Application-specific Optimization Research Articles

TRANSIENT ANALYSIS OF LIQUID-HYDROGEN TRANSFER IN CRYOGENIC STORAGE TANKS

Efficient transfer of liquid hydrogen is critical for a myriad of processes in terrestrial and extraterrestrial applications. The objective of this study is to model and quantify the amount of liquid parahydrogen vaporized during a discharging/charging process in a cryogenic storage system. The effects that are included include geometrical effects (storage tank volume), two-phase compressible flow, choking/unchoking, and wall-fluid heat transfer. A brief description of other physical boil-off mechanisms is also presented. This study provides a foundation for applicationspecific optimization. The storage system studied comprises two interconnected tanks whose transfer line has a variation in the flow area. This simulates current liquid-hydrogen storage and transportation systems. The model tracks the temperature- and pressure-time histories of liquid hydrogen as it flows between storage tanks, in both the choked and unchoked flow regimes. The transfer of liquid is induced solely by the pressure differential that exists between the two storage tanks. An iterative technique helps account for choking, which is likely to exist at the throat in two-phase flow. Cases of zero and infinite tank-wall thermal mass are also discussed. By analyzing the behavior of fundamental variables during the transient transfer of liquid hydrogen, boil-off losses may be minimized if the variables with the greatest effect on boil-off losses are controlled.

A comparative survey of the influence of small self-cleaving ribozymes on gene expression in human cell culture

ABSTRACT Self-cleaving ribozymes are versatile tools for synthetic biologists when it comes to controlling gene expression. Up to date, 12 different classes are known, and over the past decades more and more details about their structure, cleavage mechanisms and natural environments have been uncovered. However, when these motifs are applied to mammalian gene expression constructs, the outcome can often be unexpected. A variety of factors, such as surrounding sequences and positioning of the ribozyme influences the activity and hence performance of catalytic RNAs. While some information about the efficiency of individual ribozymes (each tested in specific contexts) is known, general trends obtained from standardized, comparable experiments are lacking, complicating decisions such as which ribozyme to choose and where to insert it into the target mRNA. In many cases, application-specific optimization is required, which can be very laborious. Here, we systematically compared different classes of ribozymes within the 3’-UTR of a given reporter gene. We then examined position-dependent effects of the best-performing ribozymes. Moreover, we tested additional variants of already widely used hammerhead ribozymes originating from various organisms. We were able to identify functional structures suited for aptazyme design and generated highly efficient hammerhead ribozyme variants originating from the human genome. The present dataset will aide decisions about how to apply ribozymes for affecting gene expression as well as for developing ribozyme-based switches for controlling gene expression in human cells.

Model-based design for self-sustainable sensor nodes

Long-term and maintenance-free operation is a critical feature for large-scale deployed battery-operated sensor nodes. Energy harvesting (EH) is the most promising technology to overcome the energy bottleneck of today’s sensors and to enable the vision of perpetual operation. However, relying on fluctuating environmental energy requires an application-specific analysis of the energy statistics combined with an in-depth characterization of circuits and algorithms, making design and verification complex. This article presents a model-based design (MBD) approach for EH-enabled devices accounting for the dynamic behavior of components in the power generation, conversion, storage, and discharge paths. The extension of existing compact models combined with data-driven statistical modeling of harvesting circuits allows accurate offline analysis, verification, and validation. The presented approach facilitates application-specific optimization during the development phase and reliable long-term evaluation combined with environmental datasets. Experimental results demonstrate the accuracy and flexibility of this approach: the model verification of a solar-powered wireless sensor node shows a determination coefficient (R2) of 0.992, resulting in an energy error of only -1.57% between measurement and simulation. Compared to state-of-practice methods, the MBD approach attains a reduction of the estimated state-of-charge error of up to 10.2% in a real-world scenario. MBD offers non-trivial insights on critical design choices: the analysis of the storage element selection reveals a 2–3 times too high self-discharge per capacity ratio for supercapacitors and a peak current constrain for lithium-ion polymer batteries.

