Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “cache locality”

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

Engineering the CernVM-Filesystem as a High Bandwidth Distributed Filesystem for Auxiliary Physics Data

A common use pattern in the computing models of particle physics experiments is running many distributed applications that read from a shared set of data files. We refer to this data is auxiliary data, to distinguish it from (a) event data from the detector (which tends to be different for every job), and (b) conditions data about the detector (which tends to be the same for each job in a batch of jobs). Relatively speaking, conditions data also tends to be relatively small per job where both event data and auxiliary data are larger per job. Unlike event data, auxiliary data comes from a limited working set of shared files. Since there is spatial locality of the auxiliary data access, the use case appears to be identical to that of the CernVM- Filesystem (CVMFS). However, we show that distributing auxiliary data through CVMFS causes the existing CVMFS infrastructure to perform poorly. We utilize a CVMFS client feature called 'alien cache' to cache data on existing local high-bandwidth data servers that were engineered for storing event data. This cache is shared between the worker nodes at a site and replaces caching CVMFS files on both the worker node local disks and on the site's local squids. We have tested this alien cache with the dCache NFSv4.1 interface, Lustre, and the Hadoop Distributed File System (HDFS) FUSE interface, and measured performance. In addition, we use high-bandwidth data servers at central sites to perform the CVMFS Stratum 1 function instead of the low-bandwidth web servers deployed for the CVMFS software distribution function. We have tested this using the dCache HTTP interface. As a result, we have a design for an end-to-end high-bandwidth distributed caching read-only filesystem, using existing client software already widely deployed to grid worker nodes and existing file servers already widely installed at grid sites. Files are published in a central place and are soon available on demand throughout the grid and cached locally on the site with a convenient POSIX interface. This paper discusses the details of the architecture and reports performance measurements.

Dykstra, D.↗

Locality-aware and sharing-aware cache coherence for collections of processors

A cache coherence technique for operating a multi-processor system including shared memory includes allocating a cache line of a cache memory of a processor to a memory address in the shared memory in response to execution of an instruction of a program executing on the processor. The technique includes encoding a shared information state of the cache line to indicate whether the memory address is a shared memory address shared by the processor and a second processor, or a private memory address private to the processor, in response to whether the instruction is included in a critical section of the program, the critical section being a portion of the program that confines access to shared, writeable data.

Farmahini Farahani, Amin↗

Caching data from remote memories

An approach is disclosed that caches distant memories within the storage a local node. The approach provides a memory caching infrastructure that supports virtual addressing by utilizing memory in the local node as a cache of distant memories for data granules. The data granules are accessed along with metadata and an ECC associated with the data granule. The metadata is updated to indicate storage of the selected data granule in the cache.

Johns, Charles Ray↗

Library API for Z-Order Memory Layout

This library provides a simple-to-use API for implementing an altnerative to traditional row-major order in-memory layout, one based on a Morton- order space filling curve (SFC) , specifically, a Z-order variant of the Morton order curve. The library enables programmers to, after a simple initialization step, to convert a multidimensional array from row-major to Z- order layouts, then use a single, generic API call to access data from any arbitrary (i,j,k) location from within the array, whether it it be stored in row- major or z-order format. The motivation for using a SFC in-memory layout is for improved spatial locality, which results in increased use of local high speed cache memory. The basic idea is that with row-major order layouts, a data access to some location that is nearby in index space is likely far away in physical memory, resulting in poor spatial locality and slow runtime. On the other hand, with a SFC-based layout, accesses that are nearby in index space are much more likely to also be nearby in physical memory, resulting in much better spatial locality, and better runtime performance. Numerous studies over the years have shown significant runtime performance gains are realized by using a SFC-based memory layout compared to a row-major layout, sometimes by as much as 50%, which result from the better use of the memory and cache hierarchy that are attendant with a SFC-based layout (see, for example, [Beth2012]). This library implementation is intended for use with codes that work with structured, array-based data in 2 or 3 dimensions. It is not appropriate for use with unstructured or point-based data.

Bethel, E. Wes↗

The Fabric for Frontier Experiments Project at Fermilab

The FabrIc for Frontier Experiments (FIFE) project is a new, far-reaching initiative within the Fermilab Scientific Computing Division to drive the future of computing services for experiments at FNAL and elsewhere. It is a collaborative effort between computing professionals and experiment scientists to produce an end-to-end, fully integrated set of services for computing on the grid and clouds, managing data, accessing databases, and collaborating within experiments. FIFE includes 1) easy to use job submission services for processing physics tasks on the Open Science Grid and elsewhere, 2) an extensive data management system for managing local and remote caches, cataloging, querying, moving, and tracking the use of data, 3) custom and generic database applications for calibrations, beam information, and other purposes, 4) collaboration tools including an electronic log book, speakers bureau database, and experiment membership database. All of these aspects will be discussed in detail. FIFE sets the direction of computing at Fermilab experiments now and in the future, and therefore is a major driver in the design of computing services worldwide.

