Search NASA⌕ Search

SEARCH · Search NASA

Results for “spatial accelerator”

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 19 records

Evaluating Spatial Accelerator Architectures with Tiled Matrix-Matrix Multiplication

There is a growing interest in custom spatial accelerators for machine learning applications. These accelerators employ a spatial array of processing elements (PEs) interacting via custom buffer hierarchies and networks-on-chip. The efficiency of these accelerators comes from employing optimized dataflow (i.e., spatial/temporal partitioning of data across the PEs and fine-grained scheduling) strategies to optimize data reuse. The focus of this work is to evaluate these accelerator architectures using a tiled general matrix-matrix multiplication (GEMM) kernel. To do so, we develop a framework that finds optimized mappings (dataflow and tile sizes) for a tiled GEMM for a given spatial accelerator and workload combination, leveraging an analytical cost model for runtime and energy. Our evaluations over five spatial accelerators demonstrate that the tiled GEMM mappings systematically generated by our framework achieve high performance on various GEMM workloads and accelerators.

43 PARTICLE ACCELERATORS↗

Evaluating Spatial Accelerator Architectures with Tiled Matrix-Matrix Multiplication.

There is a growing interest in custom spatial accelerators for machine learning applications. These accelerators employ a spatial array of processing elements (PEs) interacting via custom buffer hierarchies and networks-on-chip. The efficiency of these accelerators comes from employing optimized dataflow (i.e., spatial/temporal partitioning of data across the PEs and fine-grained scheduling) strategies to optimize data reuse. The focus of this work is to evaluate these accelerator architectures using a tiled general matrix-matrix multiplication (GEMM) kernel. To do so, we develop a framework that finds optimized mappings (dataflow and tile sizes) for a tiled GEMM for a given spatial accelerator and workload combination, leveraging an analytical cost model for runtime and energy. Finally, our evaluations over five spatial accelerators demonstrate that the tiled GEMM mappings systematically generated by our framework achieve high performance on various GEMM workloads and accelerators.

42 ENGINEERING↗

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↗

Union: A Unified HW-SW Co-Design Ecosystem in MLIR for Evaluating Tensor Operationson Spatial Accelerators

To meet the extreme compute demands for deep learning across commercial and scientific applications, dataflow accelerators are becoming increasingly popular. While these“domain-specific” accelerators are not fully programmable like CPUs and GPUs, they retain varying levels of flexibility with respect to data orchestration, i.e., dataflow and tiling optimizations to enhance efficiency. There are several challenges when designing new algorithms and mapping approaches to execute the algorithms for a target problem on new hardware. Previous works have addressed these challenges individually. To address this challenge as a whole, in this work, we present an HW-SW co-design ecosystem for spatial accelerators called Union within the popular MLIR compiler infrastructure. Our framework allows exploring different algorithms and their mappings on several accelerator cost models. Union also includes a plug-and-play library of accelerator cost models and mappers which can easily be extended. The algorithms and accelerator cost models are connected via a novel mapping abstraction that captures the map space of spatial accelerators which can be systematically pruned based on constraints from the hardware, workload, and mapper. We demonstrate the value of Union for the community with several case studies which examine offloading different tensor operations (CONV/GEMM/Tensor Contraction) on diverse accelerator architectures using different mapping schemes.

Jeong, Geonhwa↗

SU-E-I-37: Low-Dose Real-Time Region-Of-Interest X-Ray Fluoroscopic Imaging with a GPU-Accelerated Spatially Different Bilateral Filtering

