Search NASA⌕ Search

SEARCH · Search NASA

Results for “Memory Hierarchy”

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 73 records · Page 4

The Gremlin Framework

The Gremlin sofrware package is a performance analysis approach targeted to support the Co-Design process for future systems. It consists of a series of modules that can be used to alter a machine's behavior with the goal of emulating future machine properties. The modules can be divided into several classes; the most significant ones are detailed below. PowGre is a series of modules that help explore the power consumption properties of applications and to determine the impact of power constraints on applications. Most of them use low-level processor interfaces to directly control voltage and frequency settings as well as per nodes, socket, or memory power bounds. MemGre are memory Gremlins and implement a new performance analysis technique that captures the application's effective use of the storage capacity of different levels of the memory hierarchy as well as the bandwidth between adjacent levels. The approach models various memory components as resources and measures how much of each resource the application uses from the application's own perspective. To the application a given amount of a resource is "used" if not having this amount will degrade the application's performance. This is in contrast to the hardware-centric perspective that considers "use" as any hardware action that utilizes the resource, even if it has no effect on performance. ResGre are Gremlins that use fault injection techniques to emulate higher fault rates than currently present in today's systems. Faults can be injected through various means, including network interposition, static analysis, and code modification, or direct application notification. ResGre also includes patches to previously released LLNL codes that can counteract and react to injected failures.

Source record↗

Institute for Sustained Performance, Energy, and Resilience (SUPER): University of Utah Final Report

Over the course of this project, computational scientists working on behalf of the Department of Energy's Office of Science (DOE SC) were exploiting a new generation of petascale computing resources to make previously inaccessible discoveries in a broad range of disciplines including physics, chemistry and material science. The computational systems underpinning this work increased in performance potential from tens to hundreds of Pflop/s, and evolved significantly from those in use at the beginning of the project in terms of available concurrency; wider use of accelerators such as graphics processing units(GPUs); and changes to the memory hierarchy including the incorporation of a new generation of persistent devices (e.g., phase change memory). To ensure that DOE's computational scientists can successfully exploit this emerging generation of high performance computing (HPC) systems, the University of Southern California (USC) assembled a broad team of computer scientists with the expertise to address the most pressing challenges of DOE's computing systems: (a) end-to-end performance optimization, including single-node performance, interprocessor communication, load balancing and I/O; (b) performance portability for heterogeneous systems, including accelerators; (c) minimizing energy consumption; (d) resilient computation; and, (e) exploiting new memory technology.

97 MATHEMATICS AND COMPUTING↗

A Parallel Graph Environment for Real-World Data Analytics Workflows

Economic competitiveness and national security depend increasingly on the insightful analysis of large data sets. The diversity of real-world data sources and analytic workflows impose challenging hardware and software requirements for parallel graph platforms. The irregular nature of graph methods is not supported well by the deep memory hierarchies of conventional distributed systems, requiring new processor and runtime system designs to tolerate memory and synchronization latencies. Moreover, the efficiency of relational table operations and matrix computations are not attainable when data is stored in common graph data structures. In this paper, we present HAGGLE, a high-performance, scalable data analytics platform. The platform’s hybrid data model supports a variety of distributed, thread-safe data structures, parallel programming constructs, and persistent and streaming data. An abstract runtime layer enables us to map the stack to conventional, distributed computer systems with accelerators. The runtime uses multithreading, active messages, and data aggregation to hide memory and synchronization latencies on large-scale systems.

Graph Analytics↗

The Efficiency and the Scalability of an Explicit Operator on an IBM POWER4 System

We present an evaluation of the efficiency and the scalability of an explicit CFD operator on an IBM POWER4 system. The POWER4 architecture exhibits a common trend in HPC architectures: boosting CPU processing power by increasing the number of functional units, while hiding the latency of memory access by increasing the depth of the memory hierarchy. The overall machine performance depends on the ability of the caches-buses-fabric-memory to feed the functional units with the data to be processed. In this study we evaluate the efficiency and scalability of one explicit CFD operator on an IBM POWER4. This operator performs computations at the points of a Cartesian grid and involves a few dozen floating point numbers and on the order of 100 floating point operations per grid point. The computations in all grid points are independent. Specifically, we estimate the efficiency of the RHS operator (SP of NPB) on a single processor as the observed/peak performance ratio. Then we estimate the scalability of the operator on a single chip (2 CPUs), a single MCM (8 CPUs), 16 CPUs, and the whole machine (32 CPUs). Then we perform the same measurements for a chache-optimized version of the RHS operator. For our measurements we use the HPM (Hardware Performance Monitor) counters available on the POWER4. These counters allow us to analyze the obtained performance results.

