Astra, deployed in 2018, was the first petascale supercomputer to utilize processors based on the ARM instruction set. The system was also the first under Sandia's Vanguard program which seeks to provide an evaluation vehicle for novel technologies that with refinement could be utilized in demanding, large-scale HPC environments. In addition to ARM, several other important first-of-a-kind developments were used in the machine, including new approaches to cooling the datacenter and machine. This article documents our experiences building a power measurement and control infrastructure for Astra. While this is often beyond the control of users today, the accurate measurement, cataloging, and evaluation of power, as our experiences show, is critical to the successful deployment of a large-scale platform. While such systems exist in part for other architectures, Astra required new development to support the novel Marvell ThunderX2 processor used in compute nodes. In addition to documenting the measurement of power during system bring up and for subsequent on-going routine use, we present results associated with controlling the power usage of the processor, an area which is becoming of progressively greater interest as data centers and supercomputing sites look to improve compute/energy efficiency and find additional sources for full system optimization.
The U.S. Army Research Office (ARO), in partnership with IARPA, are investigating innovative, efficient, and scalable computer architectures that are capable of executing next-generation large scale data-analytic applications. These applications are increasingly sparse, unstructured, non-local, and heterogeneous. Under the Advanced Graphic Intelligence Logical computing Environment (AGILE) program, Performer teams will be asked to design computer architectures to meet the future needs of the DoD and the Intelligence Community (IC). This design effort will require flexible, scalable, and detailed simulation to assess the performance, efficiency, and validity of their designs. To support AGILE, Sandia National Labs will be providing the AGILE-enhanced Structural Simulation Toolkit (A-SST). This toolkit is a computer architecture simulation framework designed to support fast, parallel, and multi-scale simulation of novel architectures. This document describes the A-SST framework, some of its library of simulation models, and how it may be used by AGILE Performers.
Scientific applications run on high-performance computing (HPC) systems are critical for many national security missions within Sandia and the NNSA complex. However, these applications often face performance degradation and even failures that are challenging to diagnose. To provide unprecedented insight into these issues, the HPC Development, HPC Systems, Computational Science, and Plasma Theory & Simulation departments at Sandia crafted and completed their FY21 ASC Level 2 milestone entitled "Integrated System and Application Continuous Performance Monitoring and Analysis Capability." The milestone created a novel integrated HPC system and application monitoring and analysis capability by extending Sandia's Kokkos application portability framework, Lightweight Distributed Metric Service (LDMS) monitoring tool, and scalable storage, analysis, and visualization pipeline. The extensions to Kokkos and LDMS enable collection and storage of application data during run time, as it is generated, with negligible overhead. This data is combined with HPC system data within the extended analysis pipeline to present relevant visualizations of derived system and application metrics that can be viewed at run time or post run. This new capability was evaluated using several week-long, 290-node runs of Sandia's ElectroMagnetic Plasma In Realistic Environments ( EMPIRE ) modeling and design tool and resulted in 1TB of application data and 50TB of system data. EMPIRE developers remarked this capability was incredibly helpful for quickly assessing application health and performance alongside system state. In short, this milestone work built the foundation for expansive HPC system and application data collection, storage, analysis, visualization, and feedback framework that will increase total scientific output of Sandia's HPC users.
Sandia National Laboratories is investigating scalable architectural simulation capabilities with a focus on simulating and evaluating highly scalable supercomputers for high performance computing applications. There is a growing demand for RTL model integration to provide the capability to simulate customized node architectures and heterogeneous systems. This report describes the first steps integrating the ESSENTial Signal Simulation Enabled by Netlist Transforms (ESSENT) tool with the Structural Simulation Toolkit (SST). ESSENT can emit C++ models from models written in FIRRTL to automatically generate components. The integration workflow will automatically generate the SST component and necessary interfaces to ’plug’ the ESSENT model into the SST framework.
