Spatially resolved precise prediction of local temperature T(x,y,z) is essential to evaluate Arrhenius-activated interconnect (e.g., electromigration) and transistor reliability (e.g., NBTI, HCI, and TDDB). A 3-D finite-element modeling (FEM) do provide excellent results, but the calculation is too time-consuming for a structure that involves eight to ten layers of percolating interconnects, especially for fast turn-around reliability modeling. Here, an analytical model that can quickly/accurately determine T(x,y,z) will reduce the design time of self-heated modern IC. In this paper, we 1) develop a physics-based electrothermal compact model for ICs to predict T(x,y,z) based on the synthesis of effective medium theory, image charge theory, and Rent's rule; 2) validate our model against 3-D FEM and experimental data; and 3) predict back-end-of-line (BEOL) reliability (i.e., electromigration at each layer) based on the temperature profile. Since our analytical model predicts changes in T(x,y,z) with any given IC's configuration (e.g., interconnect wire length and number distribution, metal volume fraction in BEOL, heat sinks mechanisms, materials, and type of devices), it suggests new opportunities for optimization of performance and reliability of modern ICs.

Big Data industrial systems that address problems such as classification, information extraction, and entity matching very commonly use hand-crafted rules. Today, however, little is understood about the usage of such rules. In this paper we explore this issue. We discuss how these systems differ from those considered in academia. We describe default solutions, their limitations, and reasons for using rules. We show examples of extensive rule usage in industry. Contrary to popular perceptions, we show that there is a rich set of research challenges in rule generation, evaluation, execution, optimization, and maintenance. We discuss ongoing work at WalmartLabs and UW-Madison that illustrate these challenges. Our main conclusions are (1) using rules (together with techniques such as learning and crowdsourcing) is fundamental to building semantics-intensive Big Data systems, and (2) it is increasingly critical to address rule management, given the tens of thousands of rules industrial systems often manage today in an ad-hoc fashion.

Wafer level chip scale package (WLCSP) is true chip scale package with low cost by eliminating package substrate. The direct chip-to-board attach through solder joints provides low interconnect inductance and resistance, as well as improved thermal performance. These properties make WLCSP a packaging format well suited for 5G radio frequency (RF) applications where minimized package size and parasitics as well as thermal performance are critical. Due to the dissimilar properties between chip and board, the reliability of WLCSP can be challenging. This paper reports a reliability study of WLCSP using 45nm RFSOI technology for 5G RF applications. Dedicated test chips and boards were designed and used for board level reliability tests. The test vehicles passed bHAST and drop test, whereas it is found that temperature cycling on board (TCoB) is challenging for solder joint reliability in some cases. Thorough tests were carried out based on Kelvin test on specially designed individual bump and bump pair structures and developed fail criterion. Finite element modeling was adopted to simulate the reliability performance in different configurations. The impact on reliability performance from bump depopulation, die thickness, bump size, UBM to board pad alignment, and board wiring trace were thoroughly investigated. Based on comprehensive testing and deep understanding of the failure mechanisms, design optimizations for chip, board and interconnect were implemented. WLCSP reliability was significantly improved.

For the first time, the influence of fast pulse induced skin effect on the current distribution inside the grounded-gate NMOS (GGNMOS) is reported. The skin effect results in the current crowding at the finger edges of the GGNMOS, leading to the high photoemission and high substrate potential at those regions. This report comprehensively explains some of the decades-old unexplained physical phenomena.

As an important task in autonomous driving of intelligent vehicles, the autonomous parking technology has a broad market potential. The traditional parking algorithm based on trajectory planning often results in significant tracking errors due to the non-linearity of vehicle dynamics. While general reinforcement learning algorithms may require the assistance of domain experts due to the complexity of the parking scene. However, relying on expert demonstrations can also be problematic due to the high collection cost and potential for uneven quality of the demonstrations. To address these problems, we propose an end-to-end reinforcement learning-based parking method using expert demonstrations. Firstly, we add Gaussian noise to the collected expert demonstrations for data augmentation, which expands the diversity of samples and guides the agent to learn the expert policy quickly. Then we introduce the mixed priority sampling strategy, and employ it on the augmented expert demonstrations and experience obtained from exploration in the environment to jointly train the agent. We evaluate our method on the benchmark algorithm, Soft Actor-Critic, on the CARLA autonomous driving simulation platform with vertical and parallel parking environments. The experimental results indicate that the proposed algorithm can greatly improve the convergence rate and achieve a higher parking success rate.