Purpose: The purpose of our study is to reduce imaging radiation dose while maintaining image quality of region of interest (ROI) in X-ray fluoroscopy. A low-dose real-time ROI fluoroscopic imaging technique which includes graphics-processing-unit- (GPU-) accelerated image processing for brightness compensation and noise filtering was developed in this study. Methods: In our ROI fluoroscopic imaging, a copper filter is placed in front of the X-ray tube. The filter contains a round aperture to reduce radiation dose to outside of the aperture. To equalize the brightness difference between inner and outer ROI regions, brightness compensation was performed by use of a simple weighting method that applies selectively to the inner ROI, the outer ROI, and the boundary zone. A bilateral filtering was applied to the images to reduce relatively high noise in the outer ROI images. To speed up the calculation of our technique for real-time application, the GPU-acceleration was applied to the image processing algorithm. We performed a dosimetric measurement using an ion-chamber dosimeter to evaluate the amount of radiation dose reduction. The reduction of calculation time compared to a CPU-only computation was also measured, and the assessment of image quality in terms of image noise and spatial resolution was conducted. Results: More than 80% of dose was reduced by use of the ROI filter. The reduction rate depended on the thickness of the filter and the size of ROI aperture. The image noise outside the ROI was remarkably reduced by the bilateral filtering technique. The computation time for processing each frame image was reduced from 3.43 seconds with single CPU to 9.85 milliseconds with GPU-acceleration. Conclusion: The proposed technique for X-ray fluoroscopy can substantially reduce imaging radiation dose to the patient while maintaining image quality particularly in the ROI region in real-time.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Spatially Accelerated Winding Numbers for Curved Geometry

The generalized winding number (GWN) is a scalar field that supports robust containment queries on curved geometry, including non-watertight, overlapping, and nested boundary representations. While queries can be easily parallelized over samples, direct evaluation on parametric curves and surfaces remains costly for large and complex models. Fast, state-of-the-art GWN approaches leverage a spatial index to approximate the GWN, typically coupled with a Taylor expansion which approximates the GWN contribution for far clusters of geometric primitives. However, such methods operate only on discrete inputs such as triangle meshes and point clouds, and would introduce containment errors near boundaries if applied to curved input. We extend support for fast GWN evaluation over arbitrary collections of NURBS curves in 2D and trimmed NURBS patches in 3D via a Bounding Volume Hierarchy that stores efficiently precomputed moment data in the hierarchy nodes. When querying the hierarchy, approximations for far clusters are used alongside direct evaluation for nearby NURBS primitives, achieving sub-linear complexity while preserving the geometric features in the vicinity of the query point. Central to our performance improvements is an adaptive subdivision strategy for NURBS primitives during a preprocessing phase, creating better spatial partitions while retaining the same accuracy for containment decisions as a direct evaluation. We demonstrate the performance and accuracy of our approach across a large collection of 2D and 3D datasets.

Computer science↗

Exploring the Use of Novel Spatial Accelerators in Scientific Applications

Driven by the need to find alternative accelerators which can viably replace GPUs in next-generation Supercomputing systems, this paper proposes a methodology to enable agile application/hardware co-design. The application-first methodology provides the ability to come up with design of accelerators while working with real-world workloads, available accelerators, and system software. The iterative design process targets a set of kernels in a workload for performance estimates that can prune the design space for later phases of detailed architectural evaluations. To this effect, in this paper, a novel data-parallel device model is introduced that simulates the latency of performance-sensitive operations in an accelerator including data transfers and kernel computation using multi-core CPUs. The use of off-the-shelf simulators, such as pre-RTL simulator Aladdin or multiple tools available for exploring the design of deep neural network accelerators (e.g., Timeloop) is demonstrated for evaluation of various accelerator designs using applications with realistic inputs. Examples of multiple device configurations that are instantiable in a system are explored to evaluate the performance benefit of deploying novel accelerators. The proposed device is integrated with a programming model and system software to potentially explore the impacts of high-level programming languages/compilers and low-level effects such as task scheduling on multiple accelerators. We analyze our methodology for a set of applications that represent high-performance computing (HPC) and graph analytics. The applications include a computational chemistry kernel realized using tensor contractions, triangle counting, GraphSAGE and Breadth-first Search. These applications include kernels such as dense matrix-dense matrix multiplication, sparse matrix-spare matrix multiplication, and sparse matrix-dense vector multiplication. Our results indicate potential performance benefits and insights for system design by including accelerators that realize these kernels along-side general purpose accelerators.