In this position paper we will address challenges and opportunities relating to the design and codesign of application specific circuits. Given our background as computational scientists, our perspective is from the viewpoint of a highly motivated application developer as opposed to career computer architects
MLIR (Multi-Level Intermediate Representation), is an extensible compiler framework that supports high-level data structures and operation constructs. These higher-level code representations are particularly applicable to the artificial intelligence and machine learning (AI/ML) domain, allowing developers to more easily support upcoming heterogeneous AI/ML accelerators and develop flexible domain specific compilers/frameworks with higher-level intermediate representations (IRs) and advanced compiler optimizations. The result of using MLIR within the LLVM compiler framework is expected to yield significant improvement in the quality of generated machine code, which in turn will result in improved performance and hardware efficiency
1 The exponential growth of data has driven technology providers to develop new protocols, such as cache coherent interconnects and memory semantic fabrics, to help users and facilities leverage advances in memory technologies to satisfy these growing memory and storage demands. Using these new protocols, fabric-Attached memories (FAM) can be directly attached to a system interconnect and be easily integrated with a variety of processing elements (PEs). Moreover, systems that support FAM can be smoothly upgraded and allow multiple PEs to share the FAM memory pools using well-defined protocols. The sharing of FAM between PEs allows efficient data sharing, improves memory utilization, reduces cost by allowing flexible integration of different PEs and memory modules from several vendors, and makes it easier to upgrade the system. One promising use-case for FAMs is in High-Performance Compute (HPC) systems, where the underutilization of memory is a major challenge. However, adopting FAMs in HPC systems brings new challenges. In addition to cost, flexibility, and efficiency, one particular problem that requires rethinking is virtual memory support for security and performance. To address these challenges, this paper presents decoupled access control and address translation (DeACT), a novel virtual memory implementation that supports HPC systems equipped with FAM. Compared to the state-of-The-Art two-level translation approach, DeACT achieves speedup of up to 4.59x (1.8x on average) without compromising security.1Part of this work was done when Vamsee was working under the supervision of Amro Awad at UCF. Amro Awad is now with the ECE Department at NC State.
Non-volatile memories (NVMs) have the characteristics of both traditional storage systems (persistent) and traditional memory systems (byte-Addressable). However, they suffer from high write latency and have a limited write endurance. Researchers have proposed hybrid memory systems that combine DRAM and NVM, utilizing the lower latency of the DRAM to hide some of the shortcomings of the NVM-improving system's performance by caching resident NVM data in the DRAM. However, this can nullify the persistency of the cached pages, leading to a question of trade-offs in terms of performance and reliability. In this paper, we propose Stealth-Persist, a novel architecture support feature that allows applications that need persistence to run in the DRAM while maintaining the persistency features provided by the NVM. Stealth-Persist creates the illusion of a persistent memory for the application to use, while utilizing the DRAM for performance optimizations. Our experimental results show that Stealth-Persist improves the performance by 42.02% for persistent applications.
Programmable accelerators have become commonplace in modern computing systems. Advances in programming models and the availability of unprecedented amounts of data have created a space for massively parallel accelerators capable of maintaining context for thousands of concurrent threads resident on-chip. These threads are grouped and interleaved on a cycle-by-cycle basis among several massively parallel computing cores. One path for the design of future supercomputers relies on an ability to model the performance of these massively parallel cores at scale. The SST framework has been proven to scale up to run simulations containing tens of thousands of nodes. A previous report described the initial integration of the open-source, execution-driven GPU simulator, GPGPU-Sim, into the SST framework. This report discusses the results of the integration and how to use the new GPU component in SST. It also provides examples of what it can be used to analyze and a correlation study showing how closely the execution matches that of a Nvidia V100 GPU when running kernels and mini-apps.
Arm processors have been explored in HPC for several years, however there has not yet been a demonstration of viability for supporting large-scale production workloads. In this paper, we offer a retrospective on the process of bringing up Astra, the first Petascale supercomputer based on 64-bit Arm processors, and validating its ability to run production HPC applications. Through this process several immature technology gaps were addressed, including software stack enablement, Linux bugs at scale, thermal management issues, power management capabilities, and advanced container support. From this experience, several lessons learned are formulated that contributed to the successful deployment of Astra. These insights can be helpful to accelerate deploying and maturing other first-seen HPC technologies. With Astra now supporting many users running a diverse set of production applications at multi-thousand node scales, we believe this constitutes strong supporting evidence that Arm is a viable technology for even the largest-scale supercomputer deployments.
With many recent advances in interconnect technologies and memory interfaces, disaggregated memory systems are approaching industrial adoption. For instance, the recent Gen-Z consortium focuses on a new memory semantic protocol that enables fabric-attached memories (FAM), where the memory and other compute units can be directly attached to fabric interconnects. Decoupling of memory from compute units becomes a feasible option as the rate of data transfer increases due to the emergence of novel interconnect technologies, such as Silicon Photonic Interconnects. Disaggregated memories not only enable more efficient use of capacity (minimizes under-utilization) they also allow easy integration of evolving technologies. Additionally, they simplify the programming model at the same time allowing efficient sharing of data. However, the latency of accessing the data in these Fabric Attached disaggregated Memories (FAMs) is dependent on the latency imposed by the fabric interfaces. To reduce memory access latency and to improve the performance of FAM systems, in this paper, we explore techniques to prefetch data from FAMs to the local memory present in the node (PreFAM). We realize that since the memory access latency is high in FAMs, prefetching a cache block (64 bytes) from FAM can be inefficient, since the possibility of issuing demand requests before the completion of prefetch requests, to the same FAM locations, is high. Hence, we explore predicting and prefetching FAM blocks at a distance; prefetching blocks which are going to be accessed in future but not immediately. We show that, with prefetching, the performance of FAM architectures increases by 38.84%, while memory access latency is improved by 39.6%, with only 17.65% increase in the number of accesses to the FAM, on average. Further, by prefetching at a distance we show a performance improvement of 72.23%.
