Search NASA⌕ Search

SEARCH · Search NASA

Results for “Workload”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 613 records · Page 34

Evaluating MPI resource usage summary statistics

The Message Passing Interface (MPI) remains the dominant programming model for scientific applications running on today’s high-performance computing (HPC) systems. This dominance stems from MPI’s powerful semantics for inter-process communication that has enabled scientists to write applications for simulating important physical phenomena. MPI does not, however, specify how messages and synchronization should be carried out. Those details are typically dependent on low-level architecture details and the message characteristics of the application. Therefore, analyzing an application’s MPI resource usage is critical to tuning MPI’s performance on a particular platform. The result of this analysis is typically a discussion of the mean message sizes, queue search lengths and message arrival times for a workload or set of workloads. While a discussion of the arithmetic mean in MPI resource usage might be the most intuitive summary statistic, it is not always the most accurate in terms of representing the underlying data. In this paper, we analyze MPI resource usage for a number of key MPI workloads using an existing MPI trace collector and discrete-event simulator. Our analysis demonstrates that the average, while easy and efficient to calculate, is a useful metric for characterizing latency and bandwidth measurements, but may not be a good representation of application message sizes, match list search depths, or MPI inter-operation times. Additionally, we show that the median and mode are superior choices in many cases. We also observe that the arithmetic mean is not the best representation of central tendency for data that are drawn from distributions that are multi-modal or have heavy tails. Furthermore, the results and analysis of our work provide valuable guidance on how we, as a community, should discuss and analyze MPI resource usage data for scientific applications.

97 MATHEMATICS AND COMPUTING↗

Optimization of distributed compute resources utilization in the CMS Global Pool

The CMS Submission Infrastructure is the primary system for managing computing resources for CMS workflows, including data processing, simulation, and analysis. It integrates geographically distributed resources from Grid, HPC, and cloud providers into federated pools managed by HTCondor and Glidein- WMS, for a total of around 500k CPU cores. This system dynamically manages workloads based on priorities defined by the collaboration. Additionally, CMS scheduling strategies must be flexible to handle multiple concurrent workloads while considering changing processing demands and resource availability from various providers.Efficient utilization of vast amounts of distributed compute resources is a key element for the success of the scientific programs of the LHC experiments. Optimizing the system is essential to maximize resource efficiency and fully utilize the distributed computing power. The CMS Submission Infrastructure team thus systematically investigates sources of inefficiency in workload scheduling to reduce their impact. In addition, a strategy of pilot overloading has been introduced to compensate for other inefficiency sources, thereby optimizing resource utilization and enhancing computational throughput.

Mascheroni, Marco [UC, San Diego (main)]↗

Data center growth in the United States: Decoupling the demand for services from electricity use

Data centers are energy intensive buildings that have grown in size and number to meet the increasing demands of a digital economy. This paper presents a bottom-up model to estimate data center electricity demand in the United States over a 20 year period and examines observed and projected electricity use trends in the context of changing data center operations. Results indicate a rapidly increasing electricity demand at the turn of the century that has significantly subsided to a nearly steady annual electricity use of about 70 billion kWh in recent years. While data center workloads continue to grow exponentially, comparable increases in electricity demand have been avoided through the adoption of key energy efficiency measures and a shift towards large cloud-based service providers. Alternative projections from the model illustrate the wide range in potential electricity that could be consumed to support data centers, with the US data center workload demand estimated for 2020 requiring a total electricity use that varies by about 135 billion kWh, depending on the adoption rate of efficiency measures during this decade. While recent improvements in data center energy efficiency have been a success, the growth of data center electricity use beyond 2020 is uncertain, as modeled trends indicate that the efficiency measures of the past may not be enough for the data center workloads of the future. The results show that successful stabilization of data center electricity will require new innovations in data center efficiency to further decouple electricity demand from the ever-growing demand for data center services.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗

