Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “reconfigurable computing”

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 55 records · Page 3

Multistates and Polyamorphism in Phase-Change K 2 Sb 8 Se 13

The phase-change (PC) materials in the majority of optical data storage media in use today exhibit a fast, reversible crystal -> amorphous phase transition that allows them to be switched between on (1) and off (0) binary states. Solid-state inorganic materials with this property are relatively common, but those exhibiting an amorphous -> amorphous transition called polyamorphism are exceptionally rare. K 2 Sb 8 Se 13 (KSS) reported here is the first example of a material that has both amorphous -> amorphous polyamorphic transition and amorphous -> crystal transition at easily accessible temperatures (227 and 263 degrees C, respectively). The transitions are associated with the atomic coordinative preferences of the atoms, and all three states of K 2 Sb 8 Se 13 are stable in air at 25 degrees C and 1 atm. All three states of K 2 Sb 8 Se 13 exhibit distinct optical bandgaps, E-g = 1.25, 1.0, and 0.74 eV, for the amorphous-II, amorphous-I, and crystalline versions, respectively. The room-temperature electrical conductivity increases by more than 2 orders of magnitude from amorphous-I to -II and by another 2 orders of magnitude from amorphous-II to the crystalline state. This extraordinary behavior suggests that a new class of materials exist which could provide multistate level systems to enable higher-order computing logic circuits, reconfigurable logic devices, and optical switches.

amorphous-amorphous transition↗

A Framework for Neural Network Inference on FPGA-Centric SmartNICs

FPGA-based SmartNICs offer great potential to significantly improve the performance of high-performance computing and warehouse data processing by tightly coupling support for reconfigurable data-intensive computation with cross-node communication, thereby mitigating the von Neumann bottleneck. Existing work, however, has been generally been limited in that it assumes an accelerator model where kernels are offloaded to SmartNICs, but most control tasks are left to the CPUs. This leads to frequent waiting, inferior performance, and scaling challenges. In this work, we propose a new distributive data-centric computing framework, named FCsN, for reconfigurable SmartNIC-based systems. Through a lightweight task circulation execution model and its implementation architecture, FCsN allows the complete detaching of kernel execution, control logic, system scheduling, and network communication to the SmarNICs. This boosts performance by: (i) avoiding the control dependency with CPUs and (ii) supporting streaming kernel execution and network communication at line rate and in a very fine-grained manner. We demonstrate the efficiency and flexibility of FCsN using various types of neural network applications including graph neural networks; as these last are both irregular and data intensive they offer an especially robust demonstration. Evaluations using commonly-used neural network models and graph datasets show that a system with the support of FCsN can achieve, on average, 144 speedups over the MPI-based standard CPU baselines.

Guo, Anqi↗

Cellular computational platform and neurally inspired elements thereof

A cellular computational platform is disclosed that includes a multiplicity of functionally identical, repeating computational hardware units that are interconnected electrically and optically. Each computational hardware unit includes a reprogrammable local memory and has interconnections to other such units that have reconfigurable weights. Each computational hardware unit is configured to transmit signals into the network for broadcast in a protocol-less manner to other such units in the network, and to respond to protocol-less broadcast messages that it receives from the network. Each computational hardware unit is further configured to reprogram the local memory in response to incoming electrical and/or optical signals.