Object Intersection Captures on Interactive Apps to Drive a Crowd-sourced Replay-based Compiler Optimization

Traditional offline optimization frameworks rely on representative hardware, software, and inputs to compare different optimizations on. With application-specific optimization for mobile systems though, the idea of a representative testbench is unrealistic while creating offline inputs is non-trivial. Online approaches partially overcome these problems but they might expose users to suboptimal or even erroneous code. Therefore, our mobile code is poorly optimized, resulting in wasted performance and energy and user frustration. In this article, we introduce a novel compiler optimization approach designed for mobile applications. It requires no developer effort, it tunes applications for the user’s device and usage patterns, and it has no negative impact on the user experience. It is based on a lightweight capture and replay mechanism. Our previous work [ 46 ] captures the state accessed by any targeted code region during its online stage. By repurposing existing OS capabilities, it keeps the overhead low. In its offline stage, it replays the code region but under different optimization decisions to enable sound comparisons of different optimizations under realistic conditions. In this article, we propose a technique that further decreases the storage sizes without any additional overhead. It captures only the intersection of reachable objects and accessed heap pages. We compare this with another new approach that has minimal runtime overheads at the cost of higher capture sizes. Coupled with a search heuristic for the compiler optimization space, our capture and replay mechanism allows us to discover optimization decisions that improve performance without testing these decisions directly on the user. Finally, with crowd-sourcing we split this offline evaluation effort between several users, allowing us to discover better code in less time. We implemented a prototype system in Android based on LLVM combined with a genetic search engine and a crowd-sourcing architecture. We evaluated it on both benchmarks and real Android applications. Online captures are infrequent and introduce ~5 ms or 15 ms on average, depending on the approach used. For this negligible effect on user experience, we achieve speedups of 44% on average over the Android compiler and 35% over LLVM -O3. Our collaborative search is just 5% short of that speedup, which is impressive given the acceleration gains. The user with the highest workload concluded the search 7 \( \times \) faster.

DirectNVM: Hardware-accelerated NVMe SSDs for High-performance Embedded Computing

With data-intensive artificial intelligence (AI) and machine learning (ML) applications rapidly surging, modern high-performance embedded systems, with heterogeneous computing resources, critically demand low-latency and high-bandwidth data communication. As such, the newly emerging NVMe (Non-Volatile Memory Express) protocol, with parallel queuing, access prioritization, and optimized I/O arbitration, starts to be widely adopted as a de facto fast I/O communication interface. However, effectively leveraging the potential of modern NVMe storage proves to be nontrivial and demands fine-grained control, high processing concurrency, and application-specific optimization. Fortunately, modern FPGA devices, capable of efficient parallel processing and application-specific programmability, readily meet the underlying physical layer requirements of the NVMe protocol, therefore providing unprecedented opportunities to implementing a rich-featured NVMe middleware to benefit modern high-performance embedded computing. In this article, we present how to rethink existing accessing mechanisms of NVMe storage and devise innovative hardware-assisted solutions to accelerating NVMe data access performance for the high-performance embedded computing system. Our key idea is to exploit the massively parallel I/O queuing capability, provided by the NVMe storage system, through leveraging FPGAs’ reconfigurability and native hardware computing power to operate transparently to the main processor. Specifically, our DirectNVM system aims at providing effective hardware constructs for facilitating high-performance and scalable userspace storage applications through (1) hardening all the essential NVMe driver functionalities, therefore avoiding expensive OS syscalls and enabling zero-copy data access from the application, (2) relying on hardware for the I/O communication control instead of relying on OS-level interrupts that can significantly reduce both total I/O latency and its variance, and (3) exposing cutting-edge and application-specific weighted-round-robin I/O traffic scheduling to the userspace. To validate our design methodology, we developed a complete DirectNVM system utilizing the Xilinx Zynq MPSoC architecture that incorporates a high-performance application processor (APU) equipped with DDR4 system memory and a hardened configurable PCIe Gen3 block in its programmable logic part. We then measured the storage bandwidth and I/O latency of both our DirectNVM system and a conventional OS-based system when executing the standard FIO benchmark suite [ 2 ]. Specifically, compared against the PetaLinux built-in kernel driver code running on a Zynq MPSoC, our DirectNVM has shown to achieve up to 18.4× higher throughput and up to 4.5× lower latency. To ensure the fairness of our performance comparison, we also measured our DirectNVM system against the Intel SPDK [ 26 ], a highly optimized userspace asynchronous NVMe I/O framework running on a X86 PC system. Our experiment results have shown that our DirectNVM, even running on a considerably less powerful embedded ARM processor than a full-scale AMD processor, achieved up to 2.2× higher throughput and 1.3× lower latency. Furthermore, by experimenting with a multi-threading test case, we have demonstrated that our DirectNVM’s weighted-round-robin scheduling can significantly optimize the bandwidth allocation between latency-constraint frontend applications and other backend applications in real-time systems. Finally, we have developed a theoretical framework of performance modeling with classic queuing theory that can quantitatively define the relationship between a system’s I/O performance and its I/O implementation.

