Search

Introduction: Missions beyond low Earth orbit (LEO) will expose astronauts to ionizing radiation (IR) in the form of solar energetic particles (SEP) and galactic cosmic rays (GCR) including high atomic number and energy (HZE) nuclei. The gastrointestinal (GI) system is documented to be highly radiosensitive with even relatively low dose IR exposures capable of inducing mucosal lesions and disrupting epithelial barrier function. IR is also an established risk factor for colorectal cancer (CRC) with several studies examining long-term GI effects of SEP/GCR exposure using tumor-prone APC mouse models. Studies of acute short-term effects of modeled space radiation exposures in wildtype mouse models are more limited and necessary to better define charged particle- induced GI pathologies and test novel medical countermeasures (MCMs) to promote astronaut safety.Methods: In this study, we performed ground-based studies where male and female C57BL/6J mice were exposed to γ-rays, 50 MeV protons, or 1 GeV/n Fe-56 ions at the NASA Space Radiation Laboratory (NSRL) with histology and immunohistochemistry endpoints measured in the first 24 h post-irradiation to define immediate SEP/GCR-induced GI alterations.Results: Our data show that unlike matched γ-ray controls, acute exposures to protons and iron ions disrupts intestinal function and induces mucosal lesions, vascular congestion, epithelial barrier breakdown, and marked enlargement of mucosa-associated lymphoid tissue. We also measured kinetics of DNA double-strand break (DSB) repair using gamma-H2AX- specific antibodies and apoptosis via TUNEL labeling, noting the induction and disappearance of extranuclear cytoplasmic DNA marked by gamma-H2AX only in the charged particle-irradiated samples. We show that 18 h pre-treatment with curcumin-loaded nanolipoprotein particles (cNLPs) delivered via IV injection reduces DSB-associated foci levels and apoptosis and restore crypt villi lengths.Discussion: These data improve our understanding of physiological alterations in the GI tract immediately following exposures to modeled space radiations and demonstrates effectiveness of a promising space radiation MCM.

Significance\ud \ud Communicating in ways that motivate engagement in social distancing remains a critical global public health priority during the COVID-19 pandemic. This study tested motivational qualities of messages about social distancing (those that promoted choice and agency vs. those that were forceful and shaming) in 25,718 people in 89 countries. The autonomy-supportive message decreased feelings of defying social distancing recommendations relative to the controlling message, and the controlling message increased controlled motivation, a less effective form of motivation, relative to no message. Message type did not impact intentions to socially distance, but people’s existing motivations were related to intentions. Findings were generalizable across a geographically diverse sample and may inform public health communication strategies in this and future global health emergencies.\ud \ud \ud \ud Abstract\ud \ud Finding communication strategies that effectively motivate social distancing continues to be a global public health priority during the COVID-19 pandemic. This cross-country, preregistered experiment (n = 25,718 from 89 countries) tested hypotheses concerning generalizable positive and negative outcomes of social distancing messages that promoted personal agency and reflective choices (i.e., an autonomy-supportive message) or were restrictive and shaming (i.e., a controlling message) compared with no message at all. Results partially supported experimental hypotheses in that the controlling message increased controlled motivation (a poorly internalized form of motivation relying on shame, guilt, and fear of social consequences) relative to no message. On the other hand, the autonomy-supportive message lowered feelings of defiance compared with the controlling message, but the controlling message did not differ from receiving no message at all. Unexpectedly, messages did not influence autonomous motivation (a highly internalized form of motivation relying on one’s core values) or behavioral intentions. Results supported hypothesized associations between people’s existing autonomous and controlled motivations and self-reported behavioral intentions to engage in social distancing. Controlled motivation was associated with more defiance and less long-term behavioral intention to engage in social distancing, whereas autonomous motivation was associated with less defiance and more short- and long-term intentions to social distance. Overall, this work highlights the potential harm of using shaming and pressuring language in public health communication, with implications for the current and future global health challenges.

In order to satisfy their ever increasing capacity and compute requirements, many machine learning models are distributed across multiple nodes using space-efficient parallelism strategies. As a result, collective communications are often on the critical path, and hiding their latency by overlapping kernel-granular communication and computation is difficult due to the absence of independent computation. In this work, we propose fusing computation with communication using GPU-initiated networking, and leverage GPUs' massive parallelism to enable fine-grained overlap of the fused operations. We have developed a single, self-contained GPU kernel where workgroups (WGs) immediately communicate their results to remote GPUs when they complete their computation. Meanwhile, other WGs within the same kernel perform overlapping computation, maintaining high ALU utilization. Furthermore, we propose zero-copy optimizations for peer-to-peer GPU communication where the data computed by one GPU is directly written to the destination buffers within the peer GPUs, eliminating intermediate stores and extra buffering. Our approach leverages the emerging multi-node GPU system trend where GPUs are physically close to network with direct GPU-NIC interconnects. We demonstrate our approach by creating an embedding + All-to-All fused kernel which overlaps embedding operations and the dependent all-to-all collective in DLRM models. We evaluate our approach both using simulation and real hardware. Our evaluations show that our approach can effectively overlap All-to-All communication with embedding computations, subsequently reducing their combined execution time by 31% on average (up to 58%) for inter-node and by 25% (up to 35%) for intra-node configurations. Scale-out simulations indicate that our approach reduces DLRM execution time by ~10% for 128 node system.

