Heterogeneous Interconnection Research Articles

The nexus between technological innovation, human capital and energy efficiency: Evidence from E7 countries

Enhancing energy efficiency is very effective in supporting environmental sustainability. Financial inclusion, human capital, and technological innovation are newly cited as dominant factors for environmental sustainability in emerging economies. Against this background, this study investigated how financial inclusion and technological innovation affected energy efficiency in E7 countries from 2004 to 2021. The study identifies the diverse effect of variables at different quantiles through a novel Moments Quantile Regression (MM-QR) method to explore the heterogeneous interconnection among the variables. The MMQR outcomes suggest financial inclusion positively affects energy efficiency across all quantiles. Besides, the findings show that human capital positively affects energy efficiency in E7 countries at lower, medium, and higher quantiles. At the same time, estimated outcomes suggest that technological innovation promotes energy efficiency only at lower quantiles. This study recommends appropriate policies for decision-makers in E7 economies based on the study outcomes regarding adequately channelling human capital, financial resources, and technological innovation to promote energy efficiency and achieve sustainable development goals.

Transmission of COVID-19 in Nightlife, Household, and Health Care Settings in Tokyo, Japan, in 2020

There have been few studies on the heterogeneous interconnection of COVID-19 outbreaks occurring in different social settings using robust, surveillance epidemiological data. To describe the characteristics of COVID-19 transmission within different social settings and to evaluate settings associated with onward transmission to other settings. This is a case series study of laboratory-confirmed COVID-19 cases in Tokyo between January 23 and December 5, 2020, when vaccination was not yet implemented. Using epidemiological investigation data collected by public health centers, epidemiological links were identified and classified into 7 transmission settings: imported, nightlife, dining, workplace, household, health care, and other. The number of cases per setting and the likelihood of generating onward transmissions were compared between different transmission settings. Of the 44 054 confirmed COVID-19 cases in this study, 25 241 (57.3%) were among male patients, and the median (IQR) age of patients was 36 (26-52) years. Transmission settings were identified in 13 122 cases, including 6768 household, 2733 health care, and 1174 nightlife cases. More than 6600 transmission settings were detected, and nightlife (72 of 380 [18.9%]; P < .001) and health care (119 [36.2%]; P < .001) settings were more likely to involve 5 or more cases than dining, workplace, household, and other settings. Nightlife cases appeared in the earlier phase of the epidemic, while household and health care cases appeared later. After adjustment for transmission setting, sex, age group, presence of symptoms, and wave, household and health care cases were less likely to generate onward transmission compared with nightlife cases (household: adjusted odds ratio, 0.03; 95% CI, 0.02-0.05; health care: adjusted odds ratio, 0.57; 95% CI, 0.41-0.79). Household settings were associated with intergenerational transmission, while nonhousehold settings mainly comprised transmission between the same age group. Among 30 932 cases without identified transmission settings, cases with a history of visiting nightlife establishments were more likely to generate onward transmission to nonhousehold settings (adjusted odds ratio, 5.30 [95% CI, 4.64-6.05]; P < .001) than those without such history. In this case series study, COVID-19 cases identified in nightlife settings were associated with a higher likelihood of spreading COVID-19 than household and health care cases. Surveillance and interventions targeting nightlife settings should be prioritized to disrupt COVID-19 transmission, especially in the early stage of an epidemic.

Heterogeneous interconnection of low-loss and dense material platforms using adiabatic tapering coupler

Recently, we successfully realized amorphous silicon carbide (a-SiC) integrated photonics with optical losses as low as 0.78 dB/cm. Moreover, the deposition of a-SiC was done at 150 ℃, which enables successful lift of a-SiC as an additive step to existing photonics circuits. In this work, we present an adiabatic taper coupler which provides bidirectional lossless connection between two integrated photonics platforms: thin-film silicon nitride (Si3N4) and a-SiC. Normalized power transmission of 96.61% is presented, and the coupler enables strong confinement when coupling from weakly confined thin-film device to normal thickness device. By utilizing such a coupler as bridge, switching back and forth between Si3N4 and a-SiC platforms can be easily realized. This allow us to carry out applications including quantum interference and digital Fourier spectroscopy, in which long optical delay lines are constructed on Si3N4 and highly integrated circuits are built on a-SiC.

