Search NASA⌕ Search

SEARCH · Search NASA

Results for “parallel computer architecture”

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 1,603 records · Page 89

Fixing Amdahl's Law within the Limits of Accelerated Systems: FALLACY

Closeout report for FALLACY project. The performance of Data Model Convergence Initiative (DMC) applications on parallel machines is far below the limit set by Amdahl’s law. Whether the machine is based on many-core, GPUs, FPGAs, or a heterogeneous combination, usually the most significant bottleneck is accessing data from the memory system. Aligning with DMC’s HW/architecture thrust, this project developed a set of memory-centric tools called ‘MemGaze’ that inform the HW/SW stack about an application’s memory behavior, including data access latency and diagnosing poor data layout and data composition. Our approach uses architectural modeling and analysis of workload data accesses.

97 MATHEMATICS AND COMPUTING↗

Novel Solver Algorithms for Nearly Singular Linear Systems Arising in Combustion Modelling

Direct Numerical Simulations of realistic combustion devices are extremely challenging due to the wide separation of scales in the simulation, for example an internal combustion (IC) engine chamber, and the flame thickness of a high-pressure flame. The PeleLMeX solver uses adaptive mesh refinement (AMR) to evolve multi-species reacting flows in the low Mach number limit at the Exascale and relies on an embedded boundary (EB) approach to represent complex geometries. In that framework, the EB geometries often give rise to very small cut-cells along the boundary, which translate into extreme ill-conditioning of the pressure-projection, with eigenvalues that span 15-16 orders of magnitude. In this talk, we focus on the case of a typical IC piston bowl geometry for which we present on a novel approach towards solving these nearly singular linear systems with ILU-based, C-AMG smoothers on massively parallel architectures. In particular, we use scaling and equilibration algorithms to handle the non-normality of the upper triangular factors. This enables us to approximate the highly sequential triangular solve algorithm, embedded in the AMG smoothing-solve phase, with Jacobi iterations. This approximation can be written as a convergent Neumann series whose terms are composed of highly parallel sparse matrix vector multiplications. The result is an algorithm that substantially decreases setup and solve time, compared to state-of-the-art, for these challenging linear systems.

combustion modelling↗

Developing A Large Time Step, Robust, and Low Communication Multi-Moment PDE Integration Scheme for Exascale Applications

The Boundary Averaged Multi-moment Constrained finite-Volume (BA-MCV) method is de- rived, explained, and evaluated for 1-D transport to assess accuracy, maximum stable time step (MSTS), oscillations for discontinuous data, and parallel communication burden. The BA-MCV scheme is altered from the original MCV scheme to compute the updates of point wise cell boundary derivatives entirely locally. Then it is altered such that boundary moments are replaced with the interface upwind value. The scheme is stable at a maximum stable CFL (MSCFL) value of one no matter how high-order the scheme is, giving significantly larger time steps than Galerkin methods, for which the MSCFL decreases nearly quadratically with in- creasing order. The BA-MCV method is compared against a SE method at varying order, both using the ADER-DT time discretization. BA-MCV error for a sine wave was comparable to the same order of accuracy for a SE method. The resulting large time step, multi-moment, low communication scheme is of great interest for exascale architectures.

Norman, Matthew R.↗

Surface code architecture for donors and dots in silicon with imprecise and nonuniform qubit couplings

A scaled quantum computer with donor spins in silicon would benefit from a viable semiconductor framework and a strong inherent decoupling of the qubits from the noisy environment. Coupling neighboring spins via the natural exchange interaction according to current designs requires gate control structures with extremely small length scales. In this work, we present a silicon architecture where bismuth donors with long coherence times are coupled to electrons that can shuttle between adjacent quantum dots, thus relaxing the pitch requirements and allowing space between donors for classical control devices. An adiabatic SWAP operation within each donor/dot pair solves the scalability issues intrinsic to exchange-based two-qubit gates, as it does not rely on subnanometer precision in donor placement and is robust against noise in the control fields. In conclusion, we use this SWAP together with well established global microwave Rabi pulses and parallel electron shuttling to construct a surface code that needs minimal, feasible local control.