Cache-block compression is a highly effective technique for both reducing accesses to lower levels in the memory hierarchy (cache compression) and minimizing data transfers (link compression). While many effective cache-block compression algorithms have been proposed, the design of these algorithms is largely ad hoc and manual and relies on human recognition of patterns. In this article, we take an entirely different approach. We introduce a class of “byte-select” compression algorithms, as well as an automated methodology for generating compression algorithms in this class. We argue that, based on upper bounds within the class, the study of this class of byte-select algorithms has potential to yield algorithms with better performance than existing cache-block compression algorithms. The upper bound we establish on the compression ratio is 2X that of any existing algorithm. We then offer a generalized representation of a subset of byte-select compression algorithms and search through the resulting space guided by a set of training data traces. Using this automated process, we find efficient and effective algorithms for various hardware applications. We find that the resulting algorithms exploit novel patterns that can inform future algorithm designs. The generated byte-select algorithms are evaluated against a separate set of traces and evaluations show that Byte-Select has a 23% higher compression ratio on average. While no previous algorithm performs best for all our data sets which include CPU and GPU applications, our generated algorithms do. Using an automated hardware generator for these algorithms, we show that their decompression and compression latency is one and two cycles respectively, much lower than any existing algorithm with a competitive compression ratio.

The goal of this project was to scan and reverse engineer one of three bronze cannons (41MG86 - 11900-1) recovered from the La Belle excavation, which is currently on exhibit at the Bullock Texas State History Museum in Austin, Texas. A freeform computer aided design (CAD) model was generated based upon the topology of the three-dimensional (3-D) mesh in Geomagic Design X with a desired maximum deviation of 0.1 mm between the model and the mesh. Deviations were calculated in Geomagic Control X using the surface model as the reference data, and the mesh as the measured data. A custom patch network was subsequently designed using a series of iterative revisions until the whole of the surface model met with the specified tolerance. The 3-D surface model of the cannon will be replicated in a variety of media at variable scales for use in exhibits and for educational and promotional material for the Texas Historical Commission, the Republic of France, and their partner museums.

Florida Administrative Code (FAC) 5E-2.0311, “Performance Standards and Acceptable Test Conditions for Preventive Termite Treatments for New Construction,” approved in 2003, currently remains without change as originally written. This Code requires broad revision to include updating existing and new termite management methods and technologies, describing acceptable testing protocols that objectively evaluate termite management product efficacy, and clearly and accurately define product performance standards required for approval and use in Florida, USA. This manuscript provides modification and expansion of the existing Code, first in outline form, then followed by a proposed complete revision including detailed explanations for each outline heading and sub-heading. Efficacy requirements, acceptable test conditions, and performance standards requirements are described for termite management products, materials, and non-chemical physical exclusion barriers. We also include testing protocols for field plot and structure efficacy tests. Minimum performance standards also are described for field plots and structure tests to include: (1) termiticide applications to soil, (2) termite baiting systems, (3) pesticides applied directly to wood and building components, (4) pesticide combination mixtures, and (5) new application technologies, products, and methodologies. Although this manuscript addresses revision of a specific Florida Administrative Code, it also is appropriate for consideration by the United States Environmental Protection Agency (USEPA) as well as regulatory entities in other states responsible for pesticide regulation and consumer protection services when writing or revising their code requirements for termite management.

Contemporary GPUs are widely used for throughput-oriented data-parallel workloads and increasingly are being considered for latency-sensitive applications in datacenters. Examples include recurrent neural network (RNN) inference, network packet processing, and intelligent personal assistants. These data parallel applications have both high throughput demands and real-time deadlines (40μs-7ms). Moreover, the kernels in these applications have relatively few threads that do not fully utilize the device unless a large batch size is used. However, batching forces jobs to wait, which increases their latency, especially when realistic job arrival times are considered.Previously, programmers have managed the tradeoffs associated with concurrent, latency-sensitive jobs by using a combination of GPU streams and advanced scheduling algorithms running on the CPU host. Although GPU streams allow the accelerator to execute multiple jobs concurrently, prior state-of-the-art solutions use the relatively distant CPU host to prioritize the latency-sensitive GPU tasks. Thus, these approaches are forced to operate at a coarse granularity and cannot quickly adapt to rapidly changing program behavior.We observe that fine-grain, device-integrated kernel schedulers efficiently meet the deadlines of concurrent, latency-sensitive GPU jobs. To overcome the limitations of software-only, CPU-side approaches, we extend the GPU queue scheduler to manage real-time deadlines. We propose a novel laxity-aware scheduler (LAX) that uses information collected within the GPU to dynamically vary job priority based on how much laxity jobs have before their deadline. Compared to contemporary GPUs, 3 state-of-the-art CPU-side schedulers and 6 other advanced GPU-side schedulers, LAX meets the deadlines of 1.7X – 5.0X more jobs and provides better energy-efficiency, throughput, and 99-percentile tail latency.

Psocids are damaging stored-product pests. In this study, eggs and early-instar nymphs, adults, and all life stages of Liposcelis entomophila, L. decolor, L. bostrychophila, and L. paeta were subjected to 43, 50, or 75% (Control) relative humidity (RH) for 2, 4, 6, 8, 10, 12, 14, or 16 d at 30.0°C. All adults of these species died within 8 d at both 43 and 50% RH, except for L. bostrychophila, which required 12 d at 50% RH for 100% mortality to occur. For all life stages and eggs and early-instar nymphs, maximum survival times (times to 100% mortality) at 43 or 50% RH for L. entomophila, L. decolor, L. bostrychophila, and L. paeta, were 8 and 10 d, 8 and 12 d, 12 and 14 d, and 12 and 16 d, respectively. During this study, numbers of nymphs and adults of all species 14 d after the RH treatments increased within the 75% RH Control arenas. Different species and life stages responded differently to 43 and 50% RH, as time to kill all stages of the four psocid species was 8–12 and 10–16 d, respectively. Results indicate that using a specific RH environment may be effective in psocid management.