Frumkin, Michael↗

State recovery and lockstep execution restart in a system with multiprocessor pairing

System, method and computer program product for a multiprocessing system to offer selective pairing of processor cores for increased processing reliability. A selective pairing facility is provided that selectively connects, i.e., pairs, multiple microprocessor or processor cores to provide one highly reliable thread (or thread group). Each paired microprocessor or processor cores that provide one highly reliable thread for high-reliability connect with a system components such as a memory "nest" (or memory hierarchy), an optional system controller, and optional interrupt controller, optional I/O or peripheral devices, etc. The memory nest is attached to a selective pairing facility via a switch or a bus. Each selectively paired processor core is includes a transactional execution facility, whereing the system is configured to enable processor rollback to a previous state and reinitialize lockstep execution in order to recover from an incorrect execution when an incorrect execution has been detected by the selective pairing facility.

97 MATHEMATICS AND COMPUTING↗

Multiprocessor switch with selective pairing

System, method and computer program product for a multiprocessing system to offer selective pairing of processor cores for increased processing reliability. A selective pairing facility is provided that selectively connects, i.e., pairs, multiple microprocessor or processor cores to provide one highly reliable thread (or thread group). Each paired microprocessor or processor cores that provide one highly reliable thread for high-reliability connect with a system components such as a memory "nest" (or memory hierarchy), an optional system controller, and optional interrupt controller, optional I/O or peripheral devices, etc. The memory nest is attached to a selective pairing facility via a switch or a bus

97 MATHEMATICS AND COMPUTING↗

Mapping Dense LU Factorization on Multicore Supercomputer Nodes

Dense LU factorization is a prominent benchmark used to rank the performance of supercomputers. Many implementations use block-cyclic distributions of matrix blocks onto a two-dimensional process grid. The process grid dimensions drive a trade-off between communication and computation and are architecture- and implementation-sensitive. The critical panel factorization steps can be made less communication-bound by overlapping asynchronous collectives for pivoting with the computation of rank-k updates. By shifting the computation-communication trade-off, a modified block-cyclic distribution can beneficially exploit more available parallelism on the critical path, and reduce panel factorization's memory hierarchy contention on now-ubiquitous multicore architectures. During active panel factorization, rank-1 updates stream through memory with minimal reuse. In a column-major process grid, the performance of this access pattern degrades as too many streaming processors contend for access to memory. A block-cyclic mapping in the row-major order does not encounter this problem, but consequently sacrifices node and network locality in the critical pivoting steps. We introduce 'striding' to vary between the two extremes of row- and column-major process grids. The maximum available parallelism in the critical path work (active panel factorization, triangular solves, and subsequent broadcasts) is bounded by the length or width of the process grid. Increasing one dimension of the process grid decreases the number of distinct processes and nodes in the other dimension. To increase the harnessed parallelism in both dimensions, we start with a tall process grid. We then apply periodic 'rotation' to this grid to restore exploited parallelism along the row to previous levels. As a test-bed for further mapping experiments, we describe a dense LU implementation that allows a block distribution to be defined as a general function of block to processor. Other mappings can be tested wit...

Computer Science↗

Multiprocessor architectural study

An architectural design study was made of a multiprocessor computing system intended to meet functional and performance specifications appropriate to a manned space station application. Intermetrics, previous experience, and accumulated knowledge of the multiprocessor field is used to generate a baseline philosophy for the design of a future SUMC* multiprocessor. Interrupts are defined and the crucial questions of interrupt structure, such as processor selection and response time, are discussed. Memory hierarchy and performance is discussed extensively with particular attention to the design approach which utilizes a cache memory associated with each processor. The ability of an individual processor to approach its theoretical maximum performance is then analyzed in terms of a hit ratio. Memory management is envisioned as a virtual memory system implemented either through segmentation or paging. Addressing is discussed in terms of various register design adopted by current computers and those of advanced design.

Kosmala, A. L.↗

Instruction Roofline: An insightful visual performance model for GPUs

The Roofline performance model provides an intuitive approach to identify performance bottlenecks and guide performance optimization. However, the classic FLOP-centric approach is inappropriate for the emerging applications that perform more integer operations than floating point operations. In this article, we reintroduce our Instruction Roofline Model on NVIDIA GPUs and expand our evaluation of it. The Instruction Roofline incorporates instructions and memory transactions across all memory hierarchies together, and provides more performance insights than the FLOP-oriented Roofline Model, that is, instruction throughput, stride memory access patterns, bank conflicts, and thread predication. We use our Instruction Roofline methodology to analyze eight proxy applications: HPGMG from AMReX, Matrix Transpose benchmarks, ADEPT from MetaHipMer's sequence alignment phase, EXTENSION from MetaHipMer's local assembly phase, CUSP, cuSPARSE, cudaTensorCoreGemm, and cuBLAS. We demonstrate the ability of our methodology to understand various aspects of performance and performance bottlenecks on NVIDIA GPUs and motivate code optimizations.

