Search NASA⌕ Search

SEARCH · Search NASA

Results for “Memory Bandwidth”

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
All sourcesNASA NTRSDOE OSTI
DOE OSTI2018

Pushing memory bandwidth limitations through efficient implementations of Block-Krylov space solvers on GPUs

Lattice quantum chromodynamics simulations in nuclear physics have benefited from a tremendous number of algorithmic advances such as multigrid and eigenvector deflation. These improve the time to solution but do not alleviate the intrinsic memory-bandwidth constraints of the matrix-vector operation dominating iterative solvers. Batching this operation for multiple vectors and exploiting cache and register blocking can yield a super-linear speed up. Block-Krylov solvers can naturally take advantage of such batched matrix-vector operations, further reducing the iterations to solution by sharing the Krylov space between solves. However, practical implementations typically suffer from the quadratic scaling in the number of vector-vector operations. Here, using the QUDA library, we present an implementation of a block-CG solver on NVIDIA GPUs which reduces the memory-bandwidth complexity of vector-vector operations from quadratic to linear. We present results for the HISQ discretization, showing a 5x speedup compared to highly-optimized independent Krylov solves on NVIDIA's SaturnV cluster.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2018

High-bandwidth prefetcher for high-bandwidth memory

A method for prefetching data into a cache is provided. The method allocates an outstanding request buffer ("ORB"). The method stores in an address field of the ORB an address and a number of blocks. The method issues prefetch requests for a degree number of blocks starting at the address. When a prefetch response is received for all the prefetch requests, the method adjusts the address of the next block to prefetch and adjusts the number of blocks remaining to be retrieved and then issues prefetch requests for a degree number of blocks starting at the adjusted address. The prefetching pauses when a maximum distance between the reads of the prefetched blocks and the last prefetched block is reached. When a read request for a prefetched block is received, the method resumes prefetching when a resume criterion is satisfied.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2019

Thread Batching for High-performance Energy-efficient GPU Memory Design

Massive multi-threading in GPU imposes tremendous pressure on memory subsystems. Due to rapid growth in thread-level parallelism of GPU and slowly improved peak memory bandwidth, memory becomes a bottleneck of GPU’s performance and energy efficiency. In this article, we propose an integrated architectural scheme to optimize the memory accesses and therefore boost the performance and energy efficiency of GPU. First, we propose a thread batch enabled memory partitioning (TEMP) to improve GPU memory access parallelism. In particular, TEMP groups multiple thread blocks that share the same set of pages into a thread batch and applies a page coloring mechanism to bound each stream multiprocessor (SM) to the dedicated memory banks. After that, TEMP dispatches the thread batch to an SM to ensure high-parallel memory-access streaming from the different thread blocks. Second, a thread batch-aware scheduling (TBAS) scheme is introduced to improve the GPU memory access locality and to reduce the contention on memory controllers and interconnection networks. Experimental results show that the integration of TEMP and TBAS can achieve up to 10.3% performance improvement and 11.3% DRAM energy reduction across diverse GPU applications. We also evaluate the performance interference of the mixed CPU+GPU workloads when they are run on a heterogeneous system that employs our proposed schemes. Our results show that a simple solution can effectively ensure the efficient execution of both GPU and CPU applications.

Computer Science↗
DOE OSTI2020

Quantitative Performance Assessment of Proxy Apps and Parents (ECP Proxy App Project Milestone ADCD-504-9)

This report presents highlights of these efforts. Section 2 describes work that has been done to compare the performance of proxy applications on AMD MI60 vs. Nvidia V100 GPUs. So far only a small set of ECP proxies are running on AMD GPUs, but we will continue to expand this analysis as additional proxies become available. We find that although the MI60 and V100 have nearly the same measured memory bandwidth, memory bound proxy app kernels perform 20-30% worse on the MI60. Further work is needed to refine these comparisons to determine whether the root cause is due to differences in the hardware, software stack, platform specific optimization, or some combination of the three. Section 3 describes our continuing effort to find methods to accurately assess the similarity of proxies and parents. We have recently seen very encouraging results using a cosine similarity metric. This technique uses the angle between two vectors of hardware performance counters to characterize the similarity (or difference) between two applications or proxies. We show not only that several widely used proxies are highly similar to their parents, but also that they differ from non-related codes. We also show that cosine similarity can be used to identify gaps and redundancies in suites and even to gain insight into the effects of architectural differences between platforms. Our work on assessing the Exascale toolchain is ongoing. Our successes with performance measurement tools are evident from the data provided in this report. However, our assessments across the broader tool chain are still too incomplete to provide a meaningful report at this time. We will continue to assess tools and work with vendors and third party developers as issues are identified.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2016