Recent advances in die-stacking and 2.5D chip integration technologies introduce in-package network heterogeneities that can complicate the interconnect design. Integrating chiplets over a silicon interposer offers new opportunities of optimizing interposer topologies. However, limited by the capability of existing network-on-chip (NoC) simulators, the full potential of the interposer-based NoCs has not been exploited. In this paper, we address the shortfalls of prior NoC designs and present a new family of chiplet topologies called Kite. Kite topologies better utilize the diverse networking and frequency domains existing in new interposer systems and outperform the prior chiplet topology proposals. Kite decreased synthetic traffic latency by 7% and improved the maximum throughput by 17% on average versus Double Butterfly and Butter Donut, two previous proposals developed using less accurate modeling.

GPUs have evolved from providing highly-constrained programmability for a single kernel to using pre-emption to ensure independent forward progress for concurrently executing kernels. However, modern GPUs do not ensure independent forward progress for kernels that use fine-grain synchronization to coordinate inter-work-group execution. Enabling independent forward progress among work-groups (WGs) is challenging as pre-empted kernels may be rescheduled with fewer hardware resources. This can lead to oversubscribed execution scenarios that deadlock current hardware even for correctly written code. Prior work addresses this problem by requiring programmers to specify resource requirements and assuming static resource allocation, which adds scheduling constraints and reduces portability. We propose a family of novel hardware approaches --- trading off hardware complexity for performance --- that provide independent forward progress in the presence of fine-grain inter-WG synchronization and dynamic resource allocation. Additionally, we propose new waiting atomic instructions compatible with proposed C++20 extensions. Our final design, Autonomous WorkGroups (AWG), uses hints from regular and waiting atomics to cooperatively schedule WGs within a kernel, improving efficiency and virtualizing hardware resources. In non-oversubscribed scenarios, AWG outperforms a busy-waiting baseline (which deadlocks in oversubscribed scenarios) by 12x on average for benchmarks that use different mutexes and barriers for fine-grained, WG granularity synchronization. Furthermore, AWG outperforms other solutions that do not deadlock in the oversubscribed case, such as fixed-interval round-robin context switching or naively extending monitor/mwait to GPUs, by 2.6x and 2.2x, respectively.

Changing social and economic practices had an important role for human adaptive strategies in colonial contexts and sometimes had profound consequences for emerging societies. This study uses insights from stable-isotope analyses, as well as other historical and archaeological evidence, to investigate the social and economic roles of French animal husbandry as an adaptive strategy for the settlers taking part in La Salle’s famous expedition (1684–1688) to colonize the mouth of the Mississippi River. Stable carbon- and nitrogen-isotope analyses of pig bones and other faunal remains from the shipwreck, La Belle, and associated Fort Saint Louis on the northern coast of the Gulf of Mexico are used to evaluate specific historical accounts of colonists’ animal-husbandry practices and show that a large swine population was sustained primarily on meat from local hunting activities. In this context it is argued that, although the substantial efforts involved in raising pigs mainly on other animal products seemingly makes little economic sense, there are social explanations for what initially appears to be irrational behavior. This study provides an example of how stable-isotope analyses of animal-husbandry practices can contribute to understanding social processes through historical archaeology.

A GIS-enabled culvert design module is presented. This module employs Python programming to combine a proposed culvert location, topography, land use, and rainfall data to automatically design a culvert. The module is embedded within ESRI ArcGIS 10.4 software, providing a seamless single platform that eliminates error propagation associated with cross-platform data transfer as well as providing 95% time savings over traditional calculation methods. The module uses United States Geological Survey digital elevation data to analyze watershed topography. Runoff coefficients are determined from data available through the National Land Cover Database. Rainfall data are retrieved from the National Oceanic and Atmospheric Administration and combined with watershed and land use information to calculate peak discharge using the rational method. Peak discharge is then combined with culvert design parameters to design a single-barrel culvert. The module was used to redesign ten existing culverts along a highway in Tuscaloosa, Alabama, resulting in designs for updated land cover and rainfall conditions. Results from the techniques developed herein can be used for planning purposes and to highlight vulnerabilities in the existing infrastructure. The automation methods may be extended to other hydrologic objectives and runoff mitigation design such as open-channel design and detention or retention ponds.

General-Purpose Graphics Processing Units (GPGPUs) are employed in today's fastest supercomputers to accelerate a variety of scientific compute workloads. These workloads typically comprise of data-parallel mathematical kernels that are well suited for execution on GPUs. The hyperplane sweep operation is one such mathematical kernel that is commonly used in high-performance computing. In this paper, we characterize the conventional bulk synchronous hyperplane sweep implementation currently used by GPUs and identify significant performance improvement potential by breaking the operation into finer-grain tasks. Guided by this characterization, we propose multi-grain task decomposition and scheduling techniques to optimize the operation. We use KRIPKE as a case study that features the sweep operation, and we show that our proposed optimizations achieve 41% speedup over the bulk synchronous implementation.