Adaptive Runtime Features For Distributed Graph Algorithms

Performance of distributed graph algorithms can benefit greatly by forming rapport between algorithmic abstraction and the underlying runtime system that is responsible for scheduling work and exchanging messages. However, due to their dynamic and irregular nature of computation, distributed graph algorithms written in different programming models impose varying degree of workload pressure on the runtime. To cope with such vastly different workload characteristics, a runtime has to make several trade-offs. One such trade-off arises, for example, when the runtime scheduler has to choose among alternatives such as whether to execute algorithmic work, or progress the network by probing network buffers, or throttle sending messages (termed flow control). This tradeoff decides between optimizing the throughput of a runtime scheduler by increasing the rate of execution of algorithmic work, and reducing the latency of the network messages. Another trade-off exists when a decision has to be made about when to send aggregated messages in buffers (message coalescing). This decision chooses between trading off latency for network bandwidth and vice versa. At any instant, such trade-offs emphasize either on improving the quantity of work being executed (by maximizing the scheduler throughput) or on improving the quality of work (by prioritizing better work). However, encoding static policies for different runtime features (such as flow control, coalescing) can prevent graph algorithms from achieving their full potential, thus can undermine the actual performance of a distributed graph algorithm . In this paper, we investigate runtime support for distributed graph algorithms in the context of two paradigms: variants of wellknown Bulk-Synchronous Parallel model and asynchronous programming model. We explore generic runtime features such as message coalescing (aggregation) and flow control and show that execution policies of these features need to be adjusted over time to make a positive impact on the execution time of a distributed graph algorithm. Since synchronous and asynchronous graph algorithms have different workload characteristics, not all of such runtime features may be good candidates for adaptation. Each of these algorithmic paradigms may require different set of features to be adapted over time. We demonstrate which set of feature(s) can be useful in each case to achieve the right balance of work in the runtime layer. Existing implementation of different graph algorithms can benefit from adapting dynamic policies in the underlying runtime.

Firoz, Jesun S.↗

Coding the Computing Continuum: Fluid Function Execution in Heterogeneous Computing Environments

Advances in network technologies have greatly decreased barriers to accessing physically distributed computers. This newfound accessibility coincides with increasing hardware specialization, creating exciting new opportunities to dispatch workloads to the best resource for a specific purpose, rather than those that are closest or most easily accessible. We present Delta, a service designed to intelligently schedule function-based workloads across a distributed set of heterogeneous computing resources. Delta implements an extensible architecture in which different predictors and scheduling algorithms can be integrated to provide dynamically evolving estimates of function execution times on different resources-estimates that can be used to determine the most appropriate location for execution. We describe predictors for function runtime, data transfer time, and cold-start resource provisioning and configuration delay; dynamic learning methods that update predictor models over time; and scheduling strategies that take into account both function and endpoint information. We show that these methods can halve workload makespan when compared with a strategy that selects the fastest resource, and decrease makespan by a factor of five when compared to a round robin strategy, when deployed on a heterogeneous testbed with resources ranging from a Raspberry Pi to a GPU node in an academic cloud.

Computing continuum↗

Analyzing a Five-Year Failure Record of a Leadership-Class Supercomputer

Extreme-scale computing systems are required to solve some of the grand challenges in science and technology. From astrophysics to molecular biology, supercomputers are an essential tool to accelerate scientific discovery. However, large computing systems are prone to failures due to their complexity. It is crucial to develop an understanding of how these systems fail to design reliable supercomputing platforms for the future. This paper examines a five-year failure and workload record of a leadership-class supercomputer. To the best of our knowledge, five years represents the vast majority of the lifespan of a supercomputer. This is the first time such analysis is performed on a top 10 modern supercomputer. We performed a failure categorization and found out that: i) most errors are GPU-related, with roughly 37% of them being double-bit errors on the cards; ii) failures are not evenly spread across the physical machine, with room temperature presumably playing a major role; and iii) software errors of the system bring down several nodes concurrently. Our failure rate analysis unveils that: i) the system consistently degrades, being at least twice as reliable at the beginning, compared to the end of the period; ii) Weibull distribution closely fits the mean-time-between-failure data; and iii) hardware and software errors show a markedly different pattern. Finally, we correlated failure and workload records to reveal that: i) failure and workload records are weakly correlated, except for certain types of failures when segmented by the hours of the day; ii) several categories of failures make jobs crash within the first minutes of execution; and iii) a significant fraction of failed jobs exhaust the requested time with a disregard of when the failure occurred during execution.

