A unified deep-learning network to accurately segment insulin granules of different animal models imaged under different electron microscopy methodologies

Fault tolerance overhead of high performance computing (HPC) applications is becoming critical to the efficient utilization of HPC systems at large scale. Today's HPC applications typically tolerate fail-stop failures by check pointing. However, check pointing will lose its efficiency when system becoming very large. An alternative method is algorithm-based fault recovery which has been proved to be more efficient than check pointing. In this paper, we first point out by theoretical analysis that algorithm-based fault recovery will also lose its efficiency when systems scale up to Exa flops. Then, a more efficient algorithm-based fault tolerance scheme for HPC applications at large scale is presented. The new method has two novel skills. One is algorithm-based hot replacement, which avoids the stop-and-wait time after failure. Second is background accelerated recovery, which guarantees the system to endure multiple failures in succession. As a case study, this method is incorporated to High Performance Lin pack (HPL). Theoretical analysis shows that the fault tolerance overhead can be reduced to 2/log(p, 2) of that of algorithm-based fault recovery method (p is the number of computation processes), so that the new method will still be efficient in Exascale. Experimental results for up to 1800 processes show that the overhead of the new method is about 25% of that of algorithm-based fault recovery method, which is close to the theoretical prediction.

This paper proposes a machine learning based method which can detect certain events automatically and precisely in biomedical imaging. We detect one important and not well-defined event, which is called flash, in fluorescence images of Escherichia coli. Given a time series of images, first we propose a scheme to transform the event detection on region of interest (ROI) in images to a classification problem. Then with supervised human labeling data, we develop a feature selection technique to utilize support vector machine (SVM) to solve this classification problem. To reduce the time in training SVM model, a parallel version of SVM training is implemented. On ten stacks of fluorescence images labeled by experts, each of which owns one hundred 512 ·512 images with in total 4906 ROIs and 72056 labeled events, event detection with proposed method takes 19 seconds, while human labeling roughly costs 60 hours. With human labeling as the standard, the accuracy of our method achieves an F-value of about 0.81. This method is much faster than human detection and expects to be more precise with bigger data. It also can be expanded to a series of event detection with similar properties and improve efficiency of detection greatly.

Fault tolerance overhead of high performance computing (HPC) applications is becoming critical to the efficient utilization of HPC systems at large scale. HPC applications typically tolerate fail-stop failures by checkpointing. Another promising method is in the algorithm level, called algorithmic recovery. These two methods can achieve high efficiency when the system scale is not very large, but will both lose their effectiveness when systems approach the scale of Exaflops, where the number of processors including in system is expected to achieve one million. This paper develops a new and efficient algorithm-based fault tolerance scheme for HPC applications. When failure occurs during the execution, we do not stop to wait for the recovery of corrupted data, but replace them with the corresponding redundant data and continue the execution. A background accelerated recovery method is also proposed to rebuild redundancy to tolerate multiple times of failures during the execution. To demonstrate the feasibility of our new scheme, we have incorporated it to the High Performance Linpack. Theoretical analysis demonstrates that our new fault tolerance scheme can still be effective even when the system scale achieves the Exaflops. Experiment using SiCortex SC5832 verifies the feasibility of the scheme, and indicates that the advantage of our scheme can be observable even in a small scale.