Describes how the scheduling algorithms and schedulability analysis methods developed for periodic tasks can be extended to provide performance guarantees to semi-periodic tasks. Like periodic tasks, the requests in a semi-periodic task are released regularly. However, their computation times vary widely. We focus on systems where the total maximum utilization of the tasks on each processor is larger than one. Hence, according to the existing schedulability conditions for periodic tasks, we cannot guarantee that the semi-periodic tasks are schedulable, even though their total average utilization is very small. We describe two methods of providing probabilistic schedulability guarantees to the semi-periodic tasks. The first method, called probabilistic time-demand analysis, is a modification of the exact schedulability test for periodic tasks. The second method, called the transform-task method, transforms each task into a periodic task followed by a sporadic task. The transform-task method can provide an absolute guarantee to requests with shorter computation times and a probabilistic guarantee to the longer requests.

CORAL, the Collaboration of Oak Ridge, Argonne and Livermore, is fielding two similar IBM systems, Summit and Sierra, with NVIDIA GPUs that will replace the existing Titan and Sequoia systems. Summit and Sierra are currently ranked No. 1 and No. 3, respectively on the Top500 list. We discuss the design and key differences of the systems. Our evaluation of the systems highlights the following. Applications that fit in HBM see the most benefit and may prefer more GPUs; however, for some applications, the CPU-GPU bandwidth is more important than the number of GPUs. The node-local burst buffer scales linearly, and can achieve a 4X improvement over the parallel file system for large jobs; smaller jobs, however, may benefit from writing directly to the PFS. Finally, several CPU, network and memory bound analytics and GPU-bound deep learning codes achieve up to a 11X and 79X speedup/node, respectively over Titan.

The modern power grid is strongly integrated with its communication network. While a power system primarily consists of elements that are modeled by continuous equations, a communication system has discrete event dynamics. We model the integrated operation of these two systems with a hybrid modeling and simulation technique. Systematically combining continuous and discrete event system models is necessary for correctly simulating critical system behaviors. This paper discusses an approach based on the discrete event system specification (DEVS) that characterizes the interaction of the two systems formally to preserve simulation correctness. We demonstrate the implementation of our integrated hybrid simulation technique with detailed generator and network models in a wide-area cooperative automatic load-control scenario.

Identification of a low-level point radiation source amidst background radiation is achieved by a network of radiation sensors using a two-step approach. Based on measurements from three sensors, the geometric difference triangulation method is used to estimate the location and strength of the source. Then a sequential probability ratio test based on current measurements and estimated parameters is employed to finally decide: (1) the presence of a source with the estimated parameters, or (2) the absence of the source, or (3) the insufficiency of measurements to make a decision. This method achieves specified levels of false alarm and missed detection probabilities, while ensuring a close-to-minimal number of measurements for reaching a decision. This method minimizes the ghost-source problem of current estimation methods, and achieves a lower false alarm rate compared with current detection methods. This method is tested and demonstrated using: (1) simulations, and (2) a test-bed that utilizes the scaling properties of point radiation sources to emulate high intensity ones that cannot be easily and safely handled in laboratory experiments.