Quicksilver

Quicksilver is proxy application for the Monte Carlo Transport code, Mercury. The purpose of the Quicksilver mini-app is to create a freely distributable code base that represents the key elements of the Mercury workload by solving a simplified time-dependent neutron transport problem. As an open source code, Quicksilver will be hosted on the CZ git/stash repo alongside other mini-apps in the ACS so-design suite. It will be used by vendors and other partners for co-design activities such as identifying root causes of Mercury performance bottlenecks on future architectures (i.e., memory bandwidth, memory latency, flop intensity) and developing Quicksilver will be used within LLNL as a nimble platform to evaluate programming models, algorithms, etc. that might be useful in Mercury. Quicksilver may also become a replacement for MCB in future procurement benchmarks, as well as a way to engage vendors in the Trinity and Sierra Centers of Excellence as well as other co-design efforts.

Richards, David↗
DOE OSTI2017

Fusion PIC code performance analysis on the Cori KNL system

We study the attainable performance of Particle-In-Cell codes on the Cori KNL system by analyzing a miniature particle push application based on the fusion PIC code XGC1. We start from the most basic building blocks of a PIC code and build up the complexity to identify the kernels that cost the most in performance and focus optimization efforts there. Particle push kernels operate at high AI and are not likely to be memory bandwidth or even cache bandwidth bound on KNL. Therefore, we see only minor benefits from the high bandwidth memory available on KNL, and achieving good vectorization is shown to be the most beneficial optimization path with theoretical yield of up to 8x speedup on KNL. In practice we are able to obtain up to a 4x gain from vectorization due to limitations set by the data layout and memory latency.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Efficient GPU Implementation of Automatic Differentiation for Computational Fluid Dynamics

Many scientific and engineering applications require repeated calculations of derivatives of output functions with respect to input parameters. Automatic Differentiation (AD) is a method that automates derivative calculations and can significantly speed up code development. In Computational Fluid Dynamics (CFD), derivatives of flux functions with respect to state variables (Jacobian) are needed for efficient solutions of the nonlinear governing equations. AD of flux functions on graphics processing units (GPUs) is challenging as flux computations involve many intermediate variables that create high register pressure and require significant memory traffic because of the need to store the derivatives. This paper presents a forward-mode AD method based on multivariate dual numbers that addresses these challenges and simultaneously reduces the floating-point operation count. The dimension of the multivariate dual numbers is optimized for performance. The flux computations are restructured to minimize the number of temporary variables and reduce register pressure. For effective utilization of memory bandwidth, shared memory is used to store the local flux Jacobian. This AD implementation is compared with several other Jacobian implementations on an NVIDIA V100 GPU (V100). For three-dimensional perfect-gas compressible-flow equations implemented in a practical CFD code, the AD implementation of a flux Jacobian based on multivariate dual numbers of dimension 5 outperforms all other GPU AD implementations on V100. Its performance is comparable with the optimized hand-differentiated version. Finally, the implementation achieves 75% of the peak floating-point throughput and 61 % of the peak global device memory bandwidth usage.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2021