Community detection in graphs is a canonical social network analysis method. We consider the problem of generating suites of teras-cale synthetic social networks to compare the solution quality of parallel community-detection methods. The standard method, based on the graph generator of Lancichinetti, Fortunato, and Radicchi (LFR), has been used extensively for modest-scale graphs, but has inherent scalability limitations. We provide an alternative, based on the scalable Block Two-Level Erdos-Renyi (BTER) graph generator, that enables HPC-scale evaluation of solution quality in the style of LFR. Our approach varies community coherence, and retains other important properties. Our methods can scale real-world networks, e.g., to create a version of the Friendster network that is 512 times larger. With BTER's inherent scalability, we can generate a 15-terabyte graph (4.6B vertices, 925B edges) in just over one minute. We demonstrate our capability by showing that label-propagation community-detection algorithm can be strong-scaled with negligible solution-quality loss.
As demands for memory-intensive applications continue to grow, the memory capacity of each computing node is expected to grow at a similar pace. In high-performance computing (HPC) systems, the memory capacity per compute node is decided upon the most demanding application that would likely run on such system, and hence the average capacity per node in future HPC systems is expected to grow significantly. However, since HPC systems run many applications with different capacity demands, a large percentage of the overall memory capacity will likely be underutilized; memory modules can be thought of as private memory for its corresponding computing node. Thus, as HPC systems are moving towards the exascale era, a better utilization of memory is strongly desired. Moreover, upgrading memory system requires significant efforts. Fortunately, disaggregated memory systems promise better utilization by defining regions of global memory, typically referred to as memory blades, which can be accessed by all computing nodes in the system, thus achieving much better utilization. Disaggregated memory systems are expected to be built using dense, power-efficient memory technologies. Thus, emerging nonvolatile memories (NVMs) are placing themselves as the main building blocks for such systems. However, NVMs are slower than DRAM. Therefore, it is expected that each computing node would have a small local memory that is based on either HBM or DRAM, whereas a large shared NVM memory would be accessible by all nodes. Managing such system with global and local memory requires a novel hardware/software co-design to initiate page migration between global and local memory to maximize performance while enabling access to huge shared memory. In this paper we provide support to migrate pages, investigate such memory management aspects and the major system-level aspects that can affect design decisions in disaggregated NVM systems
Programmable accelerators have become commonplace in modern computing systems. Advances in programming models and the availability of massive amounts of data have created a space for massively parallel accelerators capable of maintaining context for thousands of concurrent threads resident on-chip. These threads are grouped and interleaved on a cycle-by-cycle basis among several massively parallel computing cores. One path for the design of future supercomputers relies on an ability to model the performance of these massively parallel cores at scale. The SST framework has been proven to scale up to run simulations containing tens of thousands of nodes. A previous report described the initial integration of the open-source, execution-driven GPU simulator, GPGPU-Sim, into the SST framework. This report discusses the results of the integration and how to use the new GPU component in SST. It also provides examples of what it can be used to analyze and a correlation study showing how closely the execution matches that of a Nvidia V100 GPU when running kernels and mini-apps.
The high performance computing industry is undergoing a period of substantial change. Not least because of fabrication and lithographic challenges in the manufacturing of next-generation processors. As such challenges mount, the industry is looking to generate higher performance from additional functionality in the micro-architecture space as well as a greater emphasis on efficiency in the design of networkon-chip resources and memory subsystems. Such variation in design opens opportunities for new entrants in the data center and server markets where varying compute-to-memory ratios can present end users with more efficient node designs for particular workloads. In this paper we compare the recently released Marvell ThunderX2 Arm processor - arguably the first high-performance computing capable Arm design available in the marketplace. We perform a set of micro-benchmarking and mini-application evaluation on the ThunderX2 comparing it with Intel's Haswell and Skylake Xeon server parts commonly used in contemporary HPC designs. Our findings show that no one processor performs the best across all benchmarks, but that the ThunderX2 excels in areas demanding high memory bandwidth due to the provisioning of more memory channels in its design. We conclude that the ThunderX2 is a serious contender in the HPC server segment and has the potential to offer supercomputing sites with a viable high-performance alternative to existing designs from established industry players.
