Search

Tier partitioning is one of the most critical stages in monolithic 3-D (M3D) integrated circuits (ICs) implementation flows. It transforms 2-D netlists into 3-D by performing tier assignment for each design instance, which directly impacts the power, performance, and area (PPA) metrics of final 3-D full-chip designs. However, the current state-of-the-art tier partitioning approach named bin-based min-cut algorithm has fundamental flaws that lead to severe drawbacks, such as timing degradation, 3-D routing overhead, and redundant monolithic intertier vias (MIVs) insertion. To overcome these issues, in this article, we propose TP-GNN, an unsupervised graph learning-based tier partitioning framework that utilizes graph neural networks (GNNs) and advanced machine learning (ML) techniques to perform tier partitioning. The proposed framework comprehends design- and technology-related parameters properly so that it is generalizable to various netlists and technologies. In addition, it can be integrated with any style of M3D design flows that require tier assignments of standard cells. In the experiments, we validate the proposed framework on seven industrial designs with two different fashions of M3D implementation flows: 1) partitioning-first (Snap3D) and 2) partitioning-last (Shrunk2D and Compact2D) flows. We demonstrate that our framework, TP-GNN, significantly improves the 3-D quality of results (QoR) across most testing designs in a large margin compared with the bin-based min-cut tier partitioning algorithm. Specifically, in OpenPiton, an RISC-V-based multicore system, we observe 27.4%, 7.7%, and 20.3% improvements in performance, wirelength, and energy-per-cycle, respectively. Finally, we perform a case study by applying the proposed framework to a heterogeneous M3D design flow, Pin3D, on a commercial CPU design and observe that TP-GNN reaches better partitioning solutions than the existing partitioning approaches for heterogeneous 3-D ICs. [ABSTRACT FROM AUTHOR]

Modern physical design flows highly depend on design space exploration to find the commercial tools’ clock tree synthesis (CTS) parameters that lead to optimized clock trees. However, such exploration is often time-consuming and computationally inefficient. In this article, we overcome this drawback by proposing a novel framework named GAN-CTS, which utilizes conditional generative adversarial network (GAN) to predict and optimize CTS outcomes. Our framework is built upon three sequential learning stages. First, to precisely characterize distinct designs, we leverage transfer learning to extract netlist features directly from placement images. Second, we perform regression learning using various methods to predict the target CTS outcomes and demonstrate that the proposed multitask learning approach achieves better accuracy than the meta-modeling method adopted by previous works. To fully benefit from the predictions made by our framework, we further quantitatively interpret the importance of each CTS input parameter subject to various design objectives through attribution-based learning. Finally, generative adversarial learning is leveraged to optimize the target clock metrics with the guidance provided by the pretrained regression model. To substantiate the generality of our framework, we perform validations on four unseen netlists that are not utilized in the training process. The experimental results conducted on real-world designs demonstrate that our framework: 1) achieves an average prediction error of 3%; 2) improves the commercial tool’s auto-generated clock tree by 20.7% in clock power, 21.5% in clock wirelength, 36.1% in the worst skew; and 3) reaches an F1-score of 0.93 in the classification task of determining successful and failed CTS runs. [ABSTRACT FROM AUTHOR]

Monolithic 3-D (M3-D) integration is an emerging technology that offers significant power, performance, and area benefits for an integrated circuit (IC) design. However, a problem with the 3-D power distribution network in such ICs is that it can lead to high power supply noise (PSN) during the capture cycles in at-speed scan testing for transition delay faults. Therefore, the failure of good chips (i.e., yield loss) resulting from the PSN-induced voltage droop is a major concern for M3-D designs. In this article, we first assess the PSN and voltage droop problems and their impact on path delays for at-speed testing of benchmark M3-D designs. Next, we present an analysis framework to identify test patterns that are most likely to lead to yield loss. We describe a test-pattern reshaping solution based on integer linear programming to make appropriate changes to the test patterns that cause yield loss. Simulation results for four M3-D benchmarks highlight the effectiveness of the proposed solution. [ABSTRACT FROM AUTHOR]

The feasibility of using InGaN/GaN multiple-quantum-well light-emitting diode arrays (LED arrays) as photodiodes (PDs) is investigated experimentally in addition to their light emitting function. Two discrete LED arrays are produced from one 4 × 4 LED array with a parallel-connected pixel configuration. Such compact designs are useful for light emission or detection at the transmitting/receiving terminals of optical wireless communication systems. Despite 4 × 2 LED arrays achieving a light output power of 67.4 mW at 250 mA, they exhibit an optical responsivity (detectivity) of 0.183 A/W (1.61 × 1012 cm Hz1/2W−1) under ultraviolet light illumination (λ = 380 nm) at zero bias. For 4 × 2 LED arrays, the presence of an appreciable ultraviolet light response, together with a high 3-dB bandwidth (∼8 MHz) for modulated light detection, allowed us to build a 15 Mbit/s directed optical link with these LEDs functioning as both the optical transmitter and the receiver. Finally, the unitary LED array-based optical link is capable of real-time transmission of digital audio signals (data rate = 6 Mbit/s) at a propagation distance of 100 cm in free space even though some of the constituent pixels are inactive for light detection. [ABSTRACT FROM AUTHOR]