HAM: Hotspot-Aware Manager for Improving Communications with 3D-Stacked Memory

merging High-Performance Computing (HPC) workloads, such as graph analytics, machine learning, and big data science, are data-intensive. Data-intensive workloads usually present fine-grained memory accesses with limited or no data locality, and thus incur frequent cache misses and low utilization of memory bandwidth. 3D-stacked memory devices such as Hybrid Memory Cube (HMC) and High Bandwidth Memory (HBM) can provide significantly higher bandwidth than conventional memory modules. However, the traditional interfaces and optimization methods for JEDEC DDR devices do not allow to fully exploit the potential performance of 3D-stacked memory with the massive amount of irregular memory accesses of data-intensive applications. In this paper, we propose a novel Hotspot-Aware Manager (HAM) infrastructure for 3D-stacked memory devices capable of optimizing memory access streams via request aggregation, hotspot detection, and in-memory prefetching. %and an associated hotspot-aware page policy. We present the HAM design and implementation, and simulate it on a system using RISC-V embedded cores with attached HMC devices. We extensively evaluate HAM with over 12 benchmarks and applications representing diverse irregular memory access patterns. The results show that, on average, HAM reduces redundant requests by 37.51\% and increases the prefetch buffer hit rate by 4.2 times, compared to a baseline streaming prefetcher. On the selected benchmark set, HAM provides performance gains of 21.81\% in average (up to 34.28\%) and power savings of 35.07\% over a standard 3D-stacked memory.

Wang, Xi↗
DOE OSTI2023

Resource-aware compression

Systems, apparatuses, and methods for implementing a multi-tiered approach to cache compression are disclosed. A cache includes a cache controller, light compressor, and heavy compressor. The decision on which compressor to use for compressing cache lines is made based on certain resource availability such as cache capacity or memory bandwidth. This allows the cache to opportunistically use complex algorithms for compression while limiting the adverse effects of high decompression latency on system performance. To address the above issue, the proposed design takes advantage of the heavy compressors for effectively reducing memory bandwidth in high bandwidth memory (HBM) interfaces as long as they do not sacrifice system performance. Accordingly, the cache combines light and heavy compressors with a decision-making unit to achieve reduced off-chip memory traffic without sacrificing system performance.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2016

Announcing Supercomputer Summit

Summit is the next leap in leadership-class computing systems for open science. With Summit we will be able to address, with greater complexity and higher fidelity, questions concerning who we are, our place on earth, and in our universe. Summit will deliver more than five times the computational performance of Titan’s 18,688 nodes, using only approximately 3,400 nodes when it arrives in 2017. Like Titan, Summit will have a hybrid architecture, and each node will contain multiple IBM POWER9 CPUs and NVIDIA Volta GPUs all connected together with NVIDIA’s high-speed NVLink. Each node will have over half a terabyte of coherent memory (high bandwidth memory + DDR4) addressable by all CPUs and GPUs plus 800GB of non-volatile RAM that can be used as a burst buffer or as extended memory. To provide a high rate of I/O throughput, the nodes will be connected in a non-blocking fat-tree using a dual-rail Mellanox EDR InfiniBand interconnect. Upon completion, Summit will allow researchers in all fields of science unprecedented access to solving some of the world’s most pressing challenges.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2020

ZFP Hardware Implementation

As core counts increase in new HPC systems with comparatively little increase in memory bandwidth, the trend is an effective decrease in memory bandwidth per core. Other bandwidth limitations in HPC systems exist between CPU and GPU memory, between system nodes, and between node memory and storage. Compression of floating-point data has the potential to reduce data movement and pressure across these communication channels. Furthermore, it has the potential to reduce the footprint of floating-point arrays stored in memory. ZFP, implemented in software, is gaining traction as an effective method in floating-point compression; however, performance gains are limited to the spare compute cycles available before reaching the bandwidth limitations of the communication channel. A hardware implementation of ZFP has the potential to raise the bar on performance. From the inception of ZFP, it was designed to accommodate a hardware implementation.

42 ENGINEERING↗
DOE OSTI2023

Runtime extension for neural network training with heterogeneous memory

Systems, apparatuses, and methods for managing buffers in a neural network implementation with heterogeneous memory are disclosed. A system includes a neural network coupled to a first memory and a second memory. The first memory is a relatively low-capacity, high-bandwidth memory while the second memory is a relatively high-capacity, low-bandwidth memory. During a forward propagation pass of the neural network, a run-time manager monitors the usage of the buffers for the various layers of the neural network. During a backward propagation pass of the neural network, the run-time manager determines how to move the buffers between the first and second memories based on the monitored buffer usage during the forward propagation pass. As a result, the run-time manager is able to reduce memory access latency for the layers of the neural network during the backward propagation pass.