Many applications have growing demands for memory, particularly in the HPC space, making the memory system a potential bottleneck of next-generation computing systems. Sharing the memory system across processor sockets and nodes becomes a compelling argument given that memory technology is scaling at a slower rate than processor technology. Moreover, as many applications rely on shared data, e.g., graph applications and database workloads, having a large number of nodes accessing shared memory allows for efficient use of resources and avoids duplicating huge files, which can be infeasible for large graphs or scientific data. As new memory technologies come on the market, the flexibility of upgrading memory and system updates become major a concern, disaggregated memory systems where memory is shared across different computing nodes, e.g., System-on-Chip (SoC), is expected to become the most common design/architecture on memory-centric systems, e.g., The Machine project from HP Labs. However, due to the nature of such systems, different users and applications compete for the available memory bandwidth, which can lead to severe contention due to memory traffic from different SoCs. In this paper, we discuss the contention problem in disaggregated memory systems and suggest mechanisms to ensure memory fairness and enforce QoS. Our simulation results show that employing our proposed QoS techniques can speed up memory response time by up to 55%.
Programmable accelerators have become commonplace in modern computing systems. Advances in programming models and the availability of massive amounts of data have created a space for massively parallel acceleration where the context for thousands of concurrent threads are resident on-chip. These threads are grouped and interleaved on a cycle-by-cycle basis among several mas- sively parallel computing cores. The design of future supercomputers relies on an ability to model the performance of these massively parallel cores at scale. To address the need for a scalable, decentralized GPU model that can model large GPUs, chiplet- based GPUs and multi-node GPUs, this report details the first steps in integrating the open-source, execution driven GPGPU-Sim into the SST framework. The first stage of this project, creates two elements: a kernel scheduler SST element accepts work from SST CPU models and schedules it to an SM-collection element that performs cycle-by-cycle timing using SSTs Mem Hierarchy to model a flexible memory system.
Despite significant advances in the porting of scientific applications to novel architectures such as compute-optimized graphics processors, many-core processor/accelerators and, even special-purpose function units, the vast majority of scientific calculations are still performed on high-performance, commodity server processors. Even in the cases of applications which have been ported to new architectures, frequent serial sections still require strong server-class processor cores to compute as fast as possible. In this paper we report on a set of benchmark studies which evaluate Intel's latest Skylake Xeon server processor. Skylake represents a significant change in the Xeon product line with wider SIMD vector units, a redesigned cache architecture, and, an increased number of memory channels. The wider vector units provide 2x improvement for some compute-intensive applications and the combined memory changes can provide close to 2x the memory bandwidth. We evaluate these new hardware features on several HPC-relevant mini-Applications and benchmarks, including, STREAM, LULESH, XSBench, HPCG and SW4Lite. Together, the new hardware functions provide up to 1.8x speedup on HPC benchmark codes when compared with the previous generation Haswell processor core, providing much greater utility to a broader range of HPC applications that rely on this class of compute node.
The Vanguard program informally began in January 2017 with the submission of a white pa- per entitled "Sandia's Vision for a 2019 Arm Testbed" to NNSA headquarters. The program proceeded in earnest in May 2017 with an announcement by Doug Wade (Director, Office of Advanced Simulation and Computing and Institutional R&D at NNSA) that Sandia Na- tional Laboratories (Sandia) would host the first Advanced Architecture Prototype platform based on the Arm architecture. In August 2017, Sandia formed a Tri-lab team chartered to develop a robust HPC software stack for Astra to support the Vanguard program goal of demonstrating the viability of Arm in supporting ASC production computing workloads. This document describes the high-level Vanguard program goals, the Vanguard-Astra project acquisition plan and procurement up to contract placement, the initial software stack environment planned for the Vanguard-Astra platform (Astra), a description of how the communities of users will utilize the platform during the transition from the open network to the classified network, and initial performance results.
The Vanguard program informally began in January 2017 with the submission of a white pa- per entitled "Sandia's Vision for a 2019 Arm Testbed" to NNSA headquarters. The program proceeded in earnest in May 2017 with an announcement by Doug Wade (Director, Office of Advanced Simulation and Computing and Institutional R&D at NNSA) that Sandia Na- tional Laboratories (Sandia) would host the first Advanced Architecture Prototype platform based on the Arm architecture. In August 2017, Sandia formed a Tri-lab team chartered to develop a robust HPC software stack for Astra to support the Vanguard program goal of demonstrating the viability of Arm in supporting ASC production computing workloads. This document describes the high-level Vanguard program goals, the Vanguard-Astra project acquisition plan and procurement up to contract placement, the initial software stack environment planned for the Vanguard-Astra platform (Astra), a description of how the communities of users will utilize the platform during the transition from the open network to the classified network, and initial performance results.
Technologies such as Multi-Channel DRAM (MCDRAM) or High Bandwidth Memory (HBM) provide significantly more bandwidth than conventional memory. This trend has raised questions about how applications should manage data transfers between levels. This paper focuses on evaluating different usage modes of the MCDRAM in Intel Knights Landing (KNL) manycore processors. We evaluate these usage modes with a sorting kernel and a sorting-based streaming benchmark. We develop a performance model for the benchmark and use experimental evidence to demonstrate the correctness of the model. The model projects near-optimal numbers of copy threads for memory bandwidth bound computations. We demonstrate on KNL up to a 1.9X speedup for sort when the problem does not fit in MCDRAM over an OpenMP GNU sort that does not use MCDRAM.