Rojas, Elvis↗

An Evaluation of the Effect of Network Cost Optimization for Leadership Class Supercomputers

Dragonfly-based networks are an extensively deployed network topology in large-scale high-performance computing due to their cost-effectiveness and efficiency. The US will soon have three Exascale supercomputers for leadership class workloads deployed using dragonfly networks. Compared to indirect networks of similar scale, the dragonfly network has considerably reduced cable lengths, cable counts, and switch counts, resulting in significant network cost savings for a given system size, however, these cost reductions result in reduced global minimal paths and more challenging routing. Additionally, large scale dragonfly networks often require a taper at the global link level, resulting in less bisection bandwidth than is achievable in other traditional non-blocking topologies of equivalent scale. While dragonfly networks have been extensively studied, they have yet to be fully evaluated in an extreme scale (i.e., exascale) system that targets capability workloads. In this paper, we present the results of the first large scale evaluation of a dragonfly network on an exascale system (Frontier) and compare its behavior to a similar scale fat-tree network on a previous generation TOP500 system (Summit). This evaluation aims to determine the effect of network cost optimizations by measuring a tapered topology’s impact on capability workloads. Our evaluation is based on a collection of synthetic microbenchmarks, mini-apps, and full scale applications. It compares the scaling efficiencies of each benchmark between the dragonfly-based Frontier and the fat-tree-based Summit systems. Our results show that a dragonfly network is $\sim \mathbf{3 0 \%}$ more cost efficient than a fat-tree topology, which amortizes to $\sim 3 \%$ of an exascale system cost. Furthermore, while tapered dragonfly networks impose significant tradeoffs, the impacts are not as broad as initially thought and are mostly seen in applications with global communication patterns, particularly all-to-all (e.g., FFT-based algorithms), but also local communication patterns (e.g., nearest-neighbor algorithms) that are sensitive to network performance variability.

Khan, Awais↗

FPDeep: Scalable Acceleration of CNN Training on Deeply-Pipelined FPGA Clusters

In this paper, we propose a framework called FPDeep, which uses a hybrid of model and layer paral- lelism to configure distributed reconfigurable clusters to train DNNs. This approach has numerous benefits. First, the design does not suffer from batch size growth. Second, novel workload and weight partitioning leads to balanced loads of both among nodes. And third, the entire system is fine-grained pipeline. This leads to high parallelism and utilization and also minimizes the time features need to be cached while waiting for back-propagation.

Wang, Tianqi↗

ARETE: Accurate Error Assessment via Machine Learning-Guided Dynamic-Timing Analysis

Nanometer circuits are increasingly prone to timing errors, escalating the need for fault injection frameworks to accurately evaluate their impact on applications. Here in this paper, we propose ARETE, a novel cross-layer, fault-injection framework that combines dynamic-binary instrumentation with machine learning-guided dynamic-timing analysis. ARETE enables accurate fault-injection into any application by estimating the location of the injecting errors via dynamic-timing analysis. To accelerate fault-injection, we develop a novel, data-aware, machine learning-based mechanism that dynamically pre-selects the error-prone instructions and limits the application of the costly dynamic-timing analysis only to them. To evaluate ARETE's accuracy, our fully automated toolflow is configured to support fault-injection based on detailed post-layout gate-level simulations as well as via existing workload-agnostic error models. Our results for various workloads, including an autonomous-driving library, show that the location and time of injected errors performed by ARETE, is 89.9% consistent with fault-injection based on full gate-level simulation. On average, ARETE executes 84.6x faster than gate-level simulation and at a cost of 3.4% loss in the program output quality estimation. When compared to the existing statistical fault-injection tools that are based on workload-agnostic error models, ARETE improves the accuracy of fault-injection rate and output quality estimation by 143.9% and 40.4% on average, respectively.

97 MATHEMATICS AND COMPUTING↗

LEISURE: Load-Balanced Network-Wide Traffic Measurement and Monitor Placement

We report that network-wide traffic measurement is of interest to network operators to uncover global network behavior for the management tasks of traffic accounting, debugging or troubleshooting, security, and traffic engineering. Increasingly, sophisticated network measurement tasks such as anomaly detection and security forensic analysis are requiring in-depth fine-grained flow-level measurements. However, performing in-depth per-flow measurements (e.g., detailed payload analysis) is often an expensive process. Given the fast-changing Internet traffic landscape and large traffic volume, a single monitor is not capable of accomplishing the measurement tasks for all applications of interest due to its resource constraint. Moreover, uncovering global network behavior requires network-wide traffic measurements at multiple monitors across the network since traffic measured at any single monitor only provides a partial view and may not be sufficient or accurate. These factors call for coordinated measurements among multiple distributed monitors. In this paper, we present a centralized optimization framework, LEISURE (Load- EqualIzed meaSUREment), for load-balancing network measurement workloads across distributed monitors. Specifically, we consider various load-balancing problems under different objectives and study their extensions to support both fixed and flexible monitor deployment scenarios. We formulate the latter flexible monitor deployment case as an MILP (Mixed Integer Linear Programming) problem and propose several heuristic algorithms to approximate the optimal solution and reduce the computation complexity. We evaluate LEISURE via detailed simulations on Abilene and GEANT network traces to show that LEISURE can achieve much better load-balanced performance (e.g., 4.75 × smaller peak workload and 70 × smaller variance in workloads) across all coordinated monitors in comparison to a naive solution (uniform assignment) to accomplish network-wide traffic measurement tasks under the fixed monitor deployment scenario. Lastly, we also show that under the flexible monitor deployment setting, our heuristic solutions can achieve almost the same load-balancing performance as the optimal solution while reducing the computation times by a factor up to 22.5 × in Abilene and 800 × in GEANT.