As the deep learning and high-performance computing markets continue to grow, hardware designers are increasingly optimizing future GPUs to run compute (a.k.a. GPGPU) workloads. A key area of optimization for these compute-oriented designs, which was not emphasized when GPUs exclusively executed graphics workloads, is inter-thread data sharing and synchronization. GPU cache coherence protocols now support these operations and are governed by a specified memory consistency model. In general, current GPU models are based on sequential consistency for data-race-free (SC for DRF), which mandates data written to memory must be globally visible only after certain synchronization points. GPU coherence protocols based on such relaxed memory models are particularly difficult to design and test due to the large number of memory accesses that may be reordered. This leaves GPU hardware designers struggling to validate the correctness of GPU cache coherence optimizations. To address this issue, this paper introduces a novel, completely autonomous random testing methodology for complex GPU cache coherence protocols. Our framework continuously generates sequences of memory requests with minimal user intervention using a mix of load, store, and atomic operations. The tester dynamically and autonomously checks each response against an expected global view of memory and immediately detects any inconsistencies in a target coherence protocol, providing designers detailed feedback on the issue. We then demonstrate the methodology on the popular cycle-level gem5 simulator by replacing its GPU core model with our unique testing framework. The results show that the GPU tester can cover 94% and 100% of all reachable state transitions in L1 and L2 caches respectively of a representative GPU coherence protocol. This coverage is 6.25% and 25% higher than the one achieved by a wide selection of 26 applications. In addition, the tester runs more than 50 times faster than those applications, which enables efficient and fast protocol debugging.

In recent years, machine intelligence (MI) applications have emerged as a major driver for the computing industry. Optimizing these workloads is important, but complicated. As memory demands grow and data movement overheads increasingly limit performance, determining the best GPU caching policy to use for a diverse range of MI workloads represents one important challenge. To study this, we evaluate 17 MI applications and characterize their behavior using a range of GPU caching strategies. In our evaluations, we find that the choice of caching policy in GPU caches involves multiple performance trade-offs and interactions, and there is no one-size-fits-all GPU caching policy for MI workloads. Based on detailed simulation results, we motivate and evaluate a set of cache optimizations that consistently match the performance of the best static GPU caching policies.

Sparse solvers are heavily used in computational fluid dynamics (CFD), computer-aided design (CAD), and other important application domains. These solvers remain challenging to execute on massively parallel architectures, due to the sequential dependencies between the fine-grained application tasks. In particular, parallel sparse solvers typically suffer from substantial scheduling and dependency-management overheads relative to the compute operations. We propose adaptive task aggregation (ATA) to efficiently execute such irregular computations on GPU architectures via hierarchical dependency management and low-latency task scheduling. On a gamut of representative problems with different data-dependency structures, ATA significantly outperforms existing GPU task-execution approaches, achieving a geometric mean speedup of 2.2X to 3.7X across different sparse kernels (with speedups of up to two orders of magnitude).

In 1995, the historical shipwreck of La Belle was discovered off the coast of Texas. One partial human skeleton was recovered from alongside cargo in the rear portion of the ship; a second (complete) skeleton was found atop coiled anchor rope in the bow. In late 2015, comprehensive forensic genetic testing began on multiple samplings from each set of remains. For the partial skeleton recovered from the ship's rear cargo area, results were obtained for 26/27 Y-STRs using traditional CE; with MPS technology, results were obtained for 18/24 Y-STRs, 56/56 ancestry-informative SNPs (aiSNPs), 22/22 phenotype-informative SNPs (piSNPs), 22/27 autosomal STRs, 4/7 X-STRs, and 94/94 identity-informative SNPs (iiSNPs). For the complete skeleton of the second individual, results were obtained for 7/17 Y-STRs using traditional CE; with MPS technology, results were obtained for 5/24 Y-STRs, 49/56 aiSNPs, 18/22 piSNPs, 15/27 autosomal STRs, 1/7 X-STRs, and 66/94 iiSNPs. Biogeographic ancestry for each set of skeletal remains was predicted using the ancestry feature and metapopulation tool of the Y-STR Haplotype Reference Database (YHRD), Haplogroup Predictor, and the Forensic Research/Reference on Genetics knowledge base (FROG-kb). Phenotype prediction was performed using piSNP data and the HIrisplex eye color and hair color DNA phenotyping webtool. mtDNA whole genome sequencing also was performed successfully. This study highlights the sensitivity of current forensic laboratory methods in recovering DNA from historical and archaeological human remains. Using advanced sequencing technology provided by MiSeq™ FGx (Verogen) and Ion S5™ (Thermo Fisher Scientific) instrumentation, degraded skeletal remains can be characterized using a panel of diverse and highly informative markers, producing data which can be useful in both forensic and genealogical investigations.

In 1932, seven burials were discovered on a Texas plantation that was originally the site of a 17th-century Caddo Indian village. Of the seven excavated graves, one set of remains (an adult male) was notably buried in a manner inconsistent with traditional Caddoan burial practices and has long been purported to be the remains of Sieur de Marle (a member of the French explorer La Salle's last expedition). Diary accounts of La Salle's expedition scribe report that Sieur de Marle died along a river near an Indian village during a trek to Canada to find help for colonists left behind at the ill-fated Fort St. Louis. Additionally, two lead projectiles recovered from the grave were ballistically analyzed and determined to be consistent with ammunition used in 17th-century weaponry. In the 1980s, anthropologists requested access to the remains for study, but the skull was missing. Cranial measurements recorded in 1940 and 1962 (by two independent anthropologists) were used to investigate the ancestry of this individual; and the Giles-Elliot (G-E) discriminant function was calculated to be 18.1, within the Anglo-European range. Dietary isotope testing on non-cranial skeletal elements determined that this unknown male's diet was rich in animal/marine protein sources, which differs appreciably from Caddo Indian populations of that time period. In order to genetically assess this individual's biogeographic ancestry and to provide further support that this individual is of European descent, mitochondrial DNA (mtDNA) sequencing was performed using the Applied Biosystems™ Precision ID mtDNA Whole Genome Panel. mtDNA sequencing of multiple sections from two different long bones yielded compiled results consistent with either Haplogroup H or R, both predominantly European mtDNA haplogroups. Further anthropological calculations were conducted using cranial measurements, FORDISC™ software, and discriminant function analysis. Two-way, four-way, and multigroup discriminant function analyses further classify this set of unidentified remains as being White (European) in origin, with posterior probabilities of 0.999, 0.881 and 0.986, respectively. Combined with historical records of Sieur de Marle's death, as well as overlays of historical and contemporary maps which demonstrate that the plantation site aligns with Joutel's diary accounts of de Marle's burial, these collective results support that these remains are of a European male and may possibly belong to this prominent member of La Salle's expedition team.