Many modern applications have memory footprints that are increasingly large, driving system memory capacities higher and higher. Moreover, these systems are often organized where the bulk of the memory is collocated with the compute capability, which necessitates the need for message passing APIs to facilitate information sharing between compute nodes. Due to the diversity of applications that must run on High-Performance Computing (HPC) systems, the memory utiliza- tion can fluctuate wildly from one application to another. And, because memory is located in the node, maintenance can become problematic because each node must be taken offline and upgraded individually. To address these issues, vendors are exploring disaggregated, memory-centric, systems. In this type of organization, there are discrete nodes,reserved solely for memory, which are shared across many compute nodes. Due to their capacity, low-power, and non-volatility, Non-Volatile Memories (NVMs) are ideal candidates for these memory nodes. This report discusses a new component for the Structural Simulation Toolkit (SST), Opal, that can be used to study the impact of using NVMs in a disaggregated system in terms of performance, security, and memory management. This page intentionally left blank.
Architectures with multiple classes of memory media are becoming a common part of mainstream supercomputer deployments. So called multi-level memories offer differing characteristics for each memory component including variation in bandwidth, latency and capacity. This paper investigates the performance of sparse matrix multiplication kernels on two leading highperformance computing architectures — Intel's Knights Landing processor and NVIDIA's Pascal GPU. We describe a data placement method and a chunking-based algorithm for our kernels that exploits the existence of the multiple memory spaces in each hardware platform. We evaluate the performance of these methods w.r.t. standard algorithms using the auto-caching mechanisms Our results show that standard algorithms that exploit cache reuse performed as well as multi-memory-aware algorithms for architectures such as Ki\iLs where the memory subsystems have similar latencies. However, for architectures such as GPUS where memory subsystems differ significantly in both bandwidth and latency, multi-memory-aware methods are crucial for good performance. In addition, our new approaches permit the user to run problems that require larger capacities than the fastest memory of each compute node without depending on the software-managed cache mechanisms.
Many applications, such as PDE based simulations and machine learning, apply BLAS/LAPACK routines to large groups of small matrices. While existing batched BLAS APIs provide meaningful speedup for this problem type, a non-canonical data layout enabling cross-matrix vectorization may provide further significant speedup. In this paper, we propose a new compact data layout that interleaves matrices in blocks according to the SIMD vector length. We combine this compact data layout with a new interface to BLAS/LAPACK routines that can be used within a hierarchical parallel application. Our layout provides up to 14x, 45x, and 27x speedup against OpenMP loops around optimized DGEMM, DTRSM and DGETRF kernels, respectively, on the Intel Knights Landing architecture. We discuss the compact batched BLAS/LAPACK implementations in two libraries, KokkosKernels and Intel® Math Kernel Library. We demonstrate the APIs in a line solver for coupled PDEs. Finally, we present detailed performance analysis of our kernels.
Triangle counting serves as a key building block for a set of important graph algorithms in network science. In this paper, we address the IEEE HPEC Static Graph Challenge problem of triangle counting, focusing on obtaining the best parallel performance on a single multicore node. Our implementation uses a linear algebra-based approach to triangle counting that has grown out of work related to our miniTri data analytics miniapplication [1] and our efforts to pose graph algorithms in the language of linear algebra. We leverage KokkosKernels to implement this approach efficiently on multicore architectures. Our performance results are competitive with the fastest known graph traversal-based approaches and are significantly faster than the Graph Challenge reference implementations, up to 670,000 times faster than the C++ reference and 10,000 times faster than the Python reference on a single Intel Haswell node.
DRAM scalability is becoming more challenging, pushing the focus of the research community towards alternative memory technologies. Many emerging non-volatile memory (NVM) devices are proving themselves to be good candidates to replace DRAM in the coming years. For example, the recently announced 3D-XPoint memory by Intel/Micron promises latencies that are comparable to DRAM, while being non-volatile and much more dense. While emerging NVMs can be fabricated in different form factors, the most promising (from a performance perspective) are NVM-based DIMMs. Unfortunately, there is a shortage of studies that explore the design options for NVM-based DIMMs. Because of the read and write asymmetries in both power consumption and latency, as well as limited write endurance, which often requires wear-leveling techniques, NVMs require a specialized controller. The fact that future on-die memory controllers are expected to handle different memory technologies pushes future hardware towards on-DIMM controllers. In this paper, we propose an architectural model for NVM-based DIMMs with internal controllers, explore their design space, evaluate different optimizations and reach out to several architectural suggestions. Finally, we make our model publicly available and integrate it with a widely used architectural simulator.
