization of Big Data to Subsurface Exploration

Since the first visualization solutions were explored for O&G, major technical improvements enabled gigabyte-sized models to be rendered and manipulated at interactive speeds. Yet, other fundamental aspects such as data access and distribution are often overlooked in the process, even to date. Data movement may often be prohibitive, either due to legal constraints (data is restricted from departing the country) or practical considerations (data is too large to be moved, is a checkpoint tightly connect to imaging processes, and requires costly resources to be manipulated). Collaborative visualization tends to be performed co-locally, following an explicit, manually conducted, data transfer to a dedicated visualization machine. We propose an alternative based upon data-centric computing. The model offers visualization as-a-service over a multitenant cloud-based environment. Remote visualization enables lightweight access and interaction with generated data readily after the processing, dismissing the need to transfer the whole dataset into analysts machine. It offloads heavy graphics processing to a cloud server featuring the necessary infrastructure to handle such voluminous data, like GPUs, GPFS, and ultimately sends only a reduced output to lightweight clients (rendered images/geometry). Visualization resources can also be shared among concurrent users in a web-based interface and combined with other data sources, like correspondent well information or velocity models, facilitating effective remote collaboration towards knowledge discovery in subsurface exploration. Ubiquitous on-the-go data access (e.g., in the exploration field itself) is thereby made possible through mobile interfaces. Concurrently, several challenges emerge with the aforementioned visualization model. The effective resource distribution of different data sources among several clients needs to benefit from the cloud execution platform. The OpenPower Foundation is an example of the future HPC platform that can be customized to O&G characteristics and be offered as a cloud model.

Towards Dynamic Task Scheduling and Reconfiguration using an Aspect Oriented Approach applied on Real-time concerns of Industrial Systems

Abstract High performance computational platforms are required by industries that make use of automatic methods to manage modern machines, which are mostly controlled by high-performance specific hardware with processing capabilities. It usually works together with CPUs, forming a powerful execution platform. On an industrial production line, distinct tasks can be assigned to be processed by different machines depending on certain conditions and production parameters. However, these conditions can change at run-time influenced mainly by machine failure and maintenance, priorities changes, and possible new better task distribution. Therefore, self-adaptive computing is a potential paradigm as it can provide flexibility to explore the machine resources and improve performance on different execution scenarios of the production line. One approach is to explore scheduling and run-time task migration among machines’ hardware towards a balancing of tasks, aiming performance and production gain. This way, the monitoring of time requirements and its crosscutting behaviour play an important role for task (re)allocation decisions. This paper introduces the use of software aspect-oriented paradigms to perform machines’ monitoring and a self-rescheduling strategy of tasks to address non-functional timing constraints. As case study, tasks for a production line of aluminium ingots are designed.