97 MATHEMATICS AND COMPUTING↗

Multi-Tenant Isolation via Reconfigurable Networks

High performance computing environments are often used for a wide variety of workloads ranging from simulation, data transformation and analysis, and complex workflows to name just a few. These systems may process data at various security levels but in so doing are often enclaved at the highest security posture. This approach places significant restrictions on the users of the system even when processing data at a lower security level and exposes data at higher levels of confidentiality to a much broader population than otherwise necessary. The traditional approach of isolation, while effective in establishing security enclaves poses significant challenges for the use of shared infrastructure in HPC environments. This report details current state-of-the-art in reconfigurable network enclaving through Software Defined Networking (SDN) and Network Function Virtualization (NFV) and their applicability to secure enclaves in HPC environments. SDN and NFV methods are based on a solid foundation of system wide virtualization. The purpose of which is very straight forward, the system administrator can deploy networks that are more amenable to customer needs, and at the same time achieve increased scalability making it easier to increase overall capacity as needed without negatively affecting functionality. The network administration of both the server system and the virtual sub-systems is simplified allowing control of the infrastructure through well-defined APIs (Application Programming Interface). While SDN and NFV technologies offer significant promise in meeting these goals, they also provide the ability to address a significant component of the multi-tenant challenge in HPC environments, namely resource isolation. Traditional HPC systems are built upon scalable high-performance networking technologies designed to meet specific application requirements. Dynamic isolation of resources within these environments has remained difficult to achieve. SDN and NFV methodology provide us with relevant concepts and available open standards based APIs that isolate compute and storage resources within an otherwise common networking infrastructure. Additionally, the integration of the networking APIs within larger system frameworks such as OpenStack provide the tools necessary to establish isolated enclaves dynamically allowing the benefits of HPC while providing a controlled security structure surrounding these systems.

97 MATHEMATICS AND COMPUTING↗

Compiling Quantum Circuits for Dynamically Field-Programmable Neutral Atoms Array Processors

Dynamically field-programmable qubit arrays (DPQA) have recently emerged as a promising platform for quantum information processing. In DPQA, atomic qubits are selectively loaded into arrays of optical traps that can be reconfigured during the computation itself. Leveraging qubit transport and parallel, entangling quantum operations, different pairs of qubits, even those initially far away, can be entangled at different stages of the quantum program execution. Such reconfigurability and non-local connectivity present new challenges for compilation, especially in the layout synthesis step which places and routes the qubits and schedules the gates. In this paper, we consider a DPQA architecture that contains multiple arrays and supports 2D array movements, representing cutting-edge experimental platforms. Within this architecture, we discretize the state space and formulate layout synthesis as a satisfiability modulo theories problem, which can be solved by existing solvers optimally in terms of circuit depth. For a set of benchmark circuits generated by random graphs with complex connectivities, our compiler OLSQ-DPQA reduces the number of two-qubit entangling gates on small problem instances by 1.7x compared to optimal compilation results on a fixed planar architecture. To further improve scalability and practicality of the method, we introduce a greedy heuristic inspired by the iterative peeling approach in classical integrated circuit routing. Using a hybrid approach that combined the greedy and optimal methods, we demonstrate that our DPQA-based compiled circuits feature reduced scaling overhead compared to a grid fixed architecture, resulting in 5.1X less two-qubit gates for 90 qubit quantum circuits. These methods enable programmable, complex quantum circuits with neutral atom quantum computers, as well as informing both future compilers and future hardware choices.

Physics↗

Thermal Management for FPGA Nodes in HPC Systems

The integration of FPGAs into large-scale computing systems is gaining attention. In these systems, real-time data handling for networking, tasks for scientific computing, and machine learning can be executed with customized datapaths on reconfigurable fabric within heterogeneous compute nodes. At the same time, thermal management, particularly battling the cooling cost and guaranteeing the reliability, is a continuing concern. The introduction of new heterogeneous components into HPC nodes only adds further complexities to thermal modeling and management. The thermal behavior of multi-FPGA systems deployed within large compute clusters is less explored. Here, we first show that the thermal behaviors of different FPGAs of the same generation can vary due to their physical locations in a rack and process variation, even though they are running the same tasks. We present a machine learning–based model to capture the thermal behavior of each individual FPGA in the cluster. We then propose two thermal management strategies guided by our thermal model. First, we mitigate thermal variation and hotspots across the cluster by proactive thermal-aware task placement. Under the tested system and benchmarks, we achieve up to 26.4° C and on average 13.3° C system temperature reduction with no performance penalty. Second, we utilize this thermal model to guide HLS parameter tuning at the task design stage to achieve improved thermal response after deployment.