Ding, N↗

Highly Efficient Compensation-Based Parallelism for Wavefront Loops on GPUs

Wavefront loops are widely used in many scientific applications, e.g., partial differential equation (PDE) solvers and sequence alignment tools. However, due to the data dependencies in wavefront loops, it is challenging to fully utilize the abundant compute units of GPUs and to reuse data through their memory hierarchy. Existing solutions can only optimize for these factors to a limited extent. For example, tiling-based methods optimize memory access but may result in load imbalance; while compensation-based methods, which change the original order of computation to expose more parallelism and then compensate for it, suffer from both global synchronization overhead and limited generality. In this paper, we first prove under which circumstances that breaking data dependencies and properly changing the sequence of computation operators in our compensation-based method does not affect the correctness of results. Based on this analysis, we design a highly efficient compensation-based parallelism on GPUs. Our method provides weighted scan-based GPU kernels to optimize the computation and combines with the tiling method to optimize memory access and synchronization. The performance results on the NVIDIA K80 and P100 GPU platforms demonstrate that our method can achieve significant improvements for four types of real-world application kernels over the state-of-the-art research.

Hou, Kaixi↗

PCM-Based Durable Write Cache for Fast Disk I/O

Flash based solid-state devices (FSSDs) have been adopted within the memory hierarchy to improve the performance of hard disk drive (HDD) based storage system. However, with the fast development of storage-class memories, new storage technologies with better performance and higher write endurance than FSSDs are emerging, e.g., phase-change memory (PCM). Understanding how to leverage these state-of-the-art storage technologies for modern computing systems is important to solve challenging data intensive computing problems. In this paper, we propose to leverage PCM for a hybrid PCM-HDD storage architecture. We identify the limitations of traditional LRU caching algorithms for PCM-based caches, and develop a novel hash-based write caching scheme called HALO to improve random write performance of hard disks. To address the limited durability of PCM devices and solve the degraded spatial locality in traditional wear-leveling techniques, we further propose novel PCM management algorithms that provide effective wear-leveling while maximizing access parallelism. We have evaluated this PCM-based hybrid storage architecture using applications with a diverse set of I/O access patterns. Our experimental results demonstrate that the HALO caching scheme leads to an average reduction of 36.8% in execution time compared to the LRU caching scheme, and that the SFC wear leveling extends the lifetime of PCM by a factor of 21.6.

99 GENERAL AND MISCELLANEOUS↗

RACB: Resource Aware Cache Bypass on GPUs

Caches are universally used in computing systems to hide long off-chip memory access latencies. Unlike CPUs, massive threads running simultaneously on GPUs bring a tremendous pressure on memory hierarchy. As a result, the limitation of cache resources becomes a bottleneck for a GPU to exploit thread-level parallelism (TLP) and memory-level parallelism (MLP) and achieve high performance. In this paper, we propose a mechanism to bypass L1D and L2 cache based on the availability of cache resources. Our proposed mechanism is based on the observation that a huge number of stalls coming from limited cache resources prohibit GPUs from providing a higher throughput. So we propose Resource Aware Cache Bypass (RACB) with minor hardware changes to eliminate such stalls to improve performance. We examine the effectiveness of this approach when applied to L1D and L2 cache separately as well as together. Evaluation results with NVIDIA Computing SDK show that RACB generally improves performance the most when applied to both L1D and L2 cache, which is up to 88.05% and on an average of 16.73%, additionally, energy is saved up to 22.35% and on an average of 5.88% with minor hardware overheads.

Dai, Hongwen↗

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↗

A fast and memory-sparing probabilistic selection algorithm for the GPU

A fast and memory-sparing probabilistic top-N selection algorithm is implemented on the GPU. This probabilistic algorithm gives a deterministic result and always terminates. The use of randomization reduces the amount of data that needs heavy processing, and so reduces both the memory requirements and the average time required for the algorithm. This algorithm is well-suited to more general parallel processors with multiple layers of memory hierarchy. Probabilistic Las Vegas algorithms of this kind are a form of stochastic optimization and can be especially useful for processors having a limited amount of fast memory available.