Development and Clinical Validation of a Finite Element Method Model Mapping Focal Intracranial Cooling.

Therapeutic hypothermia (TH) is a common and effective technique to reduce inflammation and induce neuroprotection across a variety of diseases. Focal TH of the brain can avoid the side effects of systemic cooling. The degree and extent of focal TH are a function of cooling probe design and local brain thermoregulation processes. To refine focal TH probe design, with application-specific optimization, we develop precise computational models of brain thermodynamics under intense local cooling. Here, we present a novel multiphysics in silico model that can accurately predict brain response to focal cooling. The model was parameterized from previously described values of metabolic activity, thermal conductivity, and temperature-dependent cerebral perfusion. The model was validated experimentally using data from clinical cases where local cooling was induced intracranially and brain temperatures monitored in real-time with MR thermometry. The validated model was then used to identify optimal design probe parameters to maximize volumetric TH, including considering three stratifications of cooling (mild, moderate, and profound) to produce Volume of Tissue Cooled (VOTC) maps. We report cooling radius increases in a nearly linear fashion with probe length and decreasing probe surface temperature.

Polarization switching in Hf0.5Zr0.5O2-dielectric stack: The role of dielectric layer thickness

Understanding the role of the dielectric (DE) layer in ferroelectric (FE) Hf0.5Zr0.5O2 (HZO) based devices (e.g., ferroelectric-field-effect-transistors, FE-FETs) is important to enable their application-driven optimizations. To that end, in this work, we systematically investigate the polarization switching mechanisms in FE–DE stacks and analyze their dependence on the dielectric layer thickness (TDE). First, we fabricate a HZO–Al2O3 (FE–DE) stack and experimentally demonstrate a decrease in remanent polarization and an increase in coercive voltage with an increase in TDE. As such dependencies are out of the scope of commonly used single domain polarization switching models, therefore, we argue that the consideration of the multi-domain model is essential for analyzing the polarization switching in HZO. Then, using phase-field simulations of the FE–DE stack, we show that an increase in TDE results in a larger number of reverse domains in the FE layer to suppress the depolarization field, which leads to a decrease in the remanent polarization and an increase in the coercive voltage. Furthermore, our analysis signifies that the polarization switching mechanism in HZO can be modulated from domain-nucleation based to domain-wall motion based by increasing the TDE and that can serve as a potential knob for application-specific optimization of FE-FETs. In addition, we show that the effective polarization–voltage characteristics of the FE layer in the FE–DE stack exhibit a negative slope region that leads to the charge enhancement effects in the FE–DE stack. While such effects are most commonly misinterpreted as either the transient effects or the stabilized single-domain negative capacitance effects, we demonstrate that the appearance of a negative slope in the hysteretic polarization–voltage characteristics is quasi-static in nature and that originates from the multi-domain polarization switching in the FE.

Real-parameter constrained optimization using enhanced quality-based cultural algorithm with novel influence and selection schemes

Hybridization in context to Evolutionary Computation (EC) strives to combine operators, components, and the best merits of different EC paradigms, to form a new evolutionary algorithm that enjoys a statistically superior performance, compared to its ancestors, over a wide range of application-specific optimization problems. In this paper, we propose a simple yet powerful amalgam composed of a modified Cultural Algorithm (CA) that is supported with an Enhanced Levy Flight Search (ELFS) to guide the search and further promote the harmony between the explorative and exploitative capacities of the conventional techniques. The novel amalgam, denoted by q-aCA + mIS, utilizes a balanced search scheme where it employs an adapted Influence Function (IF) with a novel quality function that establishes a harmony between the Knowledge Sources (KSs) in the Belief Space (BS), and between the BS and other components in the hybrid to produce the most suitable knowledge needed for a certain search mode. The CA framework is reinforced with an updated Selection Function (SF) that employs a successful selection strategy that uses the extended situational knowledge for the future selection of individuals. The proposed algorithm is tested using more than 50 benchmark functions that are taken from the IEEE CEC’06, and the IEEE CEC’19 competitions on constrained real-parameter optimization. Moreover, three well-known engineering design problems are used to test the validity of the algorithm for the solution of complex real-life problems. The comparative study indicates that the q-aCA + mIS algorithm was able to obtain a statistically superior performance and scalability behavior over most of the considered functions in comparison with other state-of-the-art algorithms.