Room-temperature bonding strategy by ultra-high isostatic pressing for a heterogeneous interconnection architecture

Room-temperature bonding strategy by ultra-high isostatic pressing for a heterogeneous interconnection architecture

Optimization of heterogeneous interconnection transmission based on impedance matching for 3-D IC high-frequency application

Through-silicon via (TSV) is becoming an optimizing interconnection to realize high density, three-dimensional (3-D) integration system. However, its high-frequency application is limited due to the serious parasitic effect, impedance mismatch, crosstalk and so on. In this work, the impedance mismatch characteristic between heterogeneous interconnection is analyzed, modeled and simulated. Interdigital-capacitor based impedance matching structure and a corresponding analytical method are proposed for 3-D IC broadband application. The results obtained from the model and the finite element method (FEM) simulation show that the structure can effectively decrease the transmission losses: the S21 of a matched circuit is improved from −1.4 dB to −0.25 dB at 60 GHz. Furthermore, a polynomial fitting algorithm is proposed to achieve quick and accurate impedance matching and to meet the needs of different application frequency and insertion losses in microwave interconnection systems.

A high-speed and triple-node-upset recovery latch with heterogeneous interconnection

This paper presents an Interconnection heterogeneous and High-speed Triple-node-upset Recovery Latch (IHTRL). This latch utilizes homogeneous elements to construct a 4 × 4 array in four levels. The point is heterogeneous interconnection of each level. Compared with the traditional interconnection way, the proposed heterogeneous interconnection can block the propagation of TNU effectively. The IHTRL can effectively realize TNU recovery by the blocking ability of C-elements and heterogeneous interconnection. The IHTRL utilizes the high-speed path and clock-gating technique to reduce performance penalty and power consumption. Compared with TNU recovery latches LCTNURL and TNURL, the proposed IHTRL is the best in terms of delay, area overhead and PDP. Compared with TNU tolerant and recovery latches, the IHTRL achieves 81.96% reduction in delay on average, 85.91% reduction in PDP on average and 85.46% reduction in APDP on average, at the cost of 7.14% increase in power consumption and 13.21% increase in area overhead on average. Analysis shows that the IHTRL is insensitive to PVT variations.

ACCL: Architecting Highly Scalable Distributed Training Systems With Highly Efficient Collective Communication Library

Distributed systems have been widely adopted for deep neural networks model training. However, the scalability of distributed training systems is largely bounded by the communication cost. We design a highly efficient collective communication library, namely Alibaba Collective Communication Library (ACCL), to build distributed training systems with linear scalability. ACCL provides optimized algorithms to fully make use of heterogeneous interconnects simultaneously. And the experimental results show significant performance improvement.

Campus intelligent security construction based on internet of things

As a hot issue of national and social concern, campus security construction in colleges and universities has some problems such as insufficient information perception and control, as well as insufficient ability of data collection, analysis and sharing, which is difficult to meet the needs of campus security management in the new era. Based on the internet of things technology, comprehensive use of cloud computing, big data and artificial intelligence technologies, we propose a campus intelligent security system framework, then, design the main application functions are specifically, and finally, analyze the typical application scenarios and construction effects. Our campus intelligent security system framework has the characteristics of multi-dimensional perception, heterogeneous interconnection, intelligent and efficient, comprehensive control, centralized presentation and cloud management, which can provide a safe and convenient environment for teachers and students, provide overall situation display and auxiliary decision support for university administrators, and provide comprehensive platform support for campus security management.

Human Microbe-Disease Association Prediction by a Novel Double-Ended Random Walk with Restart.

Microorganisms in the human body play a vital role in metabolism, immune defense, nutrient absorption, cancer control, and prevention of pathogen colonization. More and more biological and clinical studies have shown that the imbalance of microbial communities is closely related to the occurrence and development of various complex human diseases. Finding potential microbial-disease associations is critical for understanding the pathology of a few diseases and thus further improving disease diagnosis and prognosis. In this study, we proposed a novel computational model to predict disease-associated microbes. Specifically, we first constructed a heterogeneous interconnection network based on known microbe-disease associations deposited in a few databases, the similarity between diseases, and the similarity between microorganisms. We then predicted novel microbe-disease associations by a new method called the double-ended restart random walk model (DRWHMDA) implemented on the interconnection network. In addition, we performed case studies of colon cancer and asthma for further evaluation. The results indicate that 10 and 9 of the top 10 microorganisms predicted to be associated with colorectal cancer and asthma were validated by relevant literatures, respectively. Our method is expected to be effective in identifying disease-related microorganisms and will help to reveal the relationship between microorganisms and complex human diseases.