The presentation documented the technical approach of the team and summary of the results with sufficient detail to demonstrate both the value and the completion of the milestone. A separate SAND report was also generated with more detail to supplement the presentation.
This milestone is a tri-lab deliverable supporting ongoing Co-Design efforts impacting applications in the Integrated Codes (IC) program element Advanced Technology Development and Mitigation (ATDM) program element. In FY14, the trilabs looked at porting proxy application to technologies of interest for ATS procurements. In FY15, a milestone was completed evaluating proxy applications in multiple programming models and in FY16, a milestone was completed focusing on the migration of lessons learned back into production code development. This year, the co-design milestone focuses on extracting the knowledge gained and/or code revisions back into production applications.
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. Thus, 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. In this paper, we explore the design space for a user-controlled multi-level main memory. Our work identifies situations in which rewriting application kernels can provide significant performance gains when using near-memory. We present algorithms designed for two-level main memory, using divide-and-conquer to partition computations and streaming to exploit data locality. We consider algorithms for the fundamental application of sorting and for the data analysis kernel k-means. Our algorithms asymptotically reduce memory-block transfers under certain architectural parameter settings. We use and extend Sandia National Laboratories’ SST simulation capability to demonstrate the relationship between increased bandwidth and improved algorithmic performance. Memory access counts from simulations corroborate predicted performance improvements for our sorting algorithm. In contrast, the k-means algorithm is generally CPU bound and does not improve when using near-memory except under extreme conditions. These conditions require large instances that rule out SST simulation, but we demonstrate improvements by running on a customized machine with high and low bandwidth memory. These case studies in co-design serve as positive and cautionary templates, respectively, for the major task of optimizing the computational kernels of many fundamental applications for two-level main memory systems.
DRAM technology is the main building block of main memory, however, DRAM scaling is becoming very challenging. The main issues for DRAM scaling are the increasing error rates with each new generation, the geometric and physical constraints of scaling the capacitor part of the DRAM cells, and the high power consumption caused by the continuous need for refreshing cell values. At the same time, emerging Non- Volatile Memory (NVM) technologies, such as Phase-Change Memory (PCM), are emerging as promising replacements for DRAM. NVMs, when compared to current technologies e.g., NAND-based ash, have latencies comparable to DRAM. Additionally, NVMs are non-volatile, which eliminates the need for refresh power and enables persistent memory applications. Finally, NVMs have promising densities and the potential for multi-level cell (MLC) storage.
Emerging novel architectures for shared memory parallel computing are incorporating increasingly creative innovations to deliver higher memory performance. A notable exemplar of this phenomenon is the Multi-Channel DRAM (MCDRAM) that is included in the Intel® XeonPhi™ processors. In this paper, we examine techniques to use OpenMP to exploit the high bandwidth of MCDRAM by staging data. In particular, we implement double buffering using OpenMP sections and tasks to explicitly manage movement of data into MCDRAM. We compare our double-buffered approach to a non-buffered implementation and to Intel’s cache mode, in which the system manages the MCDRAM as a transparent cache. We also demonstrate the sensitivity of performance to parameters such as dataset size and the distribution of threads between compute and copy operations.
Exascale networks are expected to comprise a significant part of the total monetary cost and 10-20% of the power budget allocated to exascale systems. Yet, our understanding of current and emerging workloads on these networks is limited. Left ignored, this knowledge gap likely will translate into missed opportunities for (1) improved application performance and (2) decreased power and monetary costs in next generation systems. This work targets a detailed understanding and analysis of the performance and utilization of the dragonfly network topology. Using the Structural Simulation Toolkit (SST) and a range of relevant workloads on a dragonfly topology of 110,592 nodes, we examine network design tradeoffs amongst execution time, power, bandwidth, and the number of global links. Our simulations report stalled, active and idle time on a per-port level of the fabric, in order to provide a detailed picture of future networks. The results of this work show potential savings of 3-10% of the exascale power budget and provide valuable insights to researchers looking for new opportunities to improve performance and increase power efficiency of next generation HPC systems.
Multi-Level Memory (MLM) will be an increasingly common organization for main memory. Hybrid main memories that combine conventional DDR and "fast" memory will allow higher peak bandwidth at an attainable cost. However, the chief hurdle for MLM systems is the management of data placement. While user-directed placement may work for some applications, it imposes a heavy burden on the programmer. To avoid this burden while still benefiting from MLM, we propose a number of automated management policies. Our results show that several possible policies offer performance and implementation trade offs. Also, unlike conventional cache or paged memory policies, the addition policy is much more important than the replacement policy.
Managing multi-level memories will require different policies from those used for cache hierarchies, as memory technologies differ in latency, bandwidth, and volatility. To this end we analyze application data allocations and main memory accesses to determine whether an application-driven approach to managing a multi-level memory system comprising stacked and conventional DRAM is viable. Our early analysis shows that the approach is viable, but some applications may require dynamic allocations (i.e., migration) while others are amenable to static allocation.