Kirby, Michael↗

Performance Debugging and Tuning of Flash-X with Data Analysis Tools

State-of-the-art multiphysics simulations running on large scale leadership computing platforms have many variables contributing to their performance and scaling behavior. We recently encountered an interesting performance anomaly in Flash-X, a multiphysics multicomponent simulation software, when characterizing its performance behavior on several large-scale HPC platforms. The anomaly was tracked down to the interaction between the use of dynamic allocation of scratch data and data locality in the cache hierarchy. In this paper we present the details of unexpected performance variability of Flash-X, its extensive analysis using the performance measurement tool TAU to collect the data and Python data analysis libraries to explore the data, and our insights from this experience. In this process, we discovered and removed or mitigated two additional performance limiting bottlenecks for performance tuning.

Huck, Kevin↗

Locality-Aware CTA Clustering For Modern GPUs

In this paper, we proposed a novel clustering technique for tapping into the performance potential of a largely ignored type of locality: inter-CTA locality. We first demonstrated the capability of the existing GPU hardware to exploit such locality, both spatially and temporally, on L1 or L1/Tex unified cache. To verify the potential of this locality, we quantified its existence in a broad spectrum of applications and discussed its sources of origin. Based on these insights, we proposed the concept of CTA-Clustering and its associated software techniques. Finally, We evaluated these techniques on all modern generations of NVIDIA GPU architectures. The experimental results showed that our proposed clustering techniques could significantly improve on-chip cache performance.

Data Locality↗

Analyzing the suitability of contemporary 3D-stacked PIM architectures for HPC scientific applications

Scaling off-chip bandwidth is challenging due to fundamental limitations, such as a fixed pin count and plateauing signaling rates. Recently, vendors have turned to 2.5D and 3D stacking to closely integrate system components. Interestingly, these technologies can integrate a logic layer under multiple memory dies, enabling computing capability inside a memory stack. This trend in stacking is making PIM architectures commercially viable. In this work, we investigate the suitability of offloading kernels in scientific applications onto 3D stacked PIM architectures. We evaluate several hardware constraints resulted from the stacked structure. We perform extensive simulation experiments and in-depth analysis to quantify the impact of application locality in TLBs, data caches, and memory stacks. Our results also identify design optimization areas in software and hardware for HPC scientific applications.

Peng, Ivy Bo↗

Quantum Monte Carlo Endstation for Petascale Computing

The major achievements enabled by QMC Endstation grant include * Performance improvement on clusters of x86 multi-core systems, especially on Cray XT systems * New and improved methods for the wavefunction optimizations * New forms of trial wavefunctions * Implementation of the full application on NVIDIA GPUs using CUDA The scaling studies of QMCPACK on large-scale systems show excellent parallel efficiency up to 216K cores on Jaguarpf (Cray XT5). The GPU implementation shows speedups of 10-15x over the CPU implementation on older generation of x86. We have implemented hybrid OpenMP/MPI scheme in QMC to take advantage of multi-core shared memory processors of petascale systems. Our hybrid scheme has several advantages over the standard MPI-only scheme. * Memory optimized: large read-only data to store one-body orbitals and other shared properties to represent the trial wave function and many-body Hamiltonian can be shared among threads, which reduces the memory footprint of a large-scale problem. * Cache optimized: the data associated with an active Walker are in cache during the compute-intensive drift-diffusion process and the operations on an Walker are optimized for cache reuse. Thread-local objects are used to ensure the data affinity to a thread. * Load balanced: Walkers in an ensemble are evenly distributed among threads and MPI tasks. The two-level parallelism reduces the population imbalance among MPI tasks and reduces the number of point-to-point communications of large messages (serialized objects) for the Walker exchange. * Communication optimized: the communication overhead, especially for the collective operations necessary to determine ET and measure the properties of an ensemble, is significantly lowered by using less MPI tasks. The multiple forms of parallelism afforded by QMC algorithms make them ideal candidates for acceleration in the many-core paradigm. We presented the results of our effort to port the QMCPACK simulation code to the NVIDIA CUDA GPU platform. We restructured the CPU algorithms to express additional parallelism, minimize GPU-CPU communication, and efficiently utilize the GPU memory hierarchy. Using mixed precision on GT200 GPUs and MPI for intercommunication and load balancing, we observe typical full-application speedups of approximately 10x to 15x relative to quad-core Xeon CPUs alone, while reproducing the double-precision CPU results within statistical error. We developed an all-electron quantum Monte Carlo (QMC) method for solids that does not rely on pseudopotentials, and used it to construct a primary ultra-high-pressure calibration based on the equation of state of cubic boron nitride. We computed the static contribution to the free energy with the QMC method and obtained the phonon contribution from density functional theory, yielding a high-accuracy calibration up to 900 GPa usable directly in experiment. We computed the anharmonic Raman frequency shift with QMC simulations as a function of pressure and temperature, allowing optical pressure calibration. In contrast to present experimental approaches, small systematic errors in the theoretical EOS do not increase with pressure, and no extrapolation is needed. This all-electron method is applicable to first-row solids, providing a new reference for ab initio calculations of solids and benchmarks for pseudopotential accuracy. We compared experimental and theoretical results on the momentum distribution and the quasiparticle renormalization factor in sodium. From an x-ray Compton-profile measurement of the valence-electron momentum density, we derived its discontinuity at the Fermi wavevector finding an accurate measure of the renormalization factor that we compared with quantum-Monte-Carlo and G0W0 calculations performed both on crystalline sodium and on the homogeneous electron gas. Our calculated results are in good agreement with the experiment. We have been studying the heat of formation for various Kubas complexes of molecular hydrogen on Ti(1,2)ethylene-nH2 using Diffusion Monte Carlo. This work has been started and is ongoing. We are studying systems involving 1 and 2 Ti bonding sites with up to 10 hydrogen molecules in numerous configurations. This work will establish a benchmark that will test the accuracy of density functional calculations and establish the feasibility of our methods for similar systems.