Graph applications are common in scientific and enterprise computing. Recent research used graphics processing units (GPUs) to accelerate graph workloads. These applications tend to present characteristics that are challenging for SIMD execution. To achieve high performance, prior work studied individual graph problems, and designed device-specific algorithms and optimizations to achieve high performance. However, programmers have to expend significant manual effort, packing data and computation to make such solutions GPU-friendly. This usually is too complex for regular programmers, and the resultant implementations may not be portable and perform well across platforms. To address these concerns, we propose and implement a library of software building blocks with application examples, BelRed which allows programmers to build graph applications with ease. BelRed currently is built on top of the OpenCL™ framework and optimized for GPUs. It consists of fundamental linear-algebra building blocks necessary for graph processing. Developers can program graph algorithms with a set of key primitives. This paper introduces the API and presents several case studies on how to use the library for a variety of representative graph problems. We evaluate application performance on an AMD GPU and investigate optimization techniques to improve performance. We show that this framework is useful to provide satisfactory GPU acceleration of various graph applications and help reduce programming efforts significantly.

In this study, we evaluate the effect of strong sediment/bedrock impedance contrasts on soil amplification in Boston, Massachusetts, for typical sites along the Charles and Mystic Rivers. These sites can be characterized by artificial fill overlying marine sediments overlying glacial till and bedrock, where the depth to bedrock ranges from 20 to 80 m. The marine sediments generally consist of organic silts, sand, and Boston Blue Clay. We chose these sites because they represent typical foundation conditions in the City of Boston, and the soil conditions are similar to other high impedance contrast environments. The sediment/bedrock interface in this region results in an impedance ratio on the order of ten, which in turn results in a significant amplification of the ground motion. Using stratigraphic information derived from numerous boreholes across the region paired with geologic and geomorphologic constraints, we develop a depth-to-bedrock model for the greater Boston region. Using shear-wave velocity profiles from 30 locations, we develop average velocity profiles for sites mapped as artificial fill, glaciofluvial deposits, and bedrock. By pairing the depth-to-bedrock model with the surficial geology and the average shear-wave velocity profiles, we can predict soil amplification in Boston. We compare linear and equivalent-linear site response predictions for a soil layer of varying thickness over bedrock, and assess the effects of varying the bedrock shear-wave velocity (VSb) and quality factor (Q). In a moderate seismicity region like Boston, many earthquakes will result in ground motions that can be modeled with linear site response methods. We also assess the effect of bedrock depth on soil amplification for a generic soil profile in artificial fill, using both linear and equivalent-linear site response models. Finally, we assess the accuracy of the model results by comparing the predicted (linear site response) and observed site response at the Northeastern University (NEU) vertical seismometer array during the 2011 M 5.8 Mineral, Virginia, earthquake. Site response at the NEU vertical array results in amplification on the order of 10 times at a period between 0.7–0.8 s. The results from this study provide evidence that the mean short-period and mean intermediate-period amplification used in design codes (i.e., from the Fa and Fv site coefficients) may underpredict soil amplification in strong impedance contrast environments such as Boston.

We can unlock higher performance in general purpose graphics processing unit (GPGPU) chips by adhering to local environment conditions while setting a timing (clock) reference. We propose partitioning the GPU chip into smaller independent domains to enable finer control of the chip voltage frequency (V/F) settings. Post silicon (Si) analysis is used to classify the domains as fast, ’typical’, or slow. This information is then provided to the system management unit (SMU) to enable the appropriate settings for power optimization. In this paper, we present an initial analysis of power/performance associated with fine-grain domains in GPUs.

As GPUs become larger and provide an increasing number of parallel execution units, a single kernel is no longer sufficient to utilize all available resources. As a result, GPU applications are beginning to use fine-grain asynchronous kernels, which are executed in parallel and expose more concurrency. Currently, the Heterogeneous System Architecture (HSA) and Compute Unified Device Architecture (CUDA) specifications support concurrent kernel launches with the help of multiple command queues (a.k.a. HSA queues and CUDA streams, respectively). In conjunction, GPU hardware has decreased launch overheads making fine-grain kernels more attractive.Although increasing the number of command queues is good for kernel concurrency, the GPU hardware can only monitor a fixed number of queues at any given time. Therefore, if the number of command queues exceeds hardware's monitoring capability, the queues become oversubscribed and hardware has to service some of these queues sequentially. This mapping process periodically swaps between all allocated queues and limits the available concurrency to the ready kernels in the currently mapped queues. In this paper, we bring to attention the queue oversubscription challenge and demonstrate one solution, queue prioritization, which provides up to 45x speedup for NW benchmark against the baseline that swaps queues in a round-robin fashion.

Modern GPU frameworks use a two-phase compilation approach. Kernels written in a high-level language are initially compiled to an implementation agnostic intermediate language (IL), then finalized to the machine ISA only when the target GPU hardware is known. Most GPU microarchitecture simulators available to academics execute IL instructions because there is substantially less functional state associated with the instructions, and in some situations, the machine ISA’s intellectual property may not be publicly disclosed. In this paper, we demonstrate the pitfalls of evaluating GPUs using this higher-level abstraction, and make the case that several important microarchitecture interactions are only visible when executing lower-level instructions. Our analysis shows that given identical application source code and GPU microarchitecture models, execution behavior will differ significantly depending on the instruction set abstraction. For example, our analysis shows the dynamic instruction count of the machine ISA is nearly 2× that of the IL on average, but contention for vector registers is reduced by 3× due to the optimized resource utilization. In addition, our analysis highlights the deficiencies of using IL to model instruction fetching, control divergence, and value similarity. Finally, we show that simulating IL instructions adds 33% error as compared to the machine ISA when comparing absolute runtimes to real hardware.