42 ENGINEERING↗

Enabling parallel simulation of large-scale HPC network systems

Here, with the increasing complexity of today’s high-performance computing (HPC) architectures, simulation has become an indispensable tool for exploring the design space of HPC systems—in particular, networks. In order to make effective design decisions, simulations of these systems must possess the following properties: (1) have high accuracy and fidelity, (2) produce results in a timely manner, and (3) be able to analyze a broad range of network workloads. Most state-of-the-art HPC network simulation frameworks, however, are constrained in one or more of these areas. In this work, we present a simulation framework for modeling two important classes of networks used in today’s IBM and Cray supercomputers: torus and dragonfly networks. We use the Co-Design of Multi-layer Exascale Storage Architecture (CODES) simulation framework to simulate these network topologies at a flit-level detail using the Rensselaer Optimistic Simulation System (ROSS) for parallel discrete-event simulation. Our simulation framework meets all the requirements of a practical network simulation and can assist network designers in design space exploration. First, it uses validated and detailed flit-level network models to provide an accurate and high-fidelity network simulation. Second, instead of relying on serial time-stepped or traditional conservative discrete-event simulations that limit simulation scalability and efficiency, we use the optimistic event-scheduling capability of ROSS to achieve efficient and scalable HPC network simulations on today’s high-performance cluster systems. Third, our models give network designers a choice in simulating a broad range of network workloads, including HPC application workloads using detailed network traces, an ability that is rarely offered in parallel with high-fidelity network simulations

97 MATHEMATICS AND COMPUTING↗

Design and Performance Characterization of RADICAL-Pilot on Leadership-Class Platforms

Many extreme scale scientific applications have workloads comprised of a large number of individual highperformance tasks. The Pilot abstraction decouples workload specification, resource management, and task execution via job placeholders and late-binding. As such, suitable implementations of the Pilot abstraction can support the collective execution of large number of tasks on supercomputers. We introduce RADICAL-Pilot (RP) as a portable, modular and extensible Pilot enabled runtime system. We describe RP's design, architecture and implementation. We characterize its performance and show its ability to scalably execute workloads comprised of tens of thousands heterogeneous tasks on DOE and NSF leadership-class HPC platforms. Specifically, we investigate RP's weak/strong scaling with CPU/GPU, single/multi core, (non)MPI tasks and python functions when using most of ORNL Summit and TACC Frontera. RADICAL-Pilot can be used stand-alone, as well as the runtime for third-party workflow systems.