Optimization of 5G Networks for Smart Logistics

Industry 4.0 is generalizing the use of wireless connectivity in manufacturing and logistics. Specifically, in Smart Logistics, novel Industry 4.0 technologies are used to enable agile supply chains, with reduced management, energy and storage costs. Cellular networks allow connectivity throughout all the scenarios where logistics processes take place, each having their own challenges. This paper explores such scenarios and challenges, and proposes 5G technology as a global unified connectivity solution. Moreover, this paper proposes a system for exploiting the application-specific optimization capabilities of 5G networks to better cater for the needs of Smart Logistics. An application traffic modeling process is proposed, along with a proactive approach to network optimization that can improve the Quality of Service and reduce connectivity costs.

Cost-Effective App User Allocation in an Edge Computing Environment

Edge computing is a new distributed computing paradigm extending the cloud computing paradigm, offering much lower end-to-end latency, as real-time, latency-sensitive applications can now be deployed on edge servers that are much closer to end-users than distant cloud servers. In edge computing, edge user allocation (EUA) is a critical problem for any app vendors, who need to determine which edge servers will serve which users. This is to satisfy application-specific optimization objectives, e.g., maximizing users’ overall quality of experience, minimizing system costs, and so on. In this article, we focus on the cost-effectiveness of user allocation solutions with two optimization objectives. The primary one is to maximize the number of users allocated to edge servers. The secondary one is to minimize the number of required edge servers, which subsequently reduces the operating costs for app vendors. We first model this problem as a bin packing problem and introduce an approach for finding optimal solutions. However, finding optimal solutions to the <inline-formula><tex-math notation="LaTeX">$\mathcal {NP}$</tex-math></inline-formula> -hard EUA problem in large-scale scenarios is intractable. Thus, we propose a heuristic to efficiently find sub-optimal solutions to large-scale EUA problems. Extensive experiments conducted on real-world data demonstrate that our heuristic can solve the EUA problem effectively and efficiently, outperforming the state-of-the-art and baseline approaches.

RF magnetron sputtered AlCoCrCu0.5FeNi high entropy alloy (HEA) thin films with tuned microstructure and chemical composition

Abstract High entropy alloy (HEA) thin films are extensively studied recently due to their propensity to display a combination of excellent properties such as high hardness, hydrophobicity, superior oxidation and corrosion resistance. Sputter deposition of thin films comprised of several elements typically requires the use of targets containing multiple elements, making both the stoichiometry and microstructure of the resulting films strongly dependent on the process parameters. In this study, HEA thin films of AlCoCrCu0.5FeNi were deposited at various radio frequency (RF) powers, from 200 to 300 W, to reveal how the power affects composition, microstructure and physical properties. X-ray diffraction (XRD) indicated a mixed phase of FCC and BCC and transmission electron microscopy (TEM) data revealed the presence of amorphous phases between the grain boundaries. Scanning electron microscopy (SEM) and atomic force microscopy (AFM) respectively showed an increase in grain size and surface roughness with increasing RF power. X-ray photoelectron spectroscopy (XPS) confirmed the presence of Al2O3 and Cr2O3 protective oxide layers on the surface of all the HEA films. The highest concentration of Al (19.7%) was detected by energy dispersive spectroscopy (EDS) in the film deposited at the lowest power, which also had the highest nanoindentation hardness of 13 GPa. The hardness dropped to 4.5 GPa as power increased due to a reduction in Al concentration and an increase in average grain size. The film deposited at the highest power showed the highest hydrophobicity with a contact angle of 129°, while the lowest resistivity of 414 μΩ-cm was recorded for the sample deposited with the lowest power. This study shows the influence of deposition power on the microstructure and composition of the HEA thin films, demonstrating that the power can be an effective processing parameter to control various properties of AlCoCrCu0.5FeNi films. An understanding of how key process parameters affect the resulting thin film properties will enable application specific process optimization.