Modern accelerator programming frameworks, such as OpenCL, organise threads into work-groups. Remote-scope promotion (RSP) is a language extension recently proposed by AMD researchers that is designed to enable applications, for the first time, both to optimise for the common case of intra-work-group communication (using memory scopes to provide consistency only within a work-group) and to allow occasional inter-work-group communication (as required, for instance, to support the popular load-balancing idiom of work stealing). We present the first formal, axiomatic memory model of OpenCL extended with RSP. We have extended the Herd memory model simulator with support for OpenCL kernels that exploit RSP, and used it to discover bugs in several litmus tests and a work-stealing queue, that have been used previously in the study of RSP. We have also formalised the proposed GPU implementation of RSP. The formalisation process allowed us to identify bugs in the description of RSP that could result in well-synchronised programs experiencing memory inconsistencies. We present and prove sound a new implementation of RSP that incorporates bug fixes and requires less non-standard hardware than the original implementation. This work, a collaboration between academia and industry, clearly demonstrates how, when designing hardware support for a new concurrent language feature, the early application of formal tools and techniques can help to prevent errors, such as those we have found, from making it into silicon.

This article provides an overview of AMD's vision for exascale computing, and in particular, how heterogeneity will play a central role in realizing this vision. Exascale computing requires high levels of performance capabilities while staying within stringent power budgets. Using hardware optimized for specific functions is much more energy efficient than implementing those functions with general-purpose cores. However, there is a strong desire for supercomputer customers not to have to pay for custom components designed only for high-end high-performance computing systems. Therefore, high-volume GPU technology becomes a natural choice for energy-efficient data-parallel computing. To fully realize the GPU's capabilities, the authors envision exascale computing nodes that compose integrated CPUs and GPUs (that is, accelerated processing units), along with the hardware and software support to enable scientists to effectively run their scientific experiments on an exascale system. The authors discuss the hardware and software challenges in building a heterogeneous exascale system and describe ongoing research efforts at AMD to realize their exascale vision.

Distributed systems incorporate GPUs because they provide massive parallelism in an energy-efficient manner. Unfortunately, existing programming models make it difficult to route a GPU-initiated network message. The traditional coprocessor model forces programmers to manually route messages through the host CPU. Other models allow GPU-initiated communication, but are inefficient for small messages.To enable fine-grain PGAS-style communication between threads executing on different GPUs, we introduce Gravel. GPU-initiated messages are offloaded through a GPU-efficient concurrent queue to an aggregator (implemented with CPU threads), which combines messages targeting to the same destination. Gravel leverages diverged work-group-level semantics to amortize synchronization across the GPU's data-parallel lanes.Using Gravel, we can distribute six applications, each with frequent small messages, across a cluster of eight GPU-accelerated nodes. Compared to one node, these applications run 5.3x faster, on average. Furthermore, we show Gravel is more programmable and usually performs better than prior GPU networking models.

Commodity heterogeneous systems (e.g., integrated CPUs and GPUs), now support a unified, shared memory address space for all components. Because the latency of global communication in a heterogeneous system can be prohibi-tively high, heterogeneous systems (unlike homogeneous CPU systems) provide synchronization mechanisms that only guarantee ordering among a subset of threads, which we call a scope. Unfortunately, the consequences and se-mantics of these scoped operations are not yet well under-stood. Without a formal and approachable model to reason about the behavior of these operations, we risk an array of portability and performance issues. In this paper, we embrace scoped synchronization with a new class of memory consistency models that add scoped synchronization to data-race-free models like those of C++ and Java. Called sequential consistency for heterogeneous-race-free (SC for HRF), the new models guarantee SC for programs with "sufficient" synchronization (no data races) of "sufficient" scope. We discuss two such models. The first, HRF-direct, works well for programs with highly regular parallelism. The second, HRF-indirect, builds on HRF-direct by allowing synchronization using different scopes in some cases involving transitive communication. We quanti-tatively show that HRF-indirect encourages forward-looking programs with irregular parallelism by showing up to a 10% performance increase in a task runtime for GPUs.

Nanoceria is an exceptionally versatile, commercially valuable catalytic material whose properties vary dramatically from that of the bulk material. Nanoceria's redox properties can be tuned by choice of method of preparation, particle size, nature and level of dopant, particle shape and surface chemistry. The two oxidation states of the cerium element in the lattice make possible the formation of oxygen vacancies which are essential to the high reactivity of the material, its oxygen buffering capability and thus its ability to act as a catalyst for both oxidation and reduction reactions. Ceria has important commercial utility in the areas of chemical mechanical polishing and planarization, catalytic converters and diesel oxidation catalysts, intermediate temperature solid oxide fuel cells and sensors. Its potential future uses include chemical looping combustion, photolytic and thermolytic water splitting for hydrogen production and as a therapeutic agent for the treatment of certain human diseases. We have seen that the method of synthesis, particle size, stabilizing corona, and purity dictate where it is used commercially. Finally, in regards to the prescient words of Dr. Feynman, we note that while there is indeed “plenty of room at the bottom”, there quite possibly exists an optimal nanoceria size of between 2–3 nm that provides maximal reactivity and thermodynamic stability.