Mappouras, Georgios↗
DOE OSTI2017

Two-level main memory co-design: Multi-threaded algorithmic primitives, analysis, and simulation

A challenge in computer architecture is that processors often cannot be fed data from DRAM as fast as CPUs can consume it. Therefore, many applications are memory-bandwidth bound. With this motivation and the realization that traditional architectures (with all DRAM reachable only via bus) are insufficient to feed groups of modern processing units, vendors have introduced a variety of non-DDR 3D memory technologies (Hybrid Memory Cube (HMC),Wide I/O 2, High Bandwidth Memory (HBM)). These offer higher bandwidth and lower power by stacking DRAM chips on the processor or nearby on a silicon interposer. We will call these solutions “near-memory,” and if user-addressable, “scratchpad.” High-performance systems on the market now offer two levels of main memory: near-memory on package and traditional DRAM further away. In the near term we expect the latencies near-memory and DRAM to be similar. Here, it is natural to think of near-memory as another module on the DRAM level of the memory hierarchy. Vendors are expected to offer modes in which the near memory is used as cache, but we believe that this will be inefficient.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2017

Embedded ensemble propagation for improving performance, portability, and scalability of uncertainty quantification on emerging computational architectures

In this study, quantifying simulation uncertainties is a critical component of rigorous predictive simulation. A key component of this is forward propagation of uncertainties in simulation input data to output quantities of interest. Typical approaches involve repeated sampling of the simulation over the uncertain input data, and can require numerous samples when accurately propagating uncertainties from large numbers of sources. Often simulation processes from sample to sample are similar and much of the data generated from each sample evaluation could be reused. We explore a new method for implementing sampling methods that simultaneously propagates groups of samples together in an embedded fashion, which we call embedded ensemble propagation. We show how this approach takes advantage of properties of modern computer architectures to improve performance by enabling reuse between samples, reducing memory bandwidth requirements, improving memory access patterns, improving opportunities for fine-grained parallelization, and reducing communication costs. We describe a software technique for implementing embedded ensemble propagation based on the use of C++ templates and describe its integration with various scientific computing libraries within Trilinos. We demonstrate improved performance, portability and scalability for the approach applied to the simulation of partial differential equations on a variety of CPU, GPU, and accelerator architectures, including up to 131,072 cores on a Cray XK7 (Titan).

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

A compute-bound formulation of Galerkin model reduction for linear time-invariant dynamical systems

This work aims to advance computational methods for projection-based reduced-order models (ROMs) of linear time-invariant (LTI) dynamical systems. For such systems, current practice relies on ROM formulations expressing the state as a rank-1 tensor (i.e., a vector), leading to computational kernels that are memory bandwidth bound and, therefore, ill-suited for scalable performance on modern architectures. This weakness can be particularly limiting when tackling many-query studies, where one needs to run a large number of simulations. This work introduces a reformulation, called rank-2 Galerkin, of the Galerkin ROM for LTI dynamical systems which converts the nature of the ROM problem from memory bandwidth to compute bound. We present the details of the formulation and its implementation, and demonstrate its utility through numerical experiments using, as a test case, the simulation of elastic seismic shear waves in an axisymmetric domain. We quantify and analyze performance and scaling results for varying numbers of threads and problem sizes. In conclusion, we present an end-to-end demonstration of using the rank-2 Galerkin ROM for a Monte Carlo sampling study. We show that the rank-2 Galerkin ROM is one order of magnitude more efficient than the rank-1 Galerkin ROM (the current practice) and about 970 times more efficient than the full-order model, while maintaining accuracy in both the mean and statistics of the field.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2016

Evaluating the Opportunities for Multi-Level Memory – An ASC 2016 L2 Milestone

The next two Advanced Technology platforms for the ASC program will feature complex memory hierarchies – in the Trinity supercomputer being deployed in 2016, Intel’s Knights Landing processors will feature 16GB of on-package, high-bandwidth memory, combined with a larger capacity DDR4 memory and in 2018, the Sierra machine deployed at Lawrence Livermore National Laboratory will feature powerful compute nodes containing POWER9 processors with large capacity memories and an array of coherent GPU accelerators also with high bandwidth memories.

97 MATHEMATICS AND COMPUTING↗