36 MATERIALS SCIENCE↗

Pretraining Billion-Scale Geospatial Foundational Models on Frontier

As AI workloads increase in scope, generalization capability becomes challenging for small task-specific models and their demand for large amounts of labeled training samples increases. On the contrary, Foundation Models (FMs) are trained with internet-scale unlabeled data via self-supervised learning and have been shown to adapt to various tasks with minimal fine-tuning. Although large FMs have demonstrated significant impact in natural language processing and computer vision, efforts toward FMs for geospatial applications have been restricted to smaller size models, as pretraining larger models requires very large computing resources equipped with state-of-the-art hardware accelerators. Current satellite constellations collect 100+TBs of data a day, resulting in images that are billions of pixels and multimodal in nature. Such geospatial data poses unique challenges opening up new opportunities to develop FMs. We investigate billion scale FMs and HPC training profiles for geospatial applications by pretraining on publicly available data. We studied from end-to-end the performance and impact in the solution by scaling the model size. Our larger 3B parameter size model achieves up to 30% improvement in top1 scene classification accuracy when comparing a 100M parameter model. Moreover, we detail performance experiments on the Frontier supercomputer, America's first exascale system, where we study different model and data parallel approaches using PyTorch's Fully Sharded Data Parallel library. Specifically, we study variants of the Vision Transformer architecture (ViT), conducting performance analysis for ViT models with size up to 15B parameters. By discussing throughput and performance bottlenecks under different parallelism configurations, we offer insights on how to leverage such leadership-class HPC resources when developing large models for geospatial imagery applications.

Tsaris, Aristeidis (aris)↗

HARMONY: Large-Scale Architecture Search for Efficient Hybrid Language Models

As large language models scale to trillions of parameters, their computational and memory requirements present critical challenges for efficient training and deployment. While Mixture of Experts (MoE) architectures enable efficient scaling through sparse parameter activation, and state-space models like Mamba offer linear-time complexity, principled methods for combining these paradigms remain undeveloped. We introduce HARMONY (Hybrid Architecture Research for Mamba, Optimized with Neural efficiencY), a multi-objective evolutionary neural architecture search framework for discovering efficient hybrid language models that integrate Transformer attention mechanisms, Mixture-of-Experts routing, and Mamba state-space components. Through large-scale distributed search using 16,384 MI250X GPUs on the Frontier supercomputer, HARMONY explores a comprehensive design space encompassing six attention variants (MHA, MQA, GQA, MLA, SWA, and Mamba-2), variable MoE configurations with both routed and shared experts, and extensive Mamba hyperparameters. Our framework discovers heterogeneous architectures that balance training performance with computational efficiency through multi-objective optimization incorporating latency penalties and fitness-based selection. Analysis of discovered architectures reveals that optimal hybrid designs favor heterogeneous component mixing rather than homogeneous patterns, with Mamba-2 and Multi-Head Latent Attention (MLA) emerging as preferred mechanisms. Discovered architectures demonstrate superior training efficiency: our best configuration achieves a final perplexity of 1.0874 with 2.38B parameters while processing 4,320 tokens/second, outperforming significantly larger manually designed models. Full-scale evaluation shows HARMONY's top architectures achieve better loss trajectories than equivalently-sized models using state-of-the-art configurations including Mixtral, Jamba, and Samba. Additionally, we demonstrate 91% weak scaling efficiency when training discovered 36B-parameter models across 1,024 GPUs. HARMONY is released as an open framework with comprehensive tools for building and training hybrid models using expert-data-pipeline parallelism, democratizing access to automated architecture design for next-generation language models.

Herron, Emily [ORNL] (ORCID:0000000273008172)↗

Practical Aerodynamic Design Optimization Based on the Navier-Stokes Equations and a Discrete Adjoint Method