97 MATHEMATICS AND COMPUTING↗

Dynamic, Adaptive, Systems and Materials: Complex, Simple and Emergent Behaviors

This program has been funded by DoE/BES for twenty years. It has moved into and out of various subjects as it has developed, but it has retained its focus on complexity and complex systems. The project has evolved in the following way: Self-Assembly and Biomimetic Self-Assembly: All self-assembling systems depend upon a minimum of two types of interaction: a repulsion and an attraction. For the familiar molecular systems, attractive interactions are typically hydrogen bonds and electrostatic interactions. Repulsive interactions include steric effects, hydrophobic effects (in biological systems), and charge-charge repulsion. We have expanded this repertoire to include surface interactions, magnetic interactions, and others. I list these systems in the order in which we have explored them: i) A key emphasis in current work is in understanding how the movement of ions in a magnetic field (the Lorentz effect) interacts with catalytic systems. We have demonstrated that an acceleration in rate of reduction of CO 2 to CO can be accomplished by applying an external magnetic field. This acceleration is largely due to the application of the Lorentz effect on mass transport at the catalyst’s surface. ii) We have also extensively explored the influence of electrostatics, as exhibited in self-assembling systems, by tribocharging. iii) Another key system involves surface tension effects; examples include interactions between heavy particles floating at a liquid-air interface, and interacting by changes in surface area; interactions of bubbles and bubble rafts, behaviors of bubble trains in microfluidic networks, and behaviors of microorganisms in constraining environments. iv) This work has intentionally de-emphasized biological systems; but it does include some work on protein-ligand interactions and interactions among microorganisms. v) We have also explored applications of some of these effects, these explorations include bubble rafts as diffraction gratings, exploration of the structures that can be obtained by tribocharging and uses of these structures in exploring nucleation and melting of crystals. vi) Although not a major focus of this work, several other topics have emerged and offer opportunities for future work. These include the behavior of bubble trains and bubble rafts in microfluidic systems. A particularly interesting example is the formation of bubble trains that repeat in the alteration of large and small bubbles according to rules we do not presently understand, but are uniquely large-period oscillating systems. These systems offer a new route into understanding the instabilities of the type represented by oscillations. vii) We have also begun exploratory projects on magnetic levitation (especially to determine molecular density), and information storage (in molecules). Magnetic Levitation: Self-assembling and biomimetic systems require both attraction and repulsion. We have used electrostatics (tribocharging), interfacial free-energies (surface tension and related forces) and others. Potential uses include reconfigurable diffraction gratings and liquid lenses; exploration of mechanisms in tribocharging; tunneling in EGaIn junctions; and bubble trains (especially in micro-fluidic systems). Examples of systems representing these topics is included in the following papers: Complexity: Disks rotating at a water-air interface; Benard-Marangoni effects; Vortex-Crystals from spinning magnetic disks (Marangoni effects); EGaIn Electrode to study quantum tunneling; Self-Assembly of electrostatically-charged metallic spheres (electrets); Dynamically reconfigurable lens; Using computational designs of ligands for enzymes; Electrostatic self-assembly by tribocharging; Monodisperse bubble trains in microchannel systems; Inverted dripping faucet; Flames; Printing of micro-organisms to regenerate the “ink” of printing device; Using micro-organisms to move loads (“microoxen”); Motion of bacterial swarms near surfaces; Making monodisperse particles in microfluidic systems; Coding/decoding of information stored in droplet trains in microfluidic networks; Magnetic levitation; and Information storage. i) Tribocharging. The change in focus of this work on electrets from the fundamentals of charging to applications of these materials in studying self-assembly using electrostatic interactions. ii) Bubbles in Microchannels. The realization that systems of bubbles in microchannels represented a major opportunity to study complexity in a very tractable system, and the development of a semi-quantitative theory of this subject. iii) Flames. The growth of “flames” remains an exploratory subject for the research, although their currently relatively little active work involving it ongoing. iv) Systems with Microorganisms. The removal of work in biological systems from this project. Based on work supported in this program, we now have a significant project on the development of microfluidic tools for studying C. elegans (a nematode), but this work was not appropriate for a program focused on complexity, and we developed separate support for it. (It is, however, an example of successful seeding of a new area by BES.) The work on electrets has gone through a period in which a part of the program was the subject of a MURI; the focus of this work was to develop materials that did not charge electrostatically on friction or contact. The MURI is now over, and the work on dynamic self-assembly (supported by BES) is the major focus. “Flames” has also enjoyed synergistic support, in terms of a project supported by DARPA on flame suppression (in the absence of extinguishing agents, using acoustic and electrostatic interactions). This work was helpful in understanding some of the basics of flames, but is entirely distinct from the BES focus in complexity. A growing interest is in the Lorentz effect. The Lorentz effect is the force exerted on charged particles (electrons, ions, charged molecules) when they move through a perpendicular magnetic field. The Lorentz effect is almost ubiquitous in modern technology: examples of applications include electric motors, dynamos, cathode ray tubes, many batteries, and most systems that control electrical currents with magnetic forces. We have begun to explore the Lorentz effect in electrochemical systems and heterogeneous catalytic systems involving charged organic species and inorganic ions. This work is still at an early stage, but initial studies that Lorentz effects can be large when ions move through magnetic fields, or magnetic fields move in the presence of ions.