Optimal Design of Timed Antenna Arrays for SLL Reduction, Dual and Multiple Broad Nulls in the Radiation Pattern

This work is concerned with the methodology for improvement of the radiation pattern performance and eliminating the effect of unwanted interference of symmetric time modulated linear antenna array (STMLAA) in the predefined direction. This is achieved using evolutionary technique called Novel Particle Swarm Optimization with Wavelet Mutation (NPSOWM) for various application specific optimization like SLL only reduction, dual broad null and multiple broad null optimization. The optimized results obtained by using NPSOWM are validated by CST-MWS by captivating practical elements of dipole antenna and its array configuration. The numerical results show the effect of the particular (practical) antenna on the array. Various examples have taken for this purpose. A set of 16-element antenna array have designed for SLL only reduction. 16-, 32-, and 64-element STMLAA for incorporating the dual broad nulls in the angular range from 400 to 700 over the elevation plane. A set of 24-, 30-, and 32-element antenna array have occupied for multiple broad nulls. The strength of the adopted algorithm is tested for different number of array antenna configuration, which establish the efficacy of the optimization technique NPSOWM adopted for achieving the interference cancellation in the predefined direction and performance improvement of the antenna array such as simultaneous improvements of broad nulls (BNs), sidelobe level (SLL), sideband level (SBL), and the directivity. From the simulated results it is clear that NPSOWM can perform better than that of un-optimized linear antenna array with respect to BNs, SLL, and SBLs and Directivity.

Application and Product-Volume-Specific Customization of BEOL Metal Pitch

High-volume manufacturing of integrated circuits is what drives the semiconductor industry and the scaling of CMOS; however, shrinking all feature sizes is not optimal for all product volumes and applications. While scaling transistors provide improvement in performance at power in general, concerted scaling of the back-end-of-line (BEOL) interconnects provides improvement in density but at some expense of performance and quality. In this paper, we explore the benefits of relaxing the BEOL metal pitches as a tunable knob for application-specific optimization as a function of product volume. This paper demonstrates that a relaxed BEOL (r-BEOL) methodology can be applied at block level as well as chip level to provide performance and quality benefits while considering density constraints. The benefits of an r-BEOL pitch for power, timing, design for manufacturability, routing congestion, and signal integrity are considered for a 16-nm FinFET technology. Silicon results using relaxed M4–M7 metal layers are used to validate the simulation comparisons.

Optimized multi-shot imaging inspection design.

The inspection of complex-shaped components, such as those enabled by additive manufacturing, is a major challenge in industrial quality assurance. A frequently adopted approach to volumetric non-destructive evaluation is X-ray computed tomography, but this has major drawbacks. Two-dimensional radiography can overcome some of these problems, but does not generally provide an inspection that is as capable. Moreover, designing a detailed inspection for a complex-shaped component is a labour-intensive task, requiring significant expert input. In response, a computational framework for optimizing the data acquisition for an image-based inspection modality has been devised. The initial objective is to advance the capabilities of radiography, but the algorithm is, in principle, also applicable to alternative types of imaging. The algorithm exploits available prior information about the inspection and simulations of the inspection modality to allow the determination of the optimal inspection configuration, including specifically component poses with respect to the imaging system. As an intermediate output, spatial maps of inspection performance are computed, for understanding spatially varying limits of detection. Key areas of innovation concern the defect detectability evaluation for arbitrarily complex indications and the creation of an application-specific optimization algorithm. Initial trials of the algorithm are presented, with good results.

Functional cargo delivery into mouse and human fibroblasts using a versatile microfluidic device

Efficient intracellular cargo delivery is a key hurdle for the translation of many emerging stem cell and cellular reprogramming therapies. Recently, a microfluidic-based device constructed from silicon was shown to transduce macromolecules into cells via shear-induced formation of plasma membrane pores. However, the scalability and widespread application of the current platform is limited since physical deformation-mediated delivery must be optimized for each therapeutic application. Therefore, we sought to create a low-cost, versatile device that could facilitate rapid prototyping and application-specific optimization in most academic research labs. Here we describe the design and implementation of a microfluidic device constructed from Polydimethylsiloxane (PDMS) that we call Cyto-PDMS (Cytoplasmic PDMS-based Delivery and Modification System). Using a systematic Cyto-PDMS workflow, we demonstrate intracellular cargo delivery with minimal effects on cellular viability. We identify specific flow rates at which a wide range of cargo sizes (1-70kDa) can be delivered to the cell interior. As a proof-of-principle for the biological utility of Cyto-PDMS, we show (i) F-actin labeling in live human fibroblasts and (ii) intracellular delivery of recombinant Cre protein with appropriate genomic recombination in recipient fibroblasts. Taken together, our results demonstrate that Cyto-PDMS can deliver small-molecules to the cytoplasm and biologically active cargo to the nucleus without major effects on viability. We anticipate that the cost and versatility of PDMS can be leveraged to optimize delivery to a broad array of possible cell types and thus expand the potential impact of cellular therapies.