Compressible and incompressible versions of a three-dimensional unstructured mesh Reynolds-averaged Navier-Stokes flow solver have been differentiated and resulting derivatives have been verified by comparisons with finite differences and a complex-variable approach. In this implementation, the turbulence model is fully coupled with the flow equations in order to achieve this consistency. The accuracy demonstrated in the current work represents the first time that such an approach has been successfully implemented. The accuracy of a number of simplifying approximations to the linearizations of the residual have been examined. A first-order approximation to the dependent variables in both the adjoint and design equations has been investigated. The effects of a "frozen" eddy viscosity and the ramifications of neglecting some mesh sensitivity terms were also examined. It has been found that none of the approximations yielded derivatives of acceptable accuracy and were often of incorrect sign. However, numerical experiments indicate that an incomplete convergence of the adjoint system often yield sufficiently accurate derivatives, thereby significantly lowering the time required for computing sensitivity information. The convergence rate of the adjoint solver relative to the flow solver has been examined. Inviscid adjoint solutions typically require one to four times the cost of a flow solution, while for turbulent adjoint computations, this ratio can reach as high as eight to ten. Numerical experiments have shown that the adjoint solver can stall before converging the solution to machine accuracy, particularly for viscous cases. A possible remedy for this phenomenon would be to include the complete higher-order linearization in the preconditioning step, or to employ a simple form of mesh sequencing to obtain better approximations to the solution through the use of coarser meshes. An efficient surface parameterization based on a free-form deformation technique has been utilized and the resulting codes have been integrated with an optimization package. Lastly, sample optimizations have been shown for inviscid and turbulent flow over an ONERA M6 wing. Drag reductions have been demonstrated by reducing shock strengths across the span of the wing. In order for large scale optimization to become routine, the benefits of parallel architectures should be exploited. Although the flow solver has been parallelized using compiler directives. The parallel efficiency is under 50 percent. Clearly, parallel versions of the codes will have an immediate impact on the ability to design realistic configurations on fine meshes, and this effort is currently underway.

Grossman, Bernard↗

Nebo: An efficient, parallel, and portable domain-specific language for numerically solving partial differential equations

This study presents Nebo, a declarative domain-specific language embedded in C++ for discretizing partial differential equations for transport phenomena on multiple architectures. Application programmers use Nebo to write code that appears sequential but can be run in parallel, without editing the code. Currently Nebo supports single-thread execution, multi-thread execution, and many-core (GPU-based) execution. With single-thread execution, Nebo performs on par with code written by domain experts. With multi-thread execution, Nebo can linearly scale (with roughly 90% efficiency) up to 12 cores, compared to its single-thread execution. Moreover, Nebo’s many-core execution can be over 140x faster than its single-thread execution.

97 MATHEMATICS AND COMPUTING↗

The Astronomy Commons Platform: A Deployable Cloud-based Analysis Platform for Astronomy

Abstract We present a scalable, cloud-based science platform solution designed to enable next-to-the-data analyses of terabyte-scale astronomical tabular data sets. The presented platform is built on Amazon Web Services (over Kubernetes and S3 abstraction layers), utilizes Apache Spark and the Astronomy eXtensions for Spark for parallel data analysis and manipulation, and provides the familiar JupyterHub web-accessible front end for user access. We outline the architecture of the analysis platform, provide implementation details and rationale for (and against) technology choices, verify scalability through strong and weak scaling tests, and demonstrate usability through an example science analysis of data from the Zwicky Transient Facility’s 1Bn+ light-curve catalog. Furthermore, we show how this system enables an end user to iteratively build analyses (in Python) that transparently scale processing with no need for end-user interaction. The system is designed to be deployable by astronomers with moderate cloud engineering knowledge, or (ideally) IT groups. Over the past 3 yr, it has been utilized to build science platforms for the DiRAC Institute, the ZTF partnership, the LSST Solar System Science Collaboration, and the LSST Interdisciplinary Network for Collaboration and Computing, as well as for numerous short-term events (with over 100 simultaneous users). In a live demo instance, the deployment scripts, source code, and cost calculators are accessible. 4 4 http://hub.astronomycommons.org/

79 ASTRONOMY AND ASTROPHYSICS↗

Knowledge-based vision for space station object motion detection, recognition, and tracking

Computer vision, especially color image analysis and understanding, has much to offer in the area of the automation of Space Station tasks such as construction, satellite servicing, rendezvous and proximity operations, inspection, experiment monitoring, data management and training. Knowledge-based techniques improve the performance of vision algorithms for unstructured environments because of their ability to deal with imprecise a priori information or inaccurately estimated feature data and still produce useful results. Conventional techniques using statistical and purely model-based approaches lack flexibility in dealing with the variabilities anticipated in the unstructured viewing environment of space. Algorithms developed under NASA sponsorship for Space Station applications to demonstrate the value of a hypothesized architecture for a Video Image Processor (VIP) are presented. Approaches to the enhancement of the performance of these algorithms with knowledge-based techniques and the potential for deployment of highly-parallel multi-processor systems for these algorithms are discussed.