36 MATERIALS SCIENCE↗

Understanding the Design Space of Sparse/Dense Multiphase Dataflows for Mapping Graph Neural Networks on Spatial Accelerators

Graph Neural Networks (GNNs) have garnered a lot of recent interest because of their success in learning representations from graph-structured data across several critical applications in cloud and HPC. Owing to their unique compute and memory characteristics that come from an interplay between dense and sparse phases of computations, the emergence of reconfigurable dataflow (aka spatial) accelerators offers promise for acceleration by mapping optimized dataflows (i.e., computation order and parallelism) for both phases. The goal of this work is to characterize and understand the design-space of dataflow choices for running GNNs on spatial accelerators in order for the compilers to optimize the dataflow based on the workload. Specifically, we propose a taxonomy to describe all possible choices for mapping the dense and sparse phases of GNNs spatially and temporally over a spatial accelerator, capturing both the intra-phase dataflow and the inter-phase (pipelined) dataflow. Using this taxonomy, we do deep-dives into the cost and benefits of several dataflows and perform case studies on implications of hardware parameters for dataflows and value of flexibility to support pipelined execution.

97 MATHEMATICS AND COMPUTING↗

Machine Learning on Heterogeneous, Edge, and Quantum Hardware for Particle Physics (ML-HEQUPP)

The next generation of particle physics experiments will face a new era of challenges in data acquisition, due to unprecedented data rates and volumes along with extreme environments and operational constraints. Harnessing this data for scientific discovery demands real-time inference and decision-making, intelligent data reduction, and efficient processing architectures beyond current capabilities. Crucial to the success of this experimental paradigm are several emerging technologies, such as artificial intelligence and machine learning (AI/ML) and silicon microelectronics, and the advent of quantum algorithms and processing. Their intersection includes areas of research such as low-power and low-latency devices for edge computing, heterogeneous accelerator systems, reconfigurable hardware, novel codesign and synthesis strategies, readout for cryogenic or high-radiation environments, and analog computing. This white paper presents a community-driven vision to identify and prioritize research and development opportunities in hardware-based ML systems and corresponding physics applications, contributing towards a successful transition to the new data frontier of fundamental science.

Gonski, Julia [SLAC]↗

Scalable FPGA Accelerator for Deep Convolutional Neural Networks with Stochastic Streaming

Here, FPGA-based heterogeneous computing platform, due to its extreme logic reconfigurability, emerges to be a strong contender as computing fabric in modern AI. As a result, various FPGA-based accelerators for deep CNN—the key driver of modern AI—have been proposed due to their advantages of high performance, reconfigurability, and fast development round, etc. In general, the consensus among researchers is that, although FPGA-based accelerator can achieve much higher energy efficiency, its raw computing performance lags behind when compared with GPUs with similar logic density. In this paper, we develop an alternative methodology to efficiently implement CNNs with FPGAs that outperform GPUs in terms of both power consumption and performance. Our key idea is to design a scalable hardware architecture and circuit design for large-scale CNNs that leverages a stochastic-based computing principle. Specifically, there are three major performance advantages. First, all key components of our deep learning CNN are designed and implemented to compute stochastically, thus achieving excellent computing performance and energy efficiency. Second, because our proposed CNN architecture enables a stream-mode computing, all of its stages can process even the partial results from preceding stages, therefore not incurring unnecessary latency due to data dependency. Finally, our FPGA-based deep CNN also provides a superior hardware scalability when compared with conventional FPGA implementations by reducing the bandwidth requirement between layers. The results show that our proposed CNN architecture significantly outperforms all previous FPGA-based deep CNN implementation approaches. It achieves 1.58x more GOPS, 6.42x more GOPS/Slice, and 10.92x more GOPS/W when compared with state-of-the-art CNN architecture. The top-5 accuracy of stochastic VGG-16 CNN is 86.77 percent with 18.91 fps frame rate.