Machine Models and Proxy Architectures for Exascale Computing Version 2.0 Prepared by Sandia National Laboratories Albuquerque, New Mexico 87185 and Livermore, California 94550 Sandia National Laboratories is a multi-program laboratory managed and operated by Sandia Corporation, a wholly owned subsidiary of Lockheed Martin Corporation, for the U.S. Department of Energy's National Nuclear Security Administration under contract DE-AC04-94AL85000. Approved for public release; further dissemination unlimited. Issued by Sandia National Laboratories, operated for the United States Department of Energy by Sandia Corporation. NOTICE: This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government, nor any agency thereof, nor any of their employees, nor any of their contractors, subcontractors, or their employees, make any warranty, express or implied, or assume any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or rep- resent that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise, does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government, any agency thereof, or any of their contractors or subcontractors. The views and opinions expressed herein do not necessarily state or reflect those of the United States Government, any agency thereof, or any of their contractors. Printed in the United States of America. This report has been reproduced directly from the best available copy. Available to DOE and DOE contractors from U.S. Department of Energy Office of Scientific and Technical Information P.O. Box 62 Oak Ridge, TN 37831 Telephone: (865) 576-8401 Facsimile: (865) 576-5728 E-Mail: reports@adonis.osti.gov Online ordering: http://www.osti.gov/bridge Available to the public from U.S. Department of Commerce National Technical Information Service 5285 Port Royal Rd Springfield, VA 22161 Telephone: (800) 553-6847 Facsimile: (703) 605-6900 E-Mail: orders@ntis.fedworld.gov Online ordering: http://www.ntis.gov/help/ordermethods.asp?loc=7-4-0#online D E P A R T M E N T O F E N E R G Y * * U N I T E D S T A T E S O F A M E R I C A SAND2016-6049 Unlimited Release Printed Abstract Machine Models and Proxy Architectures for Exascale Computing Version 2.0 J.A. Ang 1 , R.F. Barrett 1 , R.E. Benner 1 , D. Burke 2 , C. Chan 2 , J. Cook 1 , C.S. Daley 2 , D. Donofrio 2 , S.D. Hammond 1 , K.S. Hemmert 1 , R.J. Hoekstra 1 , K. Ibrahim 2 , S.M. Kelly 1 , H. Le, V.J. Leung 1 , G. Michelogiannakis 2 , D.R. Resnick 1 , A.F. Rodrigues 1 , J. Shalf 2 , D. Stark, D. Unat, N.J. Wright 2 , G.R. Voskuilen 1 1 1 Sandia National Laboratories, P.O. Box 5800, Albuquerque, New Mexico 87185-MS 1319 2 Lawrence Berkeley National Laboratory, Berkeley, California Abstract To achieve exascale computing, fundamental hardware architectures must change. The most sig- nificant consequence of this assertion is the impact on the scientific and engineering applications that run on current high performance computing (HPC) systems, many of which codify years of scientific domain knowledge and refinements for contemporary computer systems. In order to adapt to exascale architectures, developers must be able to reason about new hardware and deter- mine what programming models and algorithms will provide the best blend of performance and energy efficiency into the future. While many details of the exascale architectures are undefined, an abstract machine model is designed to allow application developers to focus on the aspects of the machine that are important or relevant to performance and code structure. These models are intended as communication aids between application developers and hardware architects during the co-design process. We use the term proxy architecture to describe a parameterized version of an abstract machine model, with the parameters added to elucidate potential speeds and capacities of key hardware components. These more detailed architectural models are formulated to enable discussion between the developers of analytic models and simulators and computer hardware archi- tects. They allow for application performance analysis and hardware optimization opportunities. In this report our goal is to provide the application development community with a set of mod- els that can help software developers prepare for exascale. In addition, through the use of proxy architectures, we can enable a more concrete exploration of how well new and evolving applica- tion codes map onto future architectures. This second version of the document addresses system scale considerations and provides a system-level abstract machine model with proxy architecture information.
The future of memory systems isMulti-LevelMemory (MLM). In a MLM system the main memory is comprised of two or more types of memory instead of a conventional DDR- DRAM-only main memory. By combining different memory technologies, an MLM system can potentially offer more us- Able bandwidth and more capacity for a similar cost as a conventional memory system. However, substantial software and hardware design challenges must be overcome to make this potential real. It is our position that the diversity of application access pat- Terns precludes any simple "one size fits all" approach and that better tools and design processes will be needed to ful- fill the potential of MLM. Effcient implementations of MLM will require a high degree of co-design and coordination be- Tween hardware and software. The simulation framework we have built for this study can aid tool building to solve the programming challenges.