Symosek, P.↗

Integrating Cache Performance Modeling and Tuning Support in Parallelization Tools

With the resurgence of distributed shared memory (DSM) systems based on cache-coherent Non Uniform Memory Access (ccNUMA) architectures and increasing disparity between memory and processors speeds, data locality overheads are becoming the greatest bottlenecks in the way of realizing potential high performance of these systems. While parallelization tools and compilers facilitate the users in porting their sequential applications to a DSM system, a lot of time and effort is needed to tune the memory performance of these applications to achieve reasonable speedup. In this paper, we show that integrating cache performance modeling and tuning support within a parallelization environment can alleviate this problem. The Cache Performance Modeling and Prediction Tool (CPMP), employs trace-driven simulation techniques without the overhead of generating and managing detailed address traces. CPMP predicts the cache performance impact of source code level "what-if" modifications in a program to assist a user in the tuning process. CPMP is built on top of a customized version of the Computer Aided Parallelization Tools (CAPTools) environment. Finally, we demonstrate how CPMP can be applied to tune a real Computational Fluid Dynamics (CFD) application.

Waheed, Abdul↗

The high level trigger and express data production at STAR

To meet the demands of the Beam Energy Scan phase-II (BES-II) program, the STAR experiment at the Relativistic Heavy Ion Collider (RHIC) developed a dual real-time framework consisting of a High Level Trigger (HLT) and an Express Data Production system (xProduction). The HLT operates online within the Data Acquisition (DAQ) chain on a dedicated multi-core CPU cluster with the option to offload compute-intensive kernels to Xeon Phi coprocessors. It uses parallelized algorithms, such as the Cellular Automaton (CA) Track Finder, to perform rapid tracking, vertexing, and event filtering. This allows it to select events of interest in real time and provide immediate feedback on detector and beam conditions. In contrast, the xProduction workflow runs concurrently and independently of the DAQ loop. It applies near offline-quality calibration and reconstruction within hours of data collection. The xProduction input is the express data stream, whose content can be enriched by HLT trigger/priority selections under DAQ/HLT resource constraints, and it uses the STAR calibration/conditions framework, incorporating online calibration/QA information when available. This enables early preliminary physics analysis, including the reconstruction of rare signals, such as hyperons and hypernuclei. It also provides collaboration-wide access to analysis-ready datasets. Together, the HLT and xProduction systems form a complementary architecture: the HLT performs online event selection while the xProduction chain delivers high-quality results within a short amount of time. This integrated framework has enabled the prompt reconstruction of the $^5_Λ$ He hypernucleus with high statistical significance and the efficient processing of hundreds of millions of heavy-ion collision events. In conclusion, its demonstrated scalability and robustness establish a model for future high-luminosity experiments requiring both online event filtering and rapid access to analysis-quality data.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

An operating system for future aerospace vehicle computer systems

The requirements for future aerospace vehicle computer operating systems are examined in this paper. The computer architecture is assumed to be distributed with a local area network connecting the nodes. Each node is assumed to provide a specific functionality. The network provides for communication so that the overall tasks of the vehicle are accomplished. The O/S structure is based upon the concept of objects. The mechanisms for integrating node unique objects with node common objects in order to implement both the autonomy and the cooperation between nodes is developed. The requirements for time critical performance and reliability and recovery are discussed. Time critical performance impacts all parts of the distributed operating system; e.g., its structure, the functional design of its objects, the language structure, etc. Throughout the paper the tradeoffs - concurrency, language structure, object recovery, binding, file structure, communication protocol, programmer freedom, etc. - are considered to arrive at a feasible, maximum performance design. Reliability of the network system is considered. A parallel multipath bus structure is proposed for the control of delivery time for time critical messages. The architecture also supports immediate recovery for the time critical message system after a communication failure.