Mobile Edge Computing and Networking for Green and Low-Latency Internet of Things

IoT, a heterogeneous interconnection of smart devices, is a great platform to develop novel mobile applications. Resource constrained smart devices, however, often become the bottlenecks to fully realize such developments, especially when it comes to intensive-computation-oriented and low-latency-demanding applications. MEC is a promising approach to address such challenges. In this article, we focus on MEC applications for IoT, and address energy efficiency as well as offloading performance of such applications in terms of end-user experience. In this regard, we present a mobility-aware hierarchical MEC framework for green and low-latency IoT. We deploy a game theoretic approach for computation offloading in order to optimize the utility of the service providers while also reducing the energy cost and the task execution time of the smart devices. Numerical results indicate that the proposed scheme does brings significant enhancement in both energy efficiency and latency performance of MEC applications for IoT.

(Invited) Diverse Accessible Heterogeneous Integration (DAHI) Foundry at Northrop Grumman Aerospace Systems (NGAS)

The requirements for future systems are increasingly demanding not only form the point of view of performance, but also from size, weight, power and cost (SWAP-C) considerations. This translates into the need for advanced integration of additional complexity and performance of electronic functions into smaller areas and volumes. In addition, to enable higher performance of the microelectronic technology in which these functions are implemented, it is not enough to use one single semiconductor technology but to combine or integrate several high performance technologies in an efficient and cost effective way. In most cases this involves the integration of semiconductor technologies based on different substrates or different technology nodes based on the same substrate. The combination of two or more dissimilar microelectronic technologies, or heterogeneous integration, leads to significant higher design flexibility and capability. The more adaptable and more intimate is this heterogeneous integration between two or more technologies, the more flexibility is given to the designer for the selection of the technology for a specific function, or even better, for an optimum combination of different transistor technologies in the same function or cell in the design. Northrop Grumman Aerospace Systems (NGAS) under the Diverse Accessible Heterojunction Integration (DAHI) DARPA program is developing heterogeneous integration processes, design kits and thermal simulation tools to integrate submicron CMOS, InP HBT, GaN HEMT and high-Q passive technologies for advanced DoD and other government systems. We have demonstrated integration of NGAS’ TF4 and TF5 InP HBT and GaN20 GaN on SiC HEMT technologies on IBM’s 65nm CMOS and GlobalFoundries’ (GFUS) 45nm CMOS (12SOI). The NGAS DAHI integration process has several advantages. It is scalable to 200 mm and 300 mm CMOS wafers and fully compatible with all Si technologies. All the technologies to be integrated are independently fabricated in parallel for optimum cycle time and decoupled line yield. The NGAS DAHI integration approach also offers advantages from the thermal management point of view. The low thermal resistance heterogeneous interconnects (HICs) enable flexible electrical and thermal routing. The GaN HEMT is integrated with its backside facing the CMOS front side to allow better intra-chip heat spreading and large inter-chip HIC heat sink arrays. The thickness of the Silicon Carbide (SiC) substrate in the GaN HEMT chiplet is designed to be thick enough to allow optimum lateral heat spreading before reaching the thermal HIC interface. The TF InP HBT technology is integrated with its front side facing the CMOS front side for maximum density of interconnects. An additional important task in the development of the DAHI technology is the creation of an integrated DAHI process design kit (PDK). The objective of the DAHI PDK development is to have a flexible design environment that can accommodate schematics, layouts and simulations of integrated DAHI circuits with device-level integration designs of a wide variety of technologies. Integration circuit results will be presented at the conference. Acknowledgements This work is supported by DARPA DAHI Program under AFRL contract No. FA8650-13-C-7312, Daniel Green DARPA program manager and Robert Fitch AFRL COTR. The views expressed are those of the authors and do not reflect the official policy or position of the Department of Defense or the U.S. Government.

An Approximate Transfer Function Model of Two Serially Connected Heterogeneous Transmission Lines

This brief presents an approximate transfer function model of two heterogeneous transmission lines connected in series. The proposed model is simple, intuitive, and can accurately describe the frequency response of various serially connected heterogeneous interconnects which satisfy a validity condition. The exact error of the proposed model is also provided. According to our analysis, the validity condition can be clearly interpreted by the propagation and the reflection waves. The accuracy of the proposed model is verified by comparing the calculated transfer function of LC-dominant off-chip and RC-dominant on-chip examples against SPICE simulation results. In these verifications, the maximum errors between calculation and simulation are, respectively, 2.58% and 3.32% for off-chip and on-chip examples. The computation time is also reduced about 29.76 and 289.07 times compared with a single-channel model and the SPICE ${w}$ -element model. With the proposed model, designers can easily understand the channel responses of diverse serially connected heterogeneous interconnects.

