Markov Chain Monte Carlo (MCMC) is an ubiquitous stochastic method, used to draw random samples from arbitrary probability distributions, such as the ones encountered in Bayesian inference. MCMC often requires forbiddingly long runtimes to give a representative sample in problems with high dimensions and large-scale data. Field-Programmable Gate Arrays (FPGAs) have proven to be a suitable platform for MCMC acceleration due to their ability to support massive parallelism. This paper introduces an automated method, which minimizes the floating point precision of the most computationally intensive part of an FPGA-mapped MCMC sampler, while keeping the precision-related bias in the output within a user-specified tolerance. The method is based on an efficient bias estimator, proposed here, which is able to estimate the bias in the output with only few random samples. The optimization process involves FPGA pre-runs, which estimate the bias and choose the optimized precision. This precision is then used to reconfigure the FPGA for the final, long MCMC run, allowing for higher sampling throughputs. The process requires no user intervention. The method is tested on two Bayesian inference case studies: Mixture models and neural network regression. The achieved speedups over double-precision FPGA designs were 3.5x-5x (including the optimization overhead). Comparisons with a sequential CPU and a GPGPU showed speedups of 223x-446x and 16x-18x respectively.

No abstract

Knowledge of deformation and failure mechanisms at micro- to nano-length scales is important for the prediction of material behavior as well as the development of new materials with desired properties. In situ multiaxial testing with scanning electron microscopes (SEM) can reveal physical deformation mechanisms under realistic multiaxial loading conditions. Although in situ SEM testing has gained traction in recent years, there is currently no multiaxial in situ SEM testing stage available with axial-torsional loading capabilities which can generally be used in any SEM. In this study, we report the development of a multiaxial miniature testing system (MMTS) with a unique capability for performing axial-torsional testing of a tubular specimen with a 1–2 mm outer diameter, inside most SEMs. The different challenges of developing a multiaxial in situ SEM testing stage, such as small load frame size, appropriate specimen position, high vacuum compatibility of MMTS load frame components, as well as the development of installation accessories, were addressed. A custom SEM stage door was developed for the MMTS load frame. Verification tests have confirmed the successful development of the MMTS for in situ SEM testing. In addition, digital image correlation was used with recorded SEM images during the test to determine the surface strain.

Living in this modern era – the epitome of communication GSM networks is one of the mainly used architectures. But GSM architecture has its own shortcomings; the GSM network is vulnerable to various security threats. For any network to provide security to the user, the algorithms should be planned and designed in such a way that it provides cellular secrecy, data and signaling confidentiality to the concerned user. Keeping in mind the above features, the A5/1 algorithm provides network security. Initially, the A5/1 algorithm dealt with a pre-defined secret key but they still possess the threat of being decrypted by cryptanalytic attacks. Although decrypting this algorithm is not easy and requires high computational power. Such attacks lead to the necessity to modify the A5/1 algorithm; in our paper, we have proposed a better method to enhance the already existing algorithm.

No abstract