97 MATHEMATICS AND COMPUTING↗

INITIATORS AND TRIGGERING CONDITIONS FOR ADAPTIVE AUTOMATION IN ADVANCED SMALL MODULAR REACTORS

It is anticipated that Advanced Small Modular Reactors (AdvSMRs) will employ high degrees of automation. High levels of automation can enhance system performance, but often at the cost of reduced human performance. Automation can lead to human out-of the loop issues, unbalanced workload, complacency, and other problems if it is not designed properly. Researchers have proposed adaptive automation (defined as dynamic or flexible allocation of functions) as a way to get the benefits of higher levels of automation without the human performance costs. Adaptive automation has the potential to balance operator workload and enhance operator situation awareness by allocating functions to the operators in a way that is sensitive to overall workload and capabilities at the time of operation. However, there still a number of questions regarding how to effectively design adaptive automation to achieve that potential. One of those questions is related to how to initiate (or trigger) a shift in automation in order to provide maximal sensitivity to operator needs without introducing undesirable consequences (such as unpredictable mode changes). Several triggering mechanisms for shifts in adaptive automation have been proposed including: operator initiated, critical events, performance-based, physiological measurement, model-based, and hybrid methods. As part of a larger project to develop design guidance for human-automation collaboration in AdvSMRs, researchers at Idaho National Laboratory have investigated the effectiveness and applicability of each of these triggering mechanisms in the context of AdvSMR. Researchers reviewed the empirical literature on adaptive automation and assessed each triggering mechanism based on the human-system performance consequences of employing that mechanism. Researchers also assessed the practicality and feasibility of using the mechanism in the context of an AdvSMR control room. Results indicate that there are tradeoffs associated with each mechanism, but that some are more applicable to the AdvSMR domain. The two mechanisms that consistently improve performance in laboratory studies are operator initiated adaptive automation based on hierarchical task delegation and the Electroencephalogram(EEG) –based measure of engagement. Current EEG methods are intrusive and require intensive analysis; therefore it is not recommended for an AdvSMR control rooms at this time. Researchers also discuss limitations in the existing empirical literature and make recommendations for further research.

99 GENERAL AND MISCELLANEOUS↗

Correcting radiation survey data to account for increased leakage during intensity modulated radiotherapy treatments

Purpose: Intensity modulated radiotherapy (IMRT) treatments require more beam-on time and produce more linac head leakage to deliver similar doses to conventional, unmodulated, radiotherapy treatments. It is necessary to take this increased leakage into account when evaluating the results of radiation surveys around bunkers that are, or will be, used for IMRT. The recommended procedure of applying a monitor-unit based workload correction factor to secondary barrier survey measurements, to account for this increased leakage when evaluating radiation survey measurements around IMRT bunkers, can lead to potentially costly overestimation of the required barrier thickness. This study aims to provide initial guidance on the validity of reducing the value of the correction factor when applied to different radiation barriers (primary barriers, doors, maze walls, and other walls) by evaluating three different bunker designs.Methods: Radiation survey measurements of primary, scattered, and leakage radiation were obtained at each of five survey points around each of three different radiotherapy bunkers and the contribution of leakage to the total measured radiation dose at each point was evaluated. Measurements at each survey point were made with the linac gantry set to 12 equidistant positions from 0° to 330°, to assess the effects of radiation beam direction on the results.Results: For all three bunker designs, less than 0.5% of dose measured at and alongside the primary barriers, less than 25% of the dose measured outside the bunker doors and up to 100% of the dose measured outside other secondary barriers was found to be caused by linac head leakage.Conclusions: Results of this study suggest that IMRT workload corrections are unnecessary, for survey measurements made at and alongside primary barriers. Use of reduced IMRT workload correction factors is recommended when evaluating survey measurements around a bunker door, provided that a subset of the measurements used in this study are repeated for the bunker in question. Reduction of the correction factor for other secondary barrier survey measurements is not recommended unless the contribution from leakage is separately evaluated.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

SU-F-T-96: A New Paradigm of Weekly Chart Checking for Radiation Therapy Clinics

Purpose: To put forth an innovative clinical paradigm for weekly chart checking so that treatment status is periodically checked accurately and efficiently. This study also aims to help optimize the chart checking clinical workflow in a busy radiation therapy clinic. Methods: It is mandated by the Texas Administrative code to check patient charts of radiation therapy once a week or every five fractions, however it varies drastically among institutions in terms of when and how it is done. Some do it every day, but a lot of efforts are wasted on opening ineligible charts; some do it on a fixed day but the distribution of intervals between subsequent checks is not optimal. To establish an optimal chart checking procedure, a new paradigm was developed to achieve 1) charts are checked more accurately and more efficiently; 2) charts are checked on optimal days without any miss; 3) workload is evened out throughout a week when multiple physicists are involved. All active charts will be accessed by querying the R&V system. Priority is assigned to each chart based on the number of days before the next due date followed by sorting and workload distribution steps. New charts are also taken into account when distributing the workload so it is reasonably even throughout the week. Results: Our clinical workflow became more streamlined and smooth. In addition, charts get checked in a more timely fashion so that errors would get caught earlier should they occur. Conclusion: We developed a new weekly chart checking diagram. It helps physicists check charts in a timely manner, saves their time in busy clinics, and consequently reduces possible errors.