This article presents a historical overview of the late 20th-century advent of financialization, that is, the unprecedented growth of the financial sector. We summarize its origins and consequences, particularly greater income inequality. An econometric model quantifies the relationship. We conclude that along with higher unemployment and an eroding minimum wage, the growth of the U.S. financial sector has contributed to the exacerbation of inequality in recent decades.

Coptotermes Wasmann is one of the most important genera of wood-destroying insect pests, both in its native and introduced countries. Pyrethroids are among the most widely used insecticides in wood preservation around the world. Consequently, they have often been evaluated against different species of Coptotermes. However, because various test methods have been used between countries, comparing results is problematic. These field trials, using a single aboveground method of exposure, assessed a range of retentions of two pyrethroids (bifenthrin and permethrin) in Pinus radiata D. Don sapwood against two species of Coptotermes in three countries to provide directly comparable results. Coptotermes acinaciformis (Froggatt) in Australia consumed the most nontreated wood, followed by Coptotermes formosanus Shiraki in China, then C formosanus in the United States, although these data were not significantly different. Both termite species demonstrated a dose-response to wood treated with the two pyrethroids; less wood was consumed as retention increased. Overall, C. acinaciformis consumed relatively little of the treated wood. In comparison, C. formosanus consumed 20–90% of the wood treated at the lowest retentions of the pyrethroids evaluated. Results indicated that C. acinaciformis was more sensitive to pyrethroid toxicity/ repellency compared with C. formosanus. Factors that may have influenced the results are discussed. However, using a single aboveground method of exposure across three countries, that suited both species of Coptotermes, made it possible to determine unambiguously the actual differences between the species in their tolerances to the two pyrethroid insecticides.

Current trends suggest that future computing platforms will be increasingly heterogeneous. While these heterogeneous processors physically integrate disparate computing elements like CPUs and GPUs on a single chip, their programmability critically depends upon the ability to efficiently support cache coherence and shared virtual memory across tightly-integrated CPUs and GPUs. However, throughput-oriented GPUs easily overwhelm existing hardware coherence mechanisms that long kept the cache hierarchies in multi-core CPUs coherent. This paper proposes a novel solution called Software Assisted Hardware Coherence (SAHC) to scale cache coherence to future heterogeneous processors. We observe that the system software (Operating system and runtime) often has semantic knowledge about sharing patterns of data across the CPU and the GPU. This high-level knowledge can be utilized to effectively provide cache coherence across throughput-oriented GPUs and latency-sensitive CPUs in a heterogeneous processor. SAHC thus proposes a hybrid software-hardware mechanism that judiciously uses hardware coherence only when needed while using software's knowledge to filter out most of the unnecessary coherence traffic. Our evaluation suggests that SAHC can often eliminate up to 98-100% of the hardware coherence lookups, resulting up to 49% reduction in runtime.

Achieving optimal performance on heterogeneous computing systems requires a programming model that supports the execution of asynchronous, multi-stream, and out-of-order tasks in a shared memory environment. Asynchronous dependency-driven tasking is one such programming model that allows the computation to be expressed as a directed acyclic graph (DAG) and exposes fine-grain task management to the programmer. The use of DAGs to extract parallelism also enables runtimes to perform dynamic load-balancing, thereby achieving higher throughput when compared to the traditional bulk-synchronous execution. However, efficient DAG implementations require features such as user-level task dispatch, hardware signalling and local barriers to achieve low-overhead task dispatch and dependency resolution.In this paper, we demonstrate that the Heterogeneous System Architecture (HSA) exposes the above capabilities, and we validate their benefits by implementing three well-referenced applications using fine-grain tasks: Cholesky factorization, Lower Upper Decomposition (LUD), and Needleman-Wunsch (NW). HSA's user-level task dispatch and signalling capability allow work to be launched and dependencies to be managed directly by the hardware, avoiding inefficient bulk-synchronization. Our results show the HSA task-based implementations of Cholesky, LUD, and NW are representative of this emerging class of workloads and using hardware-managed tasks achieve a speedup of 3.8x, 1.6x, and 1.5x, respectively, compared to bulk-synchronous implementations.

The gem5 simulation infrastructure is the merger of the best aspects of the M5 [4] and GEMS [9] simulators. M5 provides a highly configurable simulation framework, multiple ISAs, and diverse CPU models. GEMS complements these features with a detailed and exible memory system, including support for multiple cache coherence protocols and interconnect models. Currently, gem5 supports most commercial ISAs (ARM, ALPHA, MIPS, Power, SPARC, and x86), including booting Linux on three of them (ARM, ALPHA, and x86). The project is the result of the combined efforts of many academic and industrial institutions, including AMD, ARM, HP, MIPS, Princeton, MIT, and the Universities of Michigan, Texas, and Wisconsin. Over the past ten years, M5 and GEMS have been used in hundreds of publications and have been downloaded tens of thousands of times. The high level of collaboration on the gem5 project, combined with the previous success of the component parts and a liberal BSD-like license, make gem5 a valuable full-system simulation tool.

We present results of a rock-magnetic study of soils that were affected by wildfires that burned portions of the Everglades in the Spring of 2008. Soils at sites that were extensively burned exhibit a pronounced surface magnetic enhancement effect with magnetizations of surface samples up to 16 times greater than that observed at depth (>7 cm) at these sites. The increase in magnetization results from an increased abundance of a low-coercivity phase (maghemite) that occurs at the expense of the abundance of a high-coercivity phase (goethite). These results indicate that fire-induced heating caused goethite in the surface soils to convert into a more magnetic, low-coercivity phase, such as maghemite. Goethite is an excellent adsorber of phosphorus, and therefore we hypothesize that the destruction of goethite as a result of burning may have important implications for phosphorus cycling in the Everglades ecosystem. Copyright © 2010 John Wiley & Sons, Ltd.