97 MATHEMATICS AND COMPUTING↗

FitCache: A Transparent Drop-In Framework for Multi-Tier Caching to Accelerate Distributed Deep Learning Workloads

Training in Deep learning (DL) remains highly compute- and data-intensive, with I/O becoming a critical bottleneck as models and datasets scale. Recent studies report that data loading can dominate training time, especially on large-scale HPC systems with shared parallel file systems (PFS). Existing caching approaches either rely on single-tier designs or require intrusive modifications to training pipelines, limiting their portability and effectiveness. In this work, we present FitCache, a transparent drop-in framework for multi-tier caching to accelerate distributed DL training by coordinating fast local memory (e.g., DRAM, Persistent Memory (PMem)) and NVMe as hierarchical caches atop PFS. Our design adapts to hardware diversity, i.e., if NVMe is missing, memory transparently acts as a caching tier, ensuring stable performance. FitCache transparently intercepts I/O requests and issues concurrent fetches across all tiers, returning data from the fastest responder without centralized metadata or static redirection paths. FitCache adapts to dynamic workloads and heterogeneous clusters while maintaining POSIX compatibility. Experiments on Frontier (2048 GPUs) and smaller research clusters show that FitCache reduces training time by up to 40% and per-batch I/O latency by up to 71.6% compared to Lustre Orion PFS, offering a drop-in solution for scalable DL training.

Hu, Guangxing [ORNL] (ORCID:0009000283203614)↗

Efficient I/O for Neural Network Training with Compressed Data

FanStore is a shared object store that enables efficient and scalable neural network training on supercomputers. By providing a global cache layer on node-local burst buffers using a compressed representation, it significantly enhances the processing capability of deep learning (DL) applications on existing hardware. In addition, FanStore allows POSIX-compliant file access to the compressed data in user space. We investigate the tradeoff between runtime overhead and data compression ratio using real-world datasets and applications, and propose a compressor selection algorithm to maximize storage capacity given performance constraints. We consider both asynchronous (i.e., with prefetching) and synchronous I/O strategies, and propose mechanisms for selecting compressors for both approaches. Using FanStore, the same storage hardware can host 2-13x more data for example applications without significant runtime overhead. Empirically, our experiments show that FanStore scales to 512 compute nodes with near linear performance scalability.

Zhang, Zhao↗

Shifting Between Compute and Memory Bounds: A Compression-Enabled Roofline Model

In the evolving landscape of high-performance computing, especially to fight the end of Moore’s Law and Dennard’s Scaling, the ability to shift between compute-bound and memory-bound states is critical for enhancing adaptability and flexibility to diverse system and domain-specific architectures. Such capability is vital for optimizing performance across distinguished hardware configurations, such as accelerators, memory hierarchies, and cache systems. Despite that ad hoc optimization techniques, such as compressed/approximate computation, have been enabled for compute-/data-intensive computing for improved performance in distinct hardware settings, there lacks an understanding of 1) the rational behind performance improvement; 2) capability of different optimizations; 3) what optimization to respond to specific computational and memory demands. This work proposes a compression-enabled roofline model to facilitate this adaptability with data compression techniques to balance and transform between computational and memory demands. This model enables applications to adjust in response to the specific strengths and limitations of the underlying hardware and system to optimize resource utilization. The effectiveness of this approach is demonstrated with matrix multiplication kernels on different input sizes, with turning on/off various compression techniques, including 1) low-precision floating point; 2) sparse matrix formulation; and 3) compressed arrays with ZFP. By reducing memory transfer volumes and cache misses and increasing data locality and computational intensity through compression, the specific roofline model can transform between compute and memory bounds to align more efficiently with system capabilities. This advancement not only improves overall performance but also maximizes adaptability in diverse computing environments.

