SEARCH · Search NASA
Results for “memory efficiency”
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.
Ultraefficient resistance switching between charge ordered phases in 1 T -TaS 2 with a single picosecond electrical pulse
Progress in high-performance computing demands significant advances in memory technology. Among novel memory technologies that promise efficient device operation on a sub-ns timescale, resistance switching between charge ordered phases of 1T-TaS 2 has shown to be potentially useful for development of high-speed, energy efficient nonvolatile memory devices. Measurement of the electrical operation of such devices in the picosecond regime is technically challenging and hitherto still largely unexplored. Here, we use an optoelectronic “laboratory-on-a-chip” experiment for measurement of ultrafast memory switching, enabling accurate measurement of electrical switching parameters with 100 fs temporal resolution. Photoexcitation and electro-optic sampling on a (Cd,Mn)Te substrate are used to generate and, subsequently, measure electrical pulse propagation with intra-band excitation and sub-gap probing, respectively. We demonstrate high contrast nonvolatile resistance switching from high to low resistance states of a 1T-TaS 2 device using single sub-2ps electrical pulses. Using detailed modeling, we find that the switching energy density per unit area is exceptionally small, E A =9.4 fJ/μm 2 . Finally, the speed and energy efficiency of an electronic “write” process place the 1T-TaS 2 devices into a category of their own among new generation nonvolatile memory devices.
Analyzing the trade-off between multiple memory controllers and memory channels on multi-core processor performance
Increasing the core-count on current and future processors is posing critical challenges to the memory subsystem to efficiently handle concurrent memory requests. The current trend to cope with this challenge is to increase the number of memory channels available to the processor's memory controller. In this paper we investigate the effectiveness of this approach on the performance of parallel scientific applications. Specifically, we explore the trade-off between employing multiple memory channels per memory controller and the use of multiple memory controllers. Experiments conducted on two current state-of-the-art multicore processors, a 6-core AMD Istanbul and a 4-core Intel Nehalem-EP, for a wide range of production applications shows that there is a diminishing return when increasing the number of memory channels per memory controller. In addition, we show that this performance degradation can be efficiently addressed by increasing the ratio of memory controllers to channels while keeping the number of memory channels constant. Significant performance improvements can be achieved in this scheme, up to 28%, in the case of using two memory controllers with each with one channel compared with one controller with two memory channels.
Memory access response merging in a memory hierarchy
A system and method for efficiently processing memory requests are described. A computing system includes multiple compute units, multiple caches of a memory hierarchy and a communication fabric. A compute unit generates a memory access request that misses in a higher level cache, which sends a miss request to a lower level shared cache. During servicing of the miss request, the lower level cache merges identification information of multiple memory access requests targeting a same cache line from multiple compute units into a merged memory access response. The lower level shared cache continues to insert information into the merged memory access response until the lower level shared cache is ready to issue the merged memory access response. An intermediate router in the communication fabric broadcasts the merged memory access response into multiple memory access responses to send to corresponding compute units.
Magnetic Switching in Monolayer 2D Diluted Magnetic Semiconductors via Spin‐to‐Spin Conversion
Abstract The integration of 2D van der Waals (vdW) magnets with topological insulators or heavy metals holds great potential for realizing next‐generation spintronic memory devices. However, achieving high‐efficiency spin–orbit torque (SOT) switching of monolayer vdW magnets at room temperature poses a significant challenge, particularly without an external magnetic field. Here, it is shown field‐free, deterministic, and nonvolatile SOT switching of perpendicular magnetization in the monolayer, diluted magnetic semiconductor (DMS), Fe‐doped MoS 2 (Fe:MoS 2 ) at up to 380 K with a current density of ≈7 × 10 4 A cm −2 . The in situ doping of Fe into monolayer MoS 2 via chemical vapor deposition and the geometry‐induced strain in the crystal break the rotational switching symmetry in Fe:MoS 2 , promoting field‐free SOT switching by generating out‐of‐plane spins via spin‐to‐spin conversion. An apparent anomalous Hall effect (AHE) loop shift at a zero in‐plane magnetic field verifies the existence of z spins in Fe:MoS 2 , inducing an antidamping‐like torque that facilitates field‐free SOT switching. This field‐free SOT application using a 2D ferromagnetic monolayer provides a new pathway for developing highly power‐efficient spintronic memory devices.
Efficient ICCG on a shared memory multiprocessor
Different approaches are discussed for exploiting parallelism in the ICCG (Incomplete Cholesky Conjugate Gradient) method for solving large sparse symmetric positive definite systems of equations on a shared memory parallel computer. Techniques for efficiently solving triangular systems and computing sparse matrix-vector products are explored. Three methods for scheduling the tasks in solving triangular systems are implemented on the Sequent Balance 21000. Sample problems that are representative of a large class of problems solved using iterative methods are used. We show that a static analysis to determine data dependences in the triangular solve can greatly improve its parallel efficiency. We also show that ignoring symmetry and storing the whole matrix can reduce solution time substantially.