A fundamental challenge for supercomputer architecture is that processors cannot be fed data from DRAM as fast as CPUs can consume it. Therefore, many applications are memory-bandwidth bound. As the number of cores per chip increases, and traditional DDR DRAM speeds stagnate, the problem is only getting worse. A variety of non-DDR 3D memory technologies (Wide I/O 2, HBM) offer higher bandwidth and lower power by stacking DRAM chips on the processor or nearby on a silicon interposer. However, such a packaging scheme cannot contain sufficient memory capacity for a node. It seems likely that future systems will require at least two levels of main memory: high-bandwidth, low-power memory near the processor and low-bandwidth high-capacity memory further away. This near memory will probably not have significantly faster latency than the far memory. This, combined with the large size of the near memory (multiple GB) and power constraints, may make it difficult to treat it as a standard cache. In this paper, we explore some of the design space for a user-controlled multi-level main memory. We present algorithms designed for the heterogeneous bandwidth, using streaming to exploit data locality. We consider algorithms for the fundamental application of sorting. Our algorithms asymptotically reduce memory-block transfers under certain architectural parameter settings. We use and extend Sandia National Laboratories' SST simulation capability to demonstrate the relationship between increased bandwidth and improved algorithmic performance. Memory access counts from simulations corroborate predicted performance. This co-design effort suggests implementing two-level main memory systems may improve memory performance in fundamental applications.
Sandia National Laboratories has been engaged in hardware and software codesign activities for a number of years, indeed, it might be argued that prototyping of clusters as far back as the CPLANT machines and many large capability resources including ASCI Red and RedStorm were examples of codesigned solutions. As the research supporting our codesign activities has moved closer to investigating on-node runtime behavior a nature hunger has grown for detailed analysis of both hardware and algorithm performance from the perspective of low-level operations. The Application Characterization for Exascale (APEX) LDRD was a project concieved of addressing some of these concerns. Primarily the research was to intended to focus on generating accurate and reproducible low-level performance metrics using tools that could scale to production-class code bases. Along side this research was an advocacy and analysis role associated with evaluating tools for production use, working with leading industry vendors to develop and refine solutions required by our code teams and to directly engage with production code developers to form a context for the application analysis and a bridge to the research community within Sandia. On each of these accounts significant progress has been made, particularly, as this report will cover, in the low-level analysis of operations for important classes of algorithms. This report summarizes the development of a collection of tools under the APEX research program and leaves to other SAND and L2 milestone reports the description of codesign progress with Sandia’s production users/developers.
This report provides in-depth information and analysis to help create a technical road map for developing next-generation programming models and runtime systems that support Advanced Simulation and Computing (ASC) work- load requirements. The focus herein is on asynchronous many-task (AMT) model and runtime systems, which are of great interest in the context of "Oriascale7 computing, as they hold the promise to address key issues associated with future extreme-scale computer architectures. This report includes a thorough qualitative and quantitative examination of three best-of-class AIM] runtime systems – Charm-++, Legion, and Uintah, all of which are in use as part of the Centers. The studies focus on each of the runtimes' programmability, performance, and mutability. Through the experiments and analysis presented, several overarching Predictive Science Academic Alliance Program II (PSAAP-II) Asc findings emerge. From a performance perspective, AIV runtimes show tremendous potential for addressing extreme- scale challenges. Empirical studies show an AM runtime can mitigate performance heterogeneity inherent to the machine itself and that Message Passing Interface (MP1) and AM11runtimes perform comparably under balanced conditions. From a programmability and mutability perspective however, none of the runtimes in this study are currently ready for use in developing production-ready Sandia ASC applications. The report concludes by recommending a co- design path forward, wherein application, programming model, and runtime system developers work together to define requirements and solutions. Such a requirements-driven co-design approach benefits the community as a whole, with widespread community engagement mitigating risk for both application developers developers. and high-performance computing runtime systein
For the FY15 ASC L2 Trilab Codesign milestone Sandia National Laboratories performed two main studies. The first study investigated three topics (performance, cross-platform portability and programmer productivity) when using OpenMP directives and the RAJA and Kokkos programming models available from LLNL and SNL respectively. The focus of this first study was the LULESH mini-application developed and maintained by LLNL. In the coming sections of the report the reader will find performance comparisons (and a demonstration of portability) for a variety of mini-application implementations produced during this study with varying levels of optimization. Of note is that the implementations utilized including optimizations across a number of programming models to help ensure claims that Kokkos can provide native-class application performance are valid. The second study performed during FY15 is a performance assessment of the MiniAero mini-application developed by Sandia. This mini-application was developed by the SIERRA Thermal-Fluid team at Sandia for the purposes of learning the Kokkos programming model and so is available in only a single implementation. For this report we studied its performance and scaling on a number of machines with the intent of providing insight into potential performance issues that may be experienced when similar algorithms are deployed on the forthcoming Trinity ASC ATS platform.