AI, codesign, Accelerated Computing, Modeling and ↗

Elevated carbon dioxide accelerates the spatial turnover of soil microbial communities

Although elevated CO 2 (eCO 2 ) significantly affects the -diversity, composition, function, interaction and dynamics of soil microbial communities at the local scale, little is known about eCO 2 impacts on the geographic distribution of micro-organisms regionally or globally. Here, we examined the -diversity of 110 soil microbial communities across six free air CO 2 enrichment (FACE) experimental sites using a high-throughput functional gene array. The -diversity of soil microbial communities was significantly (P<0.05) correlated with geographic distance under both CO 2 conditions, but declined significantly (P<0.05) faster at eCO 2 with a slope of -0.0250 than at ambient CO 2 (aCO 2 ) with a slope of -0.0231 although it varied within each individual site, indicating that the spatial turnover rate of soil microbial communities was accelerated under eCO 2 at a larger geographic scale (e.g. regionally). Both distance and soil properties significantly (P<0.05) contributed to the observed microbial -diversity. Furthermore, this study provides new hypotheses for further understanding their assembly mechanisms that may be especially important as global CO 2 continues to increase.

59 BASIC BIOLOGICAL SCIENCES↗

Energy-efficient specialization of functional units in a Coarse-Grained Reconfigurable Array

Functional units provide the backbone of any spatial accelerator by providing the computing resources. The desire for having rich and expensive functional units is in tension with producing a regular and energy-efficient computing fabric. This paper explores the design trade-off between complex, universal functional units and simpler, limited functional units. We show that a modest amount of specialization reduces the area-delay-energy product of an optimized architecture to 0.86x a baseline architecture. Furthermore, we provide a design guideline that allows an architect to customize the contents of the computing fabric just by examining the profile of benchmarks within the application domains. Functional units are the core of compute-intensive spatial accelerators. They perform the computation of interest with support from local storage and communication structures. Ideally, the functional units will provide rich functionality, supporting operations ranging from simple addition, to fused multiply-adds, to advanced transcendental functions and domain specific operations like add-compare-select. However, the total opportunity cost to support the more complex operations is a function of the cost of the hardware, the rate of occurrence of the operation in the application domain, and the inefficiency of emulating the operation with simpler operators. Examples of operations that are typically emulated in spatial accelerators are division and trigonometric functions, which can be solved using table-lookup based algorithms and the CORDIC algorithm. One reason to avoid having direct hardware support for complex operations in a tiled architecture like a Coarse-Grained Reconfigurable Array (CGRA) is that the expensive hardware will typically need to be replicated in some or all of the architecture's tiles. Tiled architecture are designed such that their tiles are either homogeneous or heterogeneous. Homogeneous architectures are simpler to design but heterogeneous architectures can be more efficient. Generally, CGRAs try to support a rich set of operations with the smallest possible set of hardware devices.

97 MATHEMATICS AND COMPUTING↗

Multiple quasi-monoenergetic electron beams from laser-wakefield acceleration with spatially structured laser pulse

By adjusting the focus geometry of a spatially structured laser pulse, single, double, and treble quasi-monoenergetic electron beams were generated, respectively, in laser-wakefield acceleration. Single electron beam was produced as focusing the laser pulse to a single spot. While focusing the laser pulse to two spots that are approximately equal in energy and size and intense enough to form their own filaments, two electron beams were produced. Moreover, with a proper distance between those two focal spots, three electron beams emerged with a certain probability owing to the superposition of the diffractions of those two spots. The energy spectra of the multiple electron beams are quasi-monoenergetic, which are different from that of the large energy spread beams produced due to the longitudinal multiple-injection in the single bubble.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

CMFD acceleration of spatial domain-decomposed neutron transport problems

A significant limitation to parallelizing the solution of neutron transport problems is the need for sweeps across the entirety of the problem domain. Angular domain decomposition is common practice, as the equations for each direction are independent aside from their shared scattering/fission source. Accordingly, spatial domain decomposition does not naturally arise in the transport equations and is therefore less frequent in practice. In this paper, we show that a neutron transport domain can be straightforwardly divided into independent, parallelizable sweep regions, globally linked with the standard CMFD method, with an additional update equation. We verify, theoretically (via Fourier analysis) and computationally, that the convergence properties of this method are stable and nominally as rapid as standard CMFD. (authors)

22 GENERAL STUDIES OF NUCLEAR REACTORS↗