Technical support for digital systems technology development. Task order 1: ISP contention analysis and control
Alternatives for realizing a packet-based network switch for use on a frequency division multiple access/time division multiplexed (FDMA/TDM) geostationary communication satellite were investigated. Each of the eight downlink beams supports eight directed dwells. The design needed to accommodate multicast packets with very low probability of loss due to contention. Three switch architectures were designed and analyzed. An output-queued, shared bus system yielded a functionally simple system, utilizing a first-in, first-out (FIFO) memory per downlink dwell, but at the expense of a large total memory requirement. A shared memory architecture offered the most efficiency in memory requirements, requiring about half the memory of the shared bus design. The processing requirement for the shared-memory system adds system complexity that may offset the benefits of the smaller memory. An alternative design using a shared memory buffer per downlink beam decreases circuit complexity through a distributed design, and requires at most 1000 packets of memory more than the completely shared memory design. Modifications to the basic packet switch designs were proposed to accommodate circuit-switched traffic, which must be served on a periodic basis with minimal delay. Methods for dynamically controlling the downlink dwell lengths were developed and analyzed. These methods adapt quickly to changing traffic demands, and do not add significant complexity or cost to the satellite and ground station designs. Methods for reducing the memory requirement by not requiring the satellite to store full packets were also proposed and analyzed. In addition, optimal packet and dwell lengths were computed as functions of memory size for the three switch architectures.
Optimizing Management of Persistent Data Structures in High-Performance Analytics
Large-scale data analytics workflows ingest massive input data into various data structures, including graphs and key-value datastores. These data structures undergo multiple transformations and computations and are typically reused in incremental and iterative analytics workflows. Persisting in-memory views of these data structures enables reusing them beyond the scope of a single program run while avoiding repetitive raw data ingestion overheads. Memory-mapped I/O enables persisting in-memory data structures without data serialization and deserialization overheads. However, memory-mapped I/O lacks the key feature of persisting consistent snapshots of these data structures for incremental ingestion and processing. The obstacles to efficient virtual memory snapshots using memory-mapped I/O include background writebacks outside the application’s control, and the significantly high storage footprint of such snapshots. To address these limitations, we present Privateer, a memory and storage management tool that enables storage-efficient virtual memory snapshotting while also optimizing snapshot I/O performance. Here, we integrated Privateer into Metall, a state-of-the-art persistent memory allocator for C++, and the Lightning Memory-Mapped Database (LMDB), a widely-used key-value datastore in data analytics and machine learning. Privateer optimized application performance by 1.22× when storing data structure snapshots to node-local storage, and up to 16.7× when storing snapshots to a parallel file system. Privateer also optimizes storage efficiency of incremental data structure snapshots by up to 11× using data deduplication and compression.
Distributed Saturation
The Saturation algorithm for symbolic state-space generation, has been a recent break-through in the exhaustive veri cation of complex systems, in particular globally-asyn- chronous/locally-synchronous systems. The algorithm uses a very compact Multiway Decision Diagram (MDD) encoding for states and the fastest symbolic exploration algo- rithm to date. The distributed version of Saturation uses the overall memory available on a network of workstations (NOW) to efficiently spread the memory load during the highly irregular exploration. A crucial factor in limiting the memory consumption during the symbolic state-space generation is the ability to perform garbage collection to free up the memory occupied by dead nodes. However, garbage collection over a NOW requires a nontrivial communication overhead. In addition, operation cache policies become critical while analyzing large-scale systems using the symbolic approach. In this technical report, we develop a garbage collection scheme and several operation cache policies to help on solving extremely complex systems. Experiments show that our schemes improve the performance of the original distributed implementation, SmArTNow, in terms of time and memory efficiency.
Hardware-Efficient Quantum Random Access Memory with Hybrid Quantum Acoustic Systems
Hybrid quantum systems in which acoustic resonators couple to superconducting qubits are promising quantum information platforms. High quality factors and small mode volumes make acoustic modes ideal quantum memories, while the qubit-phonon coupling enables the initialization and manipulation of quantum states. In this paper, we present a scheme for quantum computing with multimode quantum acoustic systems, and based on this scheme, propose a hardware-efficient implementation of a quantum random access memory (QRAM). Quantum information is stored in high-$\textit{Q}$ phonon modes, and couplings between modes are engineered by applying off-resonant drives to a transmon qubit. In comparison to existing proposals that involve directly exciting the qubit, this scheme can offer a substantial improvement in gate fidelity for long-lived acoustic modes. We show how these engineered phonon-phonon couplings can be used to access data in superposition according to the state of designated address modes—implementing a QRAM on a single chip.