Studies of embryonic stem cells (ESCs) reveal that these cell lines can be derived from differing stages of embryonic development. We analyzed common changes in the expression of microRNAs (miRNAs) and mRNAs in 9 different human ESC (hESC) lines during early commitment and further examined the expression of key ESC-enriched miRNAs in earlier developmental states in several species. We show that several previously defined hESC-enriched miRNA groups (the miR-302, −17, and −515 families, and the miR-371–373 cluster) and several other hESC-enriched miRNAs are down-regulated rapidly in response to differentiation. We further found that mRNAs up-regulated upon differentiation are enriched in potential target sites for these hESC-enriched miRNAs. Interestingly, we also observed that the expression of ESC-enriched miRNAs bearing identical seed sequences changed dynamically while the cells transitioned through early embryonic states. In human and monkey ESCs, as well as human-induced pluripotent stem cells (iPSCs), the miR-371–373 cluster was consistently up-regulated, while the miR-302 family was mildly down-regulated when the cells were chemically treated to regress to an earlier developmental state. Similarly, miR-302b, but not mmu-miR-295, was expressed at higher levels in murine epiblast stem cells (mEpiSC) as compared with an earlier developmental state, mouse ESCs. These results raise the possibility that the relative expression of related miRNAs might serve as diagnostic indicators in defining the developmental state of embryonic cells and other stem cell lines, such as iPSCs. These data also raise the possibility that miRNAs bearing identical seed sequences could have specific functions during separable stages of early embryonic development.

We demonstrate that the pyrrhotite that commonly occurs as inclusions in diamonds is capable of carrying strong and stable remanent magnetizations. This suggests that with suitable samples, it may be possible to use such inclusions to obtain information about the geomagnetic field, such as the paleointensity, during key intervals of geologic time. The pyrrhotite inclusions examined here exhibit a strong and pronounced anisotropy of magnetic remanence that is coupled to the crystal structure of the diamond. While this anisotropy of remanence may complicate efforts to obtain paleointensity data, we propose that it may provide a valuable way to uniquely identify individual diamonds that contain even minute inclusions of pyrrhotite.

Heterogeneous system architecture (HSA) and OpenCL define scoped synchronization to facilitate low overhead communication across a subset of threads. Scoped synchronization works well for static sharing patterns, where consumer threads are known a priori. It works poorly for dynamic sharing patterns (e.g., work stealing) where programmers cannot use a faster small scope due to the rare possibility that the work is stolen by a thread in a distant slower scope. This puts programmers in a conundrum: optimize the common case by synchronizing at a faster small scope or use work stealing at a slower large scope. In this paper, we propose to extend scoped synchronization with remote-scope promotion. This allows the most frequent sharers to synchronize through a small scope. Infrequent sharers synchronize by promoting that remote small scope to a larger shared scope. Synchronization using remote-scope promotion provides performance robustness for dynamic workloads, where the benefits provided by scoped synchronization and work stealing are hard to anticipate. Compared to a naïve baseline, static scoped synchronization alone achieves a 1.07x speedup on average and dynamic work stealing alone achieves a 1.18x speedup on average. In contrast, synchronization using remote-scope promotion achieves a robust 1.25x speedup on average, across a diverse set of graph benchmarks and inputs.

Modern graphics processing units (GPUs) include hardware- controlled caches to reduce bandwidth requirements and energy consumption. However, current GPU cache hierarchies are inefficient for general purpose GPU (GPGPU) comput- ing. GPGPU workloads tend to include data structures that would not fit in any reasonably sized caches, leading to very low cache hit rates. This problem is exacerbated by the design of current GPUs, which share small caches be- tween many threads. Caching these streaming data struc- tures needlessly burns power while evicting data that may otherwise fit into the cache. We propose a GPU cache management technique to im- prove the efficiency of small GPU caches while further re- ducing their power consumption. It adaptively bypasses the GPU cache for blocks that are unlikely to be referenced again before being evicted. This technique saves energy by avoid- ing needless insertions and evictions while avoiding cache pollution, resulting in better performance. We show that, with a 16KB L1 data cache, dynamic bypassing achieves sim- ilar performance to a double-sized L1 cache while reducing energy consumption by 25% and power by 18%. The technique is especially interesting for programs that do not use programmer-managed scratchpad memories. We give a case study to demonstrate the inefficiency of current GPU caches compared to programmer-managed scratchpad memories and show the extent to which cache bypassing can make up for the potential performance loss where the effort to program scratchpad memories is impractical.

Nanoceria is an exceptionally versatile, commercially valuable catalytic material whose properties vary dramatically from that of the bulk material. Nanoceria's redox properties can be tuned by choice of method of preparation, particle size, nature and level of dopant, particle shape and surface chemistry. The two oxidation states of the cerium element in the lattice make possible the formation of oxygen vacancies which are essential to the high reactivity of the material, its oxygen buffering capability and thus its ability to act as a catalyst for both oxidation and reduction reactions. Ceria has important commercial utility in the areas of chemical mechanical polishing and planarization, catalytic converters and diesel oxidation catalysts, intermediate temperature solid oxide fuel cells and sensors. Its potential future uses include chemical looping combustion, photolytic and thermolytic water splitting for hydrogen production and as a therapeutic agent for the treatment of certain human diseases. We have seen that the method of synthesis, particle size, stabilizing corona, and purity dictate where it is used commercially. Finally, in regards to the prescient words of Dr. Feynman, we note that while there is indeed “plenty of room at the bottom”, there quite possibly exists an optimal nanoceria size of between 2–3 nm that provides maximal reactivity and thermodynamic stability.