97 MATHEMATICS AND COMPUTING↗

Load Forecasting Based Distribution System Network Reconfiguration -- A Distributed Data-Driven Approach

In this paper, a short-term load forecasting approach based network reconfiguration is proposed in a parallel manner. Specifically, a support vector regression (SVR) based short-term load forecasting approach is designed to provide an accurate load prediction and benefit the network reconfiguration. Because of the nonconvexity of the three-phase balanced optimal power flow, a second-order cone program (SOCP) based approach is used to relax the optimal power flow problem. Then, the alternating direction method of multipliers (ADMM) is used to compute the optimal power flow in distributed manner. Considering the limited number of the switches and the increasing computation capability, the proposed network reconfiguration is solved in a parallel way. The numerical results demonstrate the feasible and effectiveness of the proposed approach.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Evaluating Proxy Apps with FPGAs

Field-programmable gate arrays (FPGAs) are becoming more promising for high performance computing (HPC). Unlike fixed hardware specialization, reconfigurability allows FPGAs to be customized for a wide variety of HPC workloads. Our ultimate goal is to understand the viability of FPGAs as accelerators for some of the key HPC codes at LLNL and identify application characteristics that can benefit from the reconfigurability of FPGAs. More specifically, this LDRD project conducted a detailed performance study of FPGAs using sparse matrix computations and identified key optimization techniques to effectively use FPGA hardware resources.

97 MATHEMATICS AND COMPUTING↗

Machine-learning-aided cognitive reconfiguration for flexible-bandwidth HPC and data center networks [Invited]

This paper proposes a machine-learning (ML)-aided cognitive approach for effective bandwidth reconfiguration in optically interconnected datacenter/high-performance computing (HPC) systems. The proposed approach relies on a Hyper-X-like architecture augmented with flexible-bandwidth photonic interconnections at large scales using a hierarchical intra/inter-POD photonic switching layout. We first formulate the problem of the connectivity graph and routing scheme optimization as a mixed-integer linear programming model. A two-phase heuristic algorithm and a joint optimization approach are devised to solve the problem with low time complexity. Then, we propose an ML-based end-to-end performance estimator design to assist the network control plane with intelligent decision making for bandwidth reconfiguration. Numerical simulations using traffic distribution profiles extracted from HPC applications traces as well as random traffic matrices verify the accuracy performance of the ML design estimator ( < <#comment/> 9 % <#comment/> error) and demonstrate up to 5 × <#comment/> throughput gain from the proposed approach compared with the baseline Hyper-X network using fixed all-to-all intra/inter-portable data center interconnects.

Chen, Xiaoliang (ORCID:0000000278056237)↗

Automatic Cloud Bursting under FermiCloud

Cloud computing is changing the infrastructure upon which scientific computing depends from supercomputers and distributed computing clusters to a more elastic cloud-based structure. The service-oriented focus and elasticity of clouds can not only facilitate technology needs of emerging business but also shorten response time and reduce operational costs of traditional scientific applications. Fermi National Accelerator Laboratory (Fermilab) is currently in the process of building its own private cloud, FermiCloud, which allows the existing grid infrastructure to use dynamically provisioned resources on FermiCloud to accommodate increased but dynamic computation demand from scientists in the domains of High Energy Physics (HEP) and other research areas. Cloud infrastructure also allows to increase a private cloud’s resource capacity through “bursting” by borrowing or renting resources from other community or commercial clouds when needed. This paper introduces a joint project on building a cloud federation to support HEP applications between Fermi National Accelerator Laboratory and Korea Institution of Science and Technology Information, with technical contributions from the Illinois Institute of Technology. In particular, this paper presents two recent accomplishments of the joint project: (a) cloud bursting automation and (b) load balancer. Automatic cloud bursting allows computer resources to be dynamically reconfigured to meet users’ demands. The load balance algorithm which the cloud bursting depends on decides when and where new resources need to be allocated. Our preliminary prototyping and experiments have shown promising success, yet, they also have opened new challenges to be studied