Hybrid Optical Wireless Network for Future SAGO-Integrated Communication Based on FSO/VLC Heterogeneous Interconnection

Free-space optics (FSO) and visible light communication (VLC) share the same optical wireless features, such as high data rate, license free, energy efficient, flexible access, and high degree of security. A hybrid optical wireless network based on FSO/VLC heterogeneous interconnection is presented here for future space-air-ground-ocean-integrated communication architecture, especially in the radio-frequency-sensitive (RF) or security-required environments. The functional modules of the hybrid network coordinator are defined, along with the implementation and deployment details. Three fundamental network-layer mechanisms are designed in the coordinator including user identification and localization, user mobility and handoff control, and routing and traffic management. An experimental platform is built to evaluate the transmission performance of the presented hybrid network, and a complete data aggregation/transmission/distribution procedure based on heterogeneous interconnection is first realized, which includes two-segment 1.0-m 450-Mb/s orthogonal-frequency-division-multiplexing-based VLC transmissions interconnected by one-segment 430-m 0.96-Gb/s on-off keying (OOK)-based nonturbulent FSO transmission, with the bit error rate under the forward-error-correction limit (3.8 × 10−3 ). The VLC transmission performances under three speed levels and FSO transmission performances under five typical air-quality conditions are experimentally demonstrated, which validate the feasibility of the proposed hybrid optical wireless network.

(Invited) Diverse Accessible Heterogeneous Integration (DAHI) Foundry at Northrop Grumman Aerospace Systems (NGAS)

The requirements for future systems are increasingly demanding not only form the point of view of performance, but also from size, weight, power and cost (SWAP-C) considerations. This translates into the need for advanced integration of additional complexity and performance of electronic functions into smaller areas and volumes. In addition, to enable higher performance of the microelectronic technology in which these functions are implemented, it is not enough to use one single semiconductor technology but to combine or integrate several high performance technologies in an efficient and cost effective way. In most cases this involves the integration of semiconductor technologies based on different substrates. The combination of two or more dissimilar microelectronic technologies, or heterogeneous integration, leads to significant higher design flexibility and capability.The more adaptable and more intimate is this heterogeneous integration between two or more technologies, the more flexibility is given to the designer for the selection of the technology for a specific function, or even better, for an optimum combination of different transistor technologies in the same function or cell in the design. Northrop Grumman Aerospace Systems (NGAS) under the Diverse Accessible Heterojunction Integration (DAHI) DARPA program is developing heterogeneous integration processes, design kits and thermal simulation tools to integrate submicron CMOS, InP HBT, GaN HEMT and high-Q passive technologies for advanced DoD and other government systems. We have demonstrated integration of NGAS’ TF4 and TF5 InP HBT and GaN20 GaN on SiC HEMT technologies on IBM’s 65nm CMOS and GlobalFoundries’ 45nm SOI. The NGAS DAHI integration process has several advantages. It is scalable to 200 mm and 300 mm CMOS wafers and fully compatible with all Si technologies. All the technologies to be integrated are independently fabricated in parallel for optimum cycle time and decoupled line yield. The NGAS DAHI integration approach also offers advantages from the thermal management point of view. The low thermal resistance heterogeneous interconnects (HICs) enable flexible electrical and thermal routing. The GaN HEMT is integrated with its backside facing the CMOS front side to allow better intra-chip heat spreading and large inter-chip HIC heat sink arrays. The thickness of the Silicon Carbide (SiC) substrate in the GaN HEMT chiplet is designed to be thick enough to allow optimum lateral heat spreading before reaching the thermal HIC interface. The TF InP HBT technology is integrated with its front side facing the CMOS front side for maximum density of interconnects. An additional very important task in the development of the DAHI technology is the creation of an integrated DAHI process design kit (PDK). The objective of the DAHI PDK development is to have a flexible design environment that can accommodate schematics, layouts and simulations of integrated DAHI circuits with device-level integration designs of a wide variety of technologies. Integration circuit results will be presented at the conference.

Electrothermal Cosimulation of 3-D Carbon-Based Heterogeneous Interconnects

Electrothermal characteristics of some novel 3-D carbon-based heterogeneous interconnects, consisting of vertical carbon nanotube bundle via and horizontal multilayer graphene, are investigated by utilizing in-house developed algorithm based on a finite-element method. With present fabrication capability, these heterogeneous interconnects can have larger electrical resistance but smaller thermal resistance in comparison with their Cu counterpart. Both the local on-chip interconnects for ballistic regime and global through-silicon via channel for diffusive regime are evaluated numerically, and their 3-D transient temperature distribution and hot spots are characterized and compared. During the electrothermal cosimulation, the anisotropic property of electrical and thermal conductivities of carbon nanomaterials is treated in an appropriate way. It is believed that this paper will be useful for the design as well as the realization of new generation carbon-based interconnects with high reliability and better thermal performance.