Foudriat, E. C.↗

Concurrent Collections (CnC): A new approach to parallel programming

A common approach in designing parallel languages is to provide some high level handles to manipulate the use of the parallel platform. This exposes some aspects of the target platform, for example, shared vs. distributed memory. It may expose some but not all types of parallelism, for example, data parallelism but not task parallelism. This approach must find a balance between the desire to provide a simple view for the domain expert and provide sufficient power for tuning. This is hard for any given architecture and harder if the language is to apply to a range of architectures. Either simplicity or power is lost. Instead of viewing the language design problem as one of providing the programmer with high level handles, we view the problem as one of designing an interface. On one side of this interface is the programmer (domain expert) who knows the application but needs no knowledge of any aspects of the platform. On the other side of the interface is the performance expert (programmer or program) who demands maximal flexibility for optimizing the mapping to a wide range of target platforms (parallel / serial, shared / distributed, homogeneous / heterogeneous, etc.) but needs no knowledge of the domain. Concurrent Collections (CnC) is based on this separation of concerns. The talk will present CnC and its benefits. About the speaker. Kathleen Knobe has focused throughout her career on parallelism especially compiler technology, runtime system design and language design. She worked at Compass (aka Massachusetts Computer Associates) from 1980 to 1991 designing compilers for a wide range of parallel platforms for Thinking Machines, MasPar, Alliant, Numerix, and several government projects. In 1991 she decided to finish her education. After graduating from MIT in 1997, she joined Digital Equipment’s Cambridge Research Lab (CRL). She stayed through the DEC/Compaq/HP mergers and when CRL was acquired and absorbed by Intel. She currently works in the Software and Services Group / Technology Pathfinding and Innovation.

97 MATHEMATICS AND COMPUTING↗

Direct 400 Vdc to 1 Vdc power conversion with input series output parallel connection for data center power supplies

Data centers consume an ever-increasing amount of electricity because of the rapid growth of cloud computing and digital information storage. A high voltage point of load (HV POL) converter is proposed to convert the 400-VDC distribution voltage to 1-VDC within a single stage to increase the power conversion efficiency. A six-phase input series output parallel (ISOP) connected structure is implemented for the HV POL. The symmetrical controlled half bridge current doubler is selected as the converter topology in the ISOP structure. The full load efficiency is improved by 4% points compared with state of the art products. A voltage compensator has been designed in order to meet the strict dynamic voltage regulation requirement. A laboratory prototype has been built, and experimental results have been provided to verify the proposed HV POL with a single power conversion stage can meet the dynamic voltage regulation requirement for an on-board power supply with higher efficiency compared to the conventional architecture.

Cui, Yutian↗

Unified Models of Turbulence and Nonlinear Wave Evolution in the Extended Solar Corona and Solar Wind

The PI (Cranmer) and Co-I (A. van Ballegooijen) made significant progress toward the goal of building a "unified model" of the dominant physical processes responsible for the acceleration of the solar wind. The approach outlined in the original proposal comprised two complementary pieces: (1) to further investigate individual physical processes under realistic coronal and solar wind conditions, and (2) to extract the dominant physical effects from simulations and apply them to a one-dimensional and time-independent model of plasma heating and acceleration. The accomplishments in the report period are thus divided into these two categories: 1a. Focused Study of Kinetic MHD Turbulence. We have developed a model of magnetohydrodynamic (MHD) turbulence in the extended solar corona that contains the effects of collisionless dissipation and anisotropic particle heating. A turbulent cascade is one possible way of generating small-scale fluctuations (easy to dissipate/heat) from a pre-existing population of low-frequency Alfven waves (difficult to dissipate/heat). We modeled the cascade as a combination of advection and diffusion in wavenumber space. The dominant spectral transfer occurs in the direction perpendicular to the background magnetic field. As expected from earlier models, this leads to a highly anisotropic fluctuation spectrum with a rapidly decaying tail in the parallel wavenumber direction. The wave power that decays to high enough frequencies to become ion cyclotron resonant depends on the relative strengths of advection and diffusion in the cascade. For the most realistic values of these parameters, though, there is insufficient power to heat protons and heavy ions. The dominant oblique waves undergo Landau damping, which implies strong parallel electron heating. We thus investigated the nonlinear evolution of the electron velocity distributions (VDFs) into parallel beams and discrete phase-space holes (similar to those seen in the terrestrial magnetosphere) which are an alternate means of heating protons via stochastic interactions similar to particle-particle collisions. 1b. Focused Study of the Multi-Mode Detailed Balance Formalism. The PI began to explore the feasibility of using the "weak turbulence," or detailed-balance theory of Tsytovich, Melrose, and others to encompass the relevant physics of the solar wind. This study did not go far, however, because if the "strong" MHD turbulence discussed above is a dominant player in the wind's acceleration region, this formalism is inherently not applicable to the corona. We will continue to study the various published approaches to the weak turbulence formalism, especially with an eye on ways to parameterize nonlinear wave reflection rates. 2. Building the Unified Model Code Architecture. We have begun developing the computational model of a time-steady open flux tube in the extended corona. The model will be "unified" in the sense that it will include (simultaneously for the first time) as many of the various proposed physical processes as possible, all on equal footing. To retain this generality, we have formulated the problem in two interconnected parts: a completely kinetic model for the particles, using the Monte Carlo approach, and a finite-difference approach for the self-consistent fluctuation spectra. The two codes are run sequentially and iteratively until complete consistency is achieved. The current version of the Monte Carlo code incorporates gravity, the zero-current electric field, magnetic mirroring, and collisions. The fluctuation code incorporates WKJ3 wave action conservation and the cascade/dissipation processes discussed above. The codes are being run for various test problems with known solutions. Planned additions to the codes include prescriptions for nonlinear wave steepening, kinetic velocity-space diffusion, and multi-mode coupling (including reflection and refraction).