Wu, Hao↗

R and DE Robotic Sensor Intern Team: Sensor Suite Mock Up Testing

Assess sensor suite performance: - Mounted on crawler typical to what would be used in tunnel; - In test bed simulating the tunnel environment. The SuperDroid fiber tethered LT-F crawler is a general-purpose surveillance robot that our team has refurbished as a testing platform for the sensor suite. This crawler was selected because its dimensions are similar to the actual crawler that will be deployed in the field. Crawler preparation included: - Communicating with vendor to resolve software and platform issues; - Configuring a laptop with the crawler control software; - Reconfiguration of communication protocol between computer and crawler; - Resolution of control board checksum errors; - Installation of a mounting fixture on the base; - In progress: addition of lighting needed for total darkness. A solution for hosting the selected sensors onto a single unit that could function standalone or be mounted on the crawler was needed. Sensor mount requirements included: - Host Lidar, panospheric cameras and IMU sensors; - Cable management for rotating Lidar; - House on-board crawler computer; - Method for affixing to crawler; - Robust and vibration proof. The Lidar sensor required a sturdy mount that ensured the axis of rotation be centered with the LiDAR's desired laser path and included a slender profile in order to not block the scanning FOV. The mounting plate holds two 180-degree cameras on either side and the servo on the center. The servo mount is located so the axis of rotation is in the center of the plate. The mount features a set of guides that hold the wires and ensure they wrap around the spool area properly. The sensor box design features thick walls for added strength, holes for mounting the sensor suite to the crawler, notches on the bottom for the different components and holes for cable routing and management. The sensor mounts were designed using CREO and printed in the R and DE Additive Manufacturing Lab. To test the performance of the sensor suite for our application it was desired to design and build a test bed simulating the H-Canyon tunnel environment. Requirements included: - Dimensions similar to the tunnel; - Wall surfaces simulating the eroded tunnel walls; - Complete darkness; - Even and uneven crawler path; - Ability to change the tunnel width to simulate erosion. The HCAEX tunnel changes in height and includes several bends ranging from approximately 30 to 80 degrees, for that reason a test bed capable of easily being changed was needed. The final design includes two sections that can be connected to provide a 16' tunnel or placed at an angle to create a tunnel bend. The test bed features caster wheels for easy movement and has a sliding inner wall that enables tunnel width to be controlled. In order to simulate the tunnel walls without adding a lot of weight to the test bed, sheets of faux stone are to be used on the inner walls. To create a completely dark tunnel, 100% black-out film will be draped over the tunnel structure. Carpet remnants in the lab will enable folds to be introduced in the floor to generate uneven surface data.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

Exploring the Performance Boundaries of a Small Reconfigurable Multi-Mission UAV through Multidisciplinary Analysis

The performance of a small reconfigurable unmanned aerial vehicle (UAV) is evaluated, combining a multidisciplinary approach in the computational analysis of additive manufactured structures, fluid dynamics, and experiments. Reconfigurable UAVs promise cost savings and efficiency, without sacrificing performance, while demonstrating versatility to fulfill different mission profiles. The use of computational fluid dynamics (CFD) in UAV design produces higher accuracy aerodynamic data, which is particularly important for complex aircraft concepts such as blended wing bodies. To address challenges relating to anisotropic materials, the Tsai–Wu failure criterion is applied to the structural analysis, using CFD solutions as load inputs. Aerodynamic performance results show the low-speed variant attains an endurance of 1 h, 48 min, whereas its high-speed counterpart is 29 min at a 66.7% higher cruise speed. Each variant serves different aspects of small UAS deployment, with low speed envisioned for high-endurance surveying, and high speed for long-range or time-critical missions such as delivery. The experimental and simulation results suggest room for design iteration, in wing area and geometry adjustments. Structural simulations demonstrated the need for airframe improvements to the low-speed configuration. This paper highlights the potential of reconfigurable UAVs to be useful across multiple industries, advocating for further research and design improvements.

42 ENGINEERING↗