Architecture of unified network based on programmable control gateway

Due to the long-term coexistence of Internet Protocol( IP) and non-IP networks,there exist interconnection and interworking difficulties in heterogeneous networks for the difference of network architectures and protocols. Focusing on the gateway,the key node of heterogeneous networks interconnection,a unified network architecture based on programmable gateway is proposed to solve this problem,which separated the control plane from the data forwarding plane and extracted the combination of the management and control functionality and application service. The unified network architecture divides the network into three layers,which were unified network collaborative management and control unit,programmable gateway,and heterogeneous subnet. The interconnection protocol stack for the typical heterogeneous networks is designed to integrate different protocols. Then the key techniques and theory for the programmable gateway are given and analyzed based on the unified network architecture,including the abstraction and description of heterogeneous networks,intelligent adaptation technology based on programmable gateway,resource management and optimization of unified network,etc. Finally,a typical application case of interconnection networks in civil aviation is given as an illustration to demonstrate the proposed network architecture.

Оптимизация отображения неоднородно взаимодействующих MPI процессов на вычислительную архитектуру

The problem of mapping the parallel task to the nodes of computing cluster is considered. MPI software with non-uniform communication and heterogeneous interconnect of computing cluster require to appropriate parallel processes mapping for optimization of data exchange. The graph mapping algorithm is developed. It uses parallel program representation as a task graph and cluster topology representation as system graph. The proposed optimization technique is tested on synthetic benchmark and on real QBox software to study its efficiency on large number of computing cores. The positive results of optimization are achieved and the summary is presented in the paper. Speedup of 17–20% is obtained on scalable benchmarks using 300–4800 parallel processes.

High‐frequency modeling of Cu‐graphene heterogeneous interconnects

SummaryOne novel interconnect scheme consisting of both Cu and graphene sheet is proposed in this paper, with the advantages of both materials exploited greatly. It is shown that the introduction of graphene layers in such heterogeneous interconnect scheme can reduce its effective resistance and thereby improve its transmission performance. On the other hand, it is also demonstrated that both coated and double‐coated structures possess better electrical performance than that of the sandwich one at high frequencies, because the graphene is placed at the interconnect surface where current is crowded. With the help of Partial Element Equivalent Circuit method, together with equivalent circuit technique, the transmission characteristics of some Cu‐graphene interconnects are captured and compared with that of Cu wire, and the advantages of such heterogeneous interconnects can be enlarged with the advanced technology. Copyright © 2015 John Wiley &amp; Sons, Ltd.

Electrothermal Characterization of Multilevel Cu-Graphene Heterogeneous Interconnects in the Presence of an Electrostatic Discharge (ESD)

Temperature responses of multilevel Cu-graphene heterogeneous interconnects under the impact of an electrostatic discharge are investigated by using our self-developed time-domain finite-element method algorithm. Corresponding to the advanced CMOS processes, all parameters of such multilevel interconnects are assessed by the ITRS. It is numerically shown that when capped with 10-nm-thick multilayer graphene, the maximum temperature of the Cu-graphene interconnect could be suppressed by 45% and 30% for 13.4- and 21-nm nodes, respectively. This study could be useful for improving the reliability of interconnects in the future nanoscale integrated circuits.

Exploring Heterogeneous NoC Design Space in Heterogeneous GPU-CPU Architectures

Computer architecture is transiting from the multicore era into the heterogeneous era in which heterogeneous architectures use on-chip networks to access shared resources and how a network is configured will likely have a significant impact on overall performance and power consumption. Recently, heterogeneous network on chip (NoC) has been proposed not only to achieve performance comparable to that of the NoCs with buffered routers but also to reduce buffer cost and energy consumption. However, heterogeneous NoC design for heterogeneous GPU-CPU architectures has not been studied in depth. This paper first evaluates the performance and power consumption of a variety of static hot-potato based heterogeneous NoCs with different buffered and bufferless router placements, which is helpful to explore the design space for heterogeneous GPU-CPU interconnection. Then it proposes Unidirectional Flow Control (UFC), a simple credit-based flow control mechanism for heterogeneous NoC in GPU-CPU architectures to control network congestion. UFC can guarantee that there are always unoccupied entries in buffered routers to receive flits coming from adjacent bufferless routers. Our evaluations show that when compared to hot-potato routing, UFC improves performance by an average of 14.1% with energy increased by an average of 5.3% only.