Cranmer, Steven R.↗

Data Locality Enhancement of Dynamic Simulations for Exascale Computing (Final Report)

The development of modern processors exhibits two trends that complicate the optimizations of modern software. The first is the increasing sensitivity of processors' throughput to irregularities in computation. With more processors produced through a massive integration of simple cores, future systems will increasingly favor regular data-level parallel computations, but deviate from the needs of applications with complex patterns. Some evidences are already shown on Graphic Processing Units (GPU): Irregular data accesses (e.g., indirect references A[D[i]]) and conditional branches are limiting many GPU applications' performance at a level an order of magnitude lower than the peak of GPU. The second hardware trend is the growing gap between memory bandwidth and the aggregate speed—that is, the sum of all cores' computing power—of a Chip Multiprocessor (CMP). Despite the capped growth of the peak CPU speed, the aggregate speed of a CMP keeps increasing as more cores get into a single chip. It is expected that by 2018, node concurrency in an exascale system will increase by hundreds of times, whereas, memory bandwidth will expand by only 10 to 20 times. Consequently, data movement and storage is expected to consume more than 70% of the total system power. Bridging this gap is difficult; the complexities of modern CMP memory hierarchy make it even harder: Data cache becomes shared among computing units, and the sharing is often non-uniform—whether two computing units share a cache depends on their proximity and the level of the cache. On the recent IBM Power7 architecture, for instance, four hardware contexts (or SMT threads) in a core share the entire memory hierarchy, all cores in one chip share an on-chip L3 cache, and cores across chips share L3 and main memory through off-chip connections. These two trends complicate the translation of computing power into performance, especially for a program with either intensive data accesses or complex patterns in data accesses or control flow paths. Unfortunately, both attributes present and will persist in a class of important applications. For instance, many scientific simulations deal with a large volume of data. And meanwhile, as most real-world processes are non-uniform and evolving (e.g., the evolution of a galaxy or the process of a drug injection), both the computations and data accesses of these programs tend to be irregular and dynamically changing. Currently, the lack of support to these applications on modern CMP severely limits their performance. On GPU, as our recent study shows and other studies echo, performance enhancement of a factor of integers is possible when memory accesses or control flows are streamlined for a set of GPU applications. On multicore CPU, our studies show that traditional locality enhancement, for being oblivious to the new features of multicore memory hierarchy, may even cause large slowdown to data-intensive dynamic applications. The severity of the issues is expected to worsen as the two hardware trends continue. Some recent studies try to match software with the trends, but in a limited scope or manner. For irregularities on GPU, most studies focus on irregularities analyzable through static analysis (e.g., data accesses in regular loops). Dynamic irregularities are harder to address because the needed analysis and transformations typically have to happen at run time. Some other research resorts to hardware extensions, an actual adoption of which is unclear for the entailed space cost and complexity. For data locality, recent years have seen some exploitations of the new memory hierarchy on multicore for performance, but most of them are on process or thread scheduling, rather than program transformations. Our recent study reveals that program-level transformations may magnify the scheduling benefits by a factor of seven, concluding that program-level transformation should play a central role for data locality enhancement on modern CMP. But research in this direction has been sparse, and most have focused on data layout or cache performance modeling, rather than program transformations to match with the new memory hierarchy features. Overall, it is still an open question how to bridge the gap between dynamic computations and the two prominent properties of modern processors. The goal of this project is to develop a set of techniques and software tools to enhance the matching between memory accesses in dynamic simulations and the prominent features of modern and future CMP, alleviating the memory performance issues for petascale and exascale computing. This report summarizes the discoveries and products produced throughout this project. It includes free launch, a new software approach to overcoming the shortcomings of both methods; coherence-free multiview, an approach that allows multiple views of a single data object to co-exist on GPU memory during a GPU kernel execution; algorithmic optimizations to data analytics problems, especially those that involve lots of distance calculations. Supporting dynamic parallelism is important for GPU to