Brain-computer interface technologies, such as steady-state visually evoked potential, P300, and motor imagery are methods of communication between the human brain and the external devices. Motor imagery-based brain-computer interfaces are popular because they avoid unnecessary external stimuli. Although feature extraction methods have been illustrated in several machine intelligent systems in motor imagery-based brain-computer interface studies, the performance remains unsatisfactory. There is increasing interest in the use of the fuzzy integrals, the Choquet and Sugeno integrals, that are appropriate for use in applications in which fusion of data must consider possible data interactions. To enhance the classification accuracy of brain-computer interfaces, we adopted fuzzy integrals, after employing the classification method of traditional brain-computer interfaces, to consider possible links between the data. Subsequently, we proposed a novel classification framework called the multimodal fuzzy fusion-based brain-computer interface system. Ten volunteers performed a motor imagery-based brain-computer interface experiment, and we acquired electroencephalography signals simultaneously. The multimodal fuzzy fusion-based brain-computer interface system enhanced performance compared with traditional brain-computer interface systems. Furthermore, when using the motor imagery-relevant electroencephalography frequency alpha and beta bands for the input features, the system achieved the highest accuracy, up to 78.81% and 78.45% with the Choquet and Sugeno integrals, respectively. Herein, we present a novel concept for enhancing brain-computer interface systems that adopts fuzzy integrals, especially in the fusion for classifying brain-computer interface commands. [ABSTRACT FROM AUTHOR]

InGaN light-emitting diodes (LEDs) grown with a thin GaN barrier and a gradually reduced well width in the multiple-quantum-well (MQW) regions close to n-GaN are proposed for use in underwater optical wireless communications. Experimentally, LED epiwafers grown by metal–organic vapor phase epitaxy were found to have comparable crystalline quality despite using a different epistructure design from InGaN MQWs. The modified MQWs facilitated hole transport to the deep quantum wells while preventing energetic electrons from accumulating at the topmost quantum wells close to p-GaN, thus improving the current-light output performance at high currents. In addition, relatively strong carrier localization in the proposed LEDs indicates more injected carriers within the modified MQWs can participate in radiative recombinations to contribute to light output. On the other hand, the available bandwidth of a line-of-sight LED-based optical link passing through a 100-cm-long water tank was found to be around 148.7 MHz using the proposed LED, which is $1.3\times $ higher than that obtained using normal LED. This type of optical link produces data transmission at rates in tap water exceeding 300 Mbit/s and enables real-time transmission of digital TV signals. [ABSTRACT FROM AUTHOR]

Visible light communications with InGaN-based light-emitting diodes (LEDs) grown on large-diameter (6-inch) and cost-effective Si (111) substrates are investigated experimentally. During epitaxial growth, the transition layers consisted of the step-graded AlGaN buffers incorporated with three low-temperature-grown (∼900 °C) AlN interlayers on AlN/Si substrates that are used to compensate for thermally induced tensile stress and to maintain a reasonable crystalline quality of GaN-on-Si LEDs. Strong light absorption from Si can be prevented by fabricating a Si substrate-free InGaN LED with a composite metal coating of Al/Ag/Al multilayer, providing improved adhesive strength and reflectivity comparable to the unitary Ag film. In comparison with GaN-on-Si LEDs, stripping Si substrates combined with the use of a highly reflective bottom mirror (Al/Ag/Al multilayer) reflected a more intense emission pattern corresponding to a 2.2 times (@ 190 mA) increase in light output power in thin-film LEDs. In addition, a 1.8 times (@ 160 mA) increase in optical channel bandwidth is achieved by using thin-film LEDs as optical transmitters. A direct line-of-sight optical link using the proposed thin-film LEDs achieved data transmission rates of up to 100 Mb/s over a distance of 100 cm, indicating that the proposed LEDs have potential for use as optical transmitters in indoor visible light communications. [ABSTRACT FROM PUBLISHER]

An experimental evaluation of a directed line-of-sight visible light communication system based on phosphor-converted white resonant-cavity light-emitting diodes (RCLEDs) is presented. With a simple dielectric coating of the transistor outline-can-packaged InGaN LEDs, a $\lambda $ /4-thick Ta2O5/SiO2 distributed Bragg reflector functioning as the top mirror of near-UV RCLEDs can be formed automatically without inducing a registration misalignment as in conventional photolithography. In addition, a passivation coating can also be created over the entire LED structure. As a result of the optical cavity effect, the completed RCLEDs exhibit improved output performance in terms of light output power, external quantum efficiency, and directionality. The resulting RCLEDs are more suitable for building a 200 Mb/s ON–OFF keying nonreturn-to-zero optical link for data transmission over a distance of 100 cm in free space. Finally, we show that real-time transmissions of digital TV signals can be accomplished using the proposed short-range ( $\sim 100$ cm) optical wireless link (data rate $\sim 150$ Mb/s) mainly consisting of a field-programmable gate array-based video encoder/decoder, an optical bandpass filter, and the phosphorescent white RCLEDs made by a simple process. [ABSTRACT FROM PUBLISHER]