Reconfigurable NanoFETs: Performance Projections for Multiple-Top-Gate Architectures

In nanowire or nanotube field-effect transistors (nanoFETs) electrostatic doping can be induced by electrical fields originating from multiple independent gates. Therefore, nanoFETs are predestined to explore reconfigurability. Solving the coupled nonlinear Poisson and drift-diffusion differential equations for the three-dimensional electrostatic potential and the one-dimensional channel charge, respectively, we predict the performance of four different reconfigurable (R) nanoFET geometries. The investigated architectures compass FETs with one (1G), two (2G), and three top-gate (3G) terminals with a moderate channel length of half a micrometer. Therefore, the theoretically investigated R-nanoFETs can be manufactured at low costs, allowing to test the performance projections claimed in this paper. The 2G R-nanoFET proved to be the most versatile architecture when no application specific optimization is attempted. However, all considered geometries offer interesting properties. Shortening the program gate with the drain simplifies the local routing and only slightly diminish the performance. A smaller footprint 1G R-nanoFET delivers comparable intrinsic gains at the cost of increased static power dissipation. Finally, a 3G R-nanoFET enables additional dynamic configuration options and faster on/off switching due to a control gate positioned at an increased distance to other metallic contacts.

(Invited) Beyond CMOS: Memristor and Its Application for Next Generation Storage and Computing

To break the Moore’s Law, beyond CMOS that is post-silicon technologies has gained great attention in recent years, especially, facing the immensely expensive computation in terms of speed and energy in the current big data environment. Novel devices and circuits and computing architectures based on them have appeared to improve the computing efficiency, such as spin device, phase change device, and memristor. Among them, memristor is considered as one of the most promising candidate for the next generation storage and computing own to its high speed, scalability, low energy, and good comparability with CMOS technology. Being found as the fourth missing circuit element by researchers from HP labs in 2008, memristor is a two-terminal device in a MIM (metal-insulator-metal) structure that store information by resistance states. Multiple resistance states can be obtained by controlling the outside bias amplitude and duration. It is organized in cross-point structure with ultra-high density – the memristor sits between the horizontal and vertical metal lines. Research of employing the memristor-based cross-point array in nonvolatile memory, which is called resistive random access memory (ReRAM) has gained great success as an alternative technology of flash memory. Besides that, memristor is also got enormous attention in novel computing architecture, such as in-memory computing and the brain-inspired computing. The brain-inspired computing, also well known as neuromorphic computing, refers to implementing the computation in neural systems by utilizing a specific VLSI hardware system. In traditional CMOS domain, many systems in digital, analog and mixed-signal formats have been presented and demonstrated, aiming at improving computing and data communication efficiency, i.e., high computation speed with low cost. The memristor device appears as a rebirth of this technology because of its above unique features and capability to perform like synapsed naturally. In this paper, we will summarize our work in memory and neuromorphic computing system designs by leveraging the memristor device. It includes three major parts: 1) Memristor devices modeling and related circuits design in resistive memory (ReRAM) technology by investigating their physical mechanism, statistical analysis, and intrinsic challenges. A weighted sensing scheme which assigns different weights to the cells on different bit lines was proposed. The area and power overhead of peripheral circuitry was effectively reduced while minimizing the amplitude of sneak paths. 2) Neuromorphic computing system designs by leveraging memristor devices and algorithm scaling in neural network and machine learning algorithms based on the similarity between memristive effect and biological synaptic behavior. First, a spiking neural network (SNN) with a rate coding model was developed in algorithm level and then mapped to hardware design for supervised learning. In addition, to further speed and accuracy improvement, another neuromorphic system adopting analog input signals with different voltage amplitude and a current sensing scheme was built. Moreover, the use of a single memristor crossbar for each neural network layer was explored. 3) The application-specific optimization for further reliability improvement of the developed neuromorphic systems. In this thesis, the impact of device failure on the memristor-based neuromorphic computing systems for cognitive applications was evaluated. Then, a retraining and a remapping design in algorithm level and hardware level were developed to rescue the large accuracy loss. Our results indicate that the ReRAM design with the weighted sensing scheme save more than 50% area and power consumption. The spiking-based design achieves more than 50% energy savings, while the average probability of failed recognition increase only 1.46% and 5.99% in the feedforward and Hopfield implementations, respectively. Our evaluation based on real device testing data shows that our rescuing design can recover 99.3% of accuracy for MNIST digit recognition when 20% random defects are considered. Our fabricated neuro-chips with CMOS technology in GlobalFoundries 130nm has more than 90% accuracy for MNIST dataset classification.