Naraparaju, Ramasoumya [University of Washington]↗

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↗

Prioritizing local and remote memory access in a non-uniform memory access architecture

A miss in a cache by a thread in a wavefront is detected. The wavefront includes a plurality of threads that are executing a memory access request concurrently on a corresponding plurality of processor cores. A priority is assigned to the thread based on whether the memory access request is addressed to a local memory or a remote memory. The memory access request for the thread is performed based on the priority. In some cases, the cache is selectively bypassed depending on whether the memory access request is addressed to the local or remote memory. A cache block is requested in response to the miss. The cache block is biased towards a least recently used position in response to requesting the cache block from the local memory and towards a most recently used position in response to requesting the cache block from the remote memory.

Boyer, Michael W.↗

MAC: Memory Access Coalescer for 3D-Stacked Memory

Emerging data-intensive applications, such as graph analytics and data mining, exhibit irregular memory access patterns. Research has shown that with these memory-bound applications, traditional cache-based processor architectures, which exploit locality and regular patterns to mitigate the memory-wall issue, are inefficient. Meantime, novel 3D-stacked memory devices, such as Hybrid Mem- ory Cube (HMC) and High Bandwidth Memory (HBM), promise significant increases in bandwidth that appear extremely appealing for memory-bound applications. However, conventional memory interfaces designed for cache-based architectures and JEDEC DDR devices fit poorly with the 3D-stacked memory, which leads to significant under-utilization of the promised high bandwidth. As a response to these issues, in this paper we propose MAC (Memory Access Coalescer), a coalescing unit for the 3D-stacked memory. We discuss the design and implementation of MAC, in the context of a custom designed cache-less architecture targeted at data-intensive, irregular applications. Through a custom simulation infrastructure based on the RISC-V toolchain, we show that MAC achieves a coalescing efficiency of 52.85% on average. It improves the performance of the memory system by 60.73% on average for a large set of irregular workloads.

Wang, Xi↗

Local rollback for fault-tolerance in parallel computing systems

A control logic device performs a local rollback in a parallel super computing system. The super computing system includes at least one cache memory device. The control logic device determines a local rollback interval. The control logic device runs at least one instruction in the local rollback interval. The control logic device evaluates whether an unrecoverable condition occurs while running the at least one instruction during the local rollback interval. The control logic device checks whether an error occurs during the local rollback. The control logic device restarts the local rollback interval if the error occurs and the unrecoverable condition does not occur during the local rollback interval.

97 MATHEMATICS AND COMPUTING↗

Tag-Split Cache for Efficient GPGPU Cache Utilization

Modern GPUs employ cache to improve memory system efficiency. However, large amount of cache space is underutilized due to irregular memory accesses and poor spatial locality which exhibited commonly in GPU applications. Our experiments show that using smaller cache lines could improve cache space utilization, but it also frequently suffers from significant performance loss by introducing large amount of extra cache requests. In this work, we propose a novel cache design named tag-split cache (TSC) that enables fine-grained cache storage to address the problem of cache space underutilization while keeping memory request number unchanged. TSC divides tag into two parts to reduce storage overhead, and it supports multiple cache line replacement in one cycle.

performance and energy optimization↗

Accelerating Application Bulk Synchronous Writes in HPC Environments

High-bandwidth storage tiers are becoming more common for their capability to absorb high-rate, bursty I/Os. Notably, the designs of these fast storage tiers differ from system to system. The variation of these layers and non-uniform methods of access can pose chal- lenges for applications seeking to run at multiple HPC facilities. Therefore, in this work, we present Spectral, a rapid-output ab- straction library to accelerate application, bulk-synchronous writes on HPC systems. We design Spectral to enable applications to use high-bandwidth storage, such as node-local storage and dis- tributed, write-caches (e.g., burst buffers) transparently without requiring modifications to the application or file system source code. The key idea is to allow applications to spend most of the time performing productive work and to not require any source code changes for maximum portability on different HPC archi- tectures. Spectral internally re-routes write-only files through available, high-performance I/O resources before ultimately mi- grating them to the shared global parallel file system. For instance, on Summit, Spectral transparently places application outputs on node-local storage and then utilizes asynchronous migration to the center-wide GPFS file system. We evaluate Spectral on the Summit HPC system (1024 nodes) using the IOR benchmark and real scientific applications. Spectral shows linear performance scaling, improving application write performance by over an order of magnitude when compared to GPFS.

Khan, Awais↗