benefit a broad range of applications. There are currently two fundamental ways for programs to exploit dynamic parallelism on GPU: a software-based approach with software-managed worklists, and a hardware-based approach through dynamic subkernel launches. Neither is satisfactory. The former is complicated to program and is often subject to some load imbalance; the latter suffers large runtime overhead. In this work, we propose free launch, a new software approach to overcoming the shortcomings of both methods. It allows programmers to use subkernel launches to express dynamic parallelism. It employs a novel compiler-based code transformation named subkernel launch removal to replace the subkernel launches with the reuse of parent threads. Coupled with an adaptive task assignment mechanism, the transformation reassigns the tasks in the subkernels to the parent threads with a good load balance. The technique requires no hardware extensions, immediately deployable on existing GPUs. It keeps the programming convenience of the subkernel launch-based approach while avoiding its large runtime overhead. Meanwhile, its superior load balancing makes it outperform manual worklist-based techniques by 3X on average. The work was published at the 48th Annual IEEE/ACM International Symposium on Microarchitecture (Micro'2015). A Graphic Processing Unit (GPU) system is typically equipped with many types of memory (e.g., global, constant, texture, shared, cache). Data placement determines what data are placed on which type of memory, essential for GPU memory performance. Prior optimizations of data placement always require a single view of a data object on memory, which limits the optimization effectiveness. In this work, we propose coherence-free multiview, an approach that allows multiple views of a single data object to co-exist on GPU memory during a GPU kernel execution. We demonstrate that under certain conditions, the multiple views can remain incoherent while facilitating enhanced data placement. We present a theorem and some compiler support to ensure the soundness of the usage of coherence-free multiview. We further develop reference-discerning data placement, a new way to enhance data placements on GPU. It enables more flexible data placements by using coherence-free multiview to leverage the slack in coherence requirement of some GPU programs. Experiments on three types of GPU systems show that, with less than 200KB space cost, the new data placement technique can provide a 1.6X average (up to 4.27X) speedup. In addition, we have examined algorithmic optimizations to data analytics problems, especially those that involve lots of distance calculations. Computing distances among data points is an essential part of many important algorithms in data analytics, graph analysis, and other domains. In each of these domains, developers have spent significant manual effort optimizing algorithms, often through novel applications of the triangle equality, in order to minimize the number of distance computations in the algorithms. In this work, we observe that many algorithms across these domains can be generalized as an instance of a generic distance-related abstraction. Based on this abstraction, we derive seven principles for correctly applying the triangular inequality to optimize distance-related algorithms. Guided by the findings, we develop Triangular OPtimizer (TOP), the first software framework that is able to automatically produce optimized algorithms that either matches or outperforms manually designed algorithms for solving distance-related problems. TOP achieves up to 237x speedups and 2.5X on average. The work has been published at the 32nd International Conference on Machine Learning (ICML'15) and the 41st International Conference on Very Large Data Bases (VLDB'15).

42 ENGINEERING↗

FastQuery: A Parallel Indexing System for Scientific Data

Modern scientific datasets present numerous data management and analysis challenges. State-of-the- art index and query technologies such as FastBit can significantly improve accesses to these datasets by augmenting the user data with indexes and other secondary information. However, a challenge is that the indexes assume the relational data model but the scientific data generally follows the array data model. To match the two data models, we design a generic mapping mechanism and implement an efficient input and output interface for reading and writing the data and their corresponding indexes. To take advantage of the emerging many-core architectures, we also develop a parallel strategy for indexing using threading technology. This approach complements our on-going MPI-based parallelization efforts. We demonstrate the flexibility of our software by applying it to two of the most commonly used scientific data formats, HDF5 and NetCDF. We present two case studies using data from a particle accelerator model and a global climate model. We also conducted a detailed performance study using these scientific datasets. The results show that FastQuery speeds up the query time by a factor of 2.5x to 50x, and it reduces the indexing time by a factor of 16 on 24 cores.

96 KNOWLEDGE MANAGEMENT AND PRESERVATION↗