Evaluating the Calling Performance of a Rare Disease NGS Panel for Single Nucleotide and Copy Number Variants.

Variant detection protocols for clinical next-generation sequencing (NGS) need application-specific optimization. Our aim was to analyze the performance of single nucleotide variant (SNV) and copy number (CNV) detection programs on an NGS panel for a rare disease. Thirty genes were sequenced in 83 patients with hereditary spastic paraplegia. The variant calls obtained with LifeScope, GATK UnifiedGenotyper and GATK HaplotypeCaller were compared with Sanger sequencing. The calling efficiency was evaluated for 187 (56 unique) SNVs and indels. Five multiexon deletions detected by multiple ligation probe assay were assessed from the NGS panel data with ExomeDepth, panelcn.MOPS and CNVPanelizer software. There were 48/51 (94%) SNVs and 1/5 (20%) indels consistently detected by all the calling algorithms. Two SNVs were not detected by any of the callers because of a rare reference allele, and one SNV in a low coverage region was only detected by two algorithms. Regarding CNVs, ExomeDepth detected 5/5 multi-exon deletions, panelcn.MOPs 4/5 and only 3/5 deletions were accurately detected by CNVPanelizer. The calling efficiency of NGS algorithms for SNVs is influenced by variant type and coverage. NGS protocols need to account for the presence of rare variants in the reference sequence as well as for ambiguities in indel calling. CNV detection algorithms can be used to identify large deletions from NGS panel data for diagnostic applications; however, sensitivity depends on coverage, selection of the reference set and deletion size. We recommend the incorporation of several variant callers in the NGS pipeline to maximize variant detection efficiency.

Essential design considerations for the resazurin reduction assay to noninvasively quantify cell expansion within perfused extracellular matrix scaffolds.

Precise measurement of cellularity within bioartificial tissues and extracellular matrix (ECM) scaffolds is necessary to augment rigorous characterization of cellular behavior, as accurate benchmarking of tissue function to cell number allows for comparison of data across experiments and between laboratories. Resazurin, a soluble dye that is reduced to highly fluorescent resorufin in proportion to the metabolic activity of a cell population, is a valuable, noninvasive tool to measure cell number. We investigated experimental conditions in which resazurin reduction is a reliable indicator of cellularity within three-dimensional (3D) ECM scaffolds. Using three renal cell populations, we demonstrate that correlation of viable cell numbers with the rate of resorufin generation may deviate from linearity at higher cell densities, lower resazurin working volumes, or longer incubation times that all contribute to depleting the pool of resazurin. In conclusion, while the resazurin reduction assay provides a powerful, noninvasive readout of metrics enumerating cellularity and growth within ECM scaffolds, assay conditions may strongly influence its applicability for accurate quantification of cell number. The approach and methodological recommendations presented herein may be used as a guide for application-specific optimization of this assay to obtain rigorous and accurate measurement of cellular content in bioengineered tissues.

Editorial

Nowadays, smart systems are becoming more relevant in a large number of critical sectors, like energy management in public spaces, healthcare, automotive, safety and security. Compared to classical embedded systems, a distinctive aspect for these systems is their smartness, which is the ability to learn from the previous experience and to seemingly react to the surrounding environment. However, this tight interaction with the physical environment implies a high level of heterogeneity in the hardware architecture. At the same time, application scenarios are becoming more complex, since an increasing amount of computation is constrained by tight performance, cost and safety requirements. Novel and comprehensive methodologies are thus required to ease the development of next-generation smart systems, with the goal of reducing design costs and time-to-market, improving the smartness of the architectures, and analyze the dynamic behavior of the resulting systems and their reactivity to unpredictable events. This special issue, entitled Innovative Design Methods for Smart Embedded Systems, tackles such challenges by providing both machine-learning techniques and application-specific optimization solutions that guarantee that the application of smart innovations meets the imposed requirements and constraints.