Multimode quantum memory based on atomic frequency combs
An efficient multimode quantum memory is a crucial resource for long-distance quantum communication based on quantum repeaters. We propose a quantum memory based on spectral shaping of an inhomogeneously broadened optical transition into an atomic frequency comb (AFC). The spectral width of the AFC allows efficient storage of multiple temporal modes without the need to increase the absorption depth of the storage material, in contrast to previously known quantum memories. Efficient readout is possible thanks to rephasing of the atomic dipoles due to the AFC structure. Long-time storage and on-demand readout is achieved by use of spin states in a lambda-type configuration. We show that an AFC quantum memory realized in solids doped with rare-earth-metal ions could store hundreds of modes or more with close to unit efficiency, for material parameters achievable today.
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.
Pushing memory bandwidth limitations through efficient implementations of Block-Krylov space solvers on GPUs
Lattice quantum chromodynamics simulations in nuclear physics have benefited from a tremendous number of algorithmic advances such as multigrid and eigenvector deflation. These improve the time to solution but do not alleviate the intrinsic memory-bandwidth constraints of the matrix-vector operation dominating iterative solvers. Batching this operation for multiple vectors and exploiting cache and register blocking can yield a super-linear speed up. Block-Krylov solvers can naturally take advantage of such batched matrix-vector operations, further reducing the iterations to solution by sharing the Krylov space between solves. However, practical implementations typically suffer from the quadratic scaling in the number of vector-vector operations. Here, using the QUDA library, we present an implementation of a block-CG solver on NVIDIA GPUs which reduces the memory-bandwidth complexity of vector-vector operations from quadratic to linear. We present results for the HISQ discretization, showing a 5x speedup compared to highly-optimized independent Krylov solves on NVIDIA's SaturnV cluster.
Very Large Scale Optimization
The purpose of this research under the NASA Small Business Innovative Research program was to develop algorithms and associated software to solve very large nonlinear, constrained optimization tasks. Key issues included efficiency, reliability, memory, and gradient calculation requirements. This report describes the general optimization problem, ten candidate methods, and detailed evaluations of four candidates. The algorithm chosen for final development is a modern recreation of a 1960s external penalty function method that uses very limited computer memory and computational time. Although of lower efficiency, the new method can solve problems orders of magnitude larger than current methods. The resulting BIGDOT software has been demonstrated on problems with 50,000 variables and about 50,000 active constraints. For unconstrained optimization, it has solved a problem in excess of 135,000 variables. The method includes a technique for solving discrete variable problems that finds a "good" design, although a theoretical optimum cannot be guaranteed. It is very scalable in that the number of function and gradient evaluations does not change significantly with increased problem size. Test cases are provided to demonstrate the efficiency and reliability of the methods and software.
Perpendicular spin transfer torque magnetic random access memories with high spin torque efficiency and thermal stability for embedded applications (invited)
Magnetic random access memories based on the spin transfer torque phenomenon (STT-MRAMs) have become one of the leading candidates for next generation memory applications. Among the many attractive features of this technology are its potential for high speed and endurance, read signal margin, low power consumption, scalability, and non-volatility. In this paper, we discuss our recent results on perpendicular STT-MRAM stack designs that show STT efficiency higher than 5 k{sub B}T/μA, energy barriers higher than 100 k{sub B}T at room temperature for sub-40 nm diameter devices, and tunnel magnetoresistance higher than 150%. We use both single device data and results from 8 Mb array to demonstrate data retention sufficient for automotive applications. Moreover, we also demonstrate for the first time thermal stability up to 400 °C exceeding the requirement of Si CMOS back-end processing, thus opening the realm of non-volatile embedded memory to STT-MRAM technology.
Devices for Efficient Analog In-Memory Neural Computing at the Edge.
Abstract not provided.
Exploring Hybrid Memory for GPU Energy Efficiency through Software-Hardware Co-Design
Explore the source record for details and available documents.
Performance Potential of Mixed Data Management Modes for Heterogeneous Memory Systems
Many high-performance systems now include different types of memory devices within the same compute platform to meet strict performance and cost constraints. Such heterogeneous memory systems often include an upper-level tier with better performance, but limited capacity, and lower-level tiers with higher capacity, but less bandwidth and longer latencies for reads and writes. To utilize the different memory layers efficiently, current systems rely on hardware-directed, memory -side caching or they provide facilities in the operating system (OS) that allow applications to make their own data-tier assignments. Since these data management options each come with their own set of trade-offs, many systems also include mixed data management configurations that allow applications to employ hardware- and software-directed management simultaneously, but for different portions of their address space. Despite the opportunity to address limitations of stand-alone data management options, such mixed management modes are under-utilized in practice, and have not been evaluated in prior studies of complex memory hardware. In this work, we develop custom program profiling, configurations, and policies to study the potential of mixed data management modes to outperform hardware- or software-based management schemes alone. Our experiments, conducted on an Intel ® Knights Landing platform with high-bandwidth memory, demonstrate that the mixed data management mode achieves the same or better performance than the best stand-alone option for five memory intensive benchmark applications (run separately and in isolation), resulting in an average speedup compared to the best stand-alone policy of over 10 %, on average.