60 APPLIED LIFE SCIENCES↗

Exploiting Internal Parallelism for Address Translation in Solid-State Drives

Solid-state Drives (SSDs) have changed the landscape of storage systems and present a promising storage solution for data-intensive applications due to their low latency, high bandwidth, and low power consumption compared to traditional hard disk drives. SSDs achieve these desirable characteristics using internal parallelism—parallel access to multiple internal flash memory chips—and a Flash Translation Layer (FTL) that determines where data are stored on those chips so that they do not wear out prematurely. However, current state-of-the-art cache-based FTLs like the Demand-based Flash Translation Layer (DFTL) do not allow IO schedulers to take full advantage of internal parallelism, because they impose a tight coupling between the logical-to-physical address translation and the data access. In this study to address this limitation, we introduce a new FTL design called Parallel-DFTL that works with the DFTL to decouple address translation operations from data accesses. Parallel-DFTL separates address translation and data access operations into different queues, allowing the SSD to use concurrent flash accesses for both types of operations. We also present a Parallel-LRU cache replacement algorithm to improve the concurrency of address translation operations. To compare Parallel-DFTL against existing FTL approaches, we present a Parallel-DFTL performance model and compare its predictions against those for DFTL and an ideal page-mapping approach. We also implemented the Parallel-DFTL approach in an SSD simulator using real device parameters, and used trace-driven simulation to evaluate Parallel-DFTL’s efficacy. Our evaluation results show that Parallel-DFTL improved the overall performance by up to 32% for the real IO workloads we tested, and by up to two orders of magnitude with synthetic test workloads. Finally, we also found that Parallel-DFTL is able to achieve reasonable performance with a very small cache size and that it provides the best benefit for those workloads with large request size or with high write ratio.

97 MATHEMATICS AND COMPUTING↗

Fast and Scalable Sparse Triangular Solver for Multi-GPU Based HPC Architectures

Designing efficient and scalable sparse linear algebra kernels on modern multi-GPU based HPC systems is a daunting task due to significant irregular memory references and workload imbalance across the GPUs. This is particularly the case for \textit{Sparse Triangular Solver (SpTRSV)} which introduces additional two-dimensional computation dependencies among subsequent computation steps. Dependency information is exchanged and shared among GPUs, thus warrant for efficient memory allocation, data partitioning, and workload distribution as well as fine-grained communication and synchronization support. In this work, we demonstrate that directly adopting unified memory can adversely affect the performance of SpTRSV on multi-GPU architectures, despite linking via fast interconnect like NVLinks and NVSwitches. Alternatively, we employ the latest NVSHMEM technology based on Partitioned Global Address Space programming model to enable efficient fine-grained communication and drastic synchronization overhead reduction. Furthermore, to handle workload imbalance, we propose a malleable task-pool execution model which can further enhance the utilization of GPUs. By applying these techniques, our experiments on the NVIDIA multi-GPU supernode V100-DGX-1 and DGX-2 systems demonstrate that our design can achieve on average 3.53x (up to 9.86x) speedup on a DGX-1 system and 3.66x (up to 9.64x) speedup on a DGX-2 system with 4-GPUs over the Unified-Memory design. The comprehensive sensitivity and scalability studies also show that the proposed zero-copy SpTRSV is able to fully utilize the computing and communication resources of the multi-GPU system.

Xie, Chenhao↗

Lifting and Dropping VMs to Dynamically Transition Between Time- and Space-sharing for Large-Scale HPC Systems

As HPC environments increasingly integrate with edge based systems, system architectures will need to handle a broader class of workloads and scheduling requirements. One result of this shift will be the need to simultaneously support bulk-synchronous parallel (BSP) and on-demand service based applications on the same infrastructure. This in turn will require that future resource management approaches utilize both space-shared as well as time-shared resource scheduling strategies. In this work we introduce the concept of "VM-lifting'' (and its inverse "VM-Dropping'') which allows dynamically switching an HPC workload between space-shared and time-shared scheduling regimes. Our work targets co-kernel based HPC system software environments, in which multiple specialized OS kernels execute natively on dedicated physical resource partitions inside a single compute node. With VM-lifting, a native co-kernel can be migrated at runtime to and from locally hosted Virtual Machine Environments due to changing scheduling requirements of the node. This allows an HPC node to be dynamically (re-)configured as either a time-shared Infrastructure-as-a-Service (IaaS) resource or a dedicated space shared resource based on the current workload demands. We have implemented this approach in the context of the Hobbes Exascale System Software stack and have demonstrated that a node can be reconfigured with minimal impact on the running applications.

Gordon, Nick↗