99 GENERAL AND MISCELLANEOUS↗

Testing New Programming Paradigms with NAS Parallel Benchmarks

Over the past decade, high performance computing has evolved rapidly, not only in hardware architectures but also with increasing complexity of real applications. Technologies have been developing to aim at scaling up to thousands of processors on both distributed and shared memory systems. Development of parallel programs on these computers is always a challenging task. Today, writing parallel programs with message passing (e.g. MPI) is the most popular way of achieving scalability and high performance. However, writing message passing programs is difficult and error prone. Recent years new effort has been made in defining new parallel programming paradigms. The best examples are: HPF (based on data parallelism) and OpenMP (based on shared memory parallelism). Both provide simple and clear extensions to sequential programs, thus greatly simplify the tedious tasks encountered in writing message passing programs. HPF is independent of memory hierarchy, however, due to the immaturity of compiler technology its performance is still questionable. Although use of parallel compiler directives is not new, OpenMP offers a portable solution in the shared-memory domain. Another important development involves the tremendous progress in the internet and its associated technology. Although still in its infancy, Java promisses portability in a heterogeneous environment and offers possibility to "compile once and run anywhere." In light of testing these new technologies, we implemented new parallel versions of the NAS Parallel Benchmarks (NPBs) with HPF and OpenMP directives, and extended the work with Java and Java-threads. The purpose of this study is to examine the effectiveness of alternative programming paradigms. NPBs consist of five kernels and three simulated applications that mimic the computation and data movement of large scale computational fluid dynamics (CFD) applications. We started with the serial version included in NPB2.3. Optimization of memory and cache usage was applied to several benchmarks, noticeably BT and SP, resulting in better sequential performance. In order to overcome the lack of an HPF performance model and guide the development of the HPF codes, we employed an empirical performance model for several primitives found in the benchmarks. We encountered a few limitations of HPF, such as lack of supporting the "REDISTRIBUTION" directive and no easy way to handle irregular computation. The parallelization with OpenMP directives was done at the outer-most loop level to achieve the largest granularity. The performance of six HPF and OpenMP benchmarks is compared with their MPI counterparts for the Class-A problem size in the figure in next page. These results were obtained on an SGI Origin2000 (195MHz) with MIPSpro-f77 compiler 7.2.1 for OpenMP and MPI codes and PGI pghpf-2.4.3 compiler with MPI interface for HPF programs.

Jin, H.↗

Roofline model toolkit: A practical tool for architectural and program analysis

We present preliminary results of the Roofline Toolkit for multicore, many core, and accelerated architectures. This paper focuses on the processor architecture characterization engine, a collection of portable instrumented micro benchmarks implemented with Message Passing Interface (MPI), and OpenMP used to express thread-level parallelism. These benchmarks are specialized to quantify the behavior of different architectural features. Compared to previous work on performance characterization, these microbenchmarks focus on capturing the performance of each level of the memory hierarchy, along with thread-level parallelism, instruction-level parallelism and explicit SIMD parallelism, measured in the context of the compilers and run-time environments. We also measure sustained PCIe throughput with four GPU memory managed mechanisms. By combining results from the architecture characterization with the Roofline model based solely on architectural specifications, this work offers insights for performance prediction of current and future architectures and their software systems. To that end, we instrument three applications and plot their resultant performance on the corresponding Roofline model when run on a Blue Gene/Q architecture.

97 MATHEMATICS AND COMPUTING↗

Experimental evaluation of multiprocessor cache-based error recovery

Several variations of cache-based checkpointing for rollback error recovery in shared-memory multiprocessors have been recently developed. By modifying the cache replacement policy, these techniques use the inherent redundancy in the memory hierarchy to periodically checkpoint the computation state. Three schemes, different in the manner in which they avoid rollback propagation, are evaluated. By simulation with address traces from parallel applications running on an Encore Multimax shared-memory multiprocessor, the performance effect of integrating the recovery schemes in the cache coherence protocol are evaluated. The results indicate that the cache-based schemes can provide checkpointing capability with low performance overhead but uncontrollable high variability in the checkpoint interval.

Janssens, Bob↗

Fast analysis of molecular dynamics trajectories with graphics processing units-Radial distribution function histogramming

The calculation of radial distribution functions (RDFs) from molecular dynamics trajectory data is a common and computationally expensive analysis task. The rate limiting step in the calculation of the RDF is building a histogram of the distance between atom pairs in each trajectory frame. Here we present an implementation of this histogramming scheme for multiple graphics processing units (GPUs). The algorithm features a tiling scheme to maximize the reuse of data at the fastest levels of the GPU's memory hierarchy and dynamic load balancing to allow high performance on heterogeneous configurations of GPUs. Several versions of the RDF algorithm are presented, utilizing the specific hardware features found on different generations of GPUs. We take advantage of larger shared memory and atomic memory operations available on state-of-the-art GPUs to accelerate the code significantly. The use of atomic memory operations allows the fast, limited-capacity on-chip memory to be used much more efficiently, resulting in a fivefold increase in performance compared to the version of the algorithm without atomic operations. The ultimate version of the algorithm running in parallel on four NVIDIA GeForce GTX 480 (Fermi) GPUs was found to be 92 times faster than a multithreaded implementation running on an Intel Xeon 5550 CPU. On this multi-GPU hardware, the RDF between two selections of 1,000,000 atoms each can be calculated in 26.9 s per frame. The multi-GPU RDF algorithms described here are implemented in VMD, a widely used and freely available software package for molecular dynamics visualization and analysis.

97 MATHEMATICS AND COMPUTING↗