Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “processor performance factors”

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 37 records · Page 2

Improved Dynamic Modeling of the Cascade Distillation Subsystem and Analysis of Factors Affecting Its Performance

The Cascade Distillation Subsystem (CDS) is a rotary multistage distiller being developed to serve as the primary processor for wastewater recovery during long-duration space missions. The CDS could be integrated with a system similar to the International Space Station Water Processor Assembly to form a complete water recovery system for future missions. A preliminary chemical process simulation was previously developed using Aspen Custom Modeler® (ACM), but it could not simulate thermal startup and lacked detailed analysis of several key internal processes, including heat transfer between stages. This paper describes modifications to the ACM simulation of the CDS that improve its capabilities and the accuracy of its predictions. Notably, the modified version can be used to model thermal startup and predicts the total energy consumption of the CDS. The simulation has been validated for both NaC1 solution and pretreated urine feeds and no longer requires retuning when operating parameters change. The simulation was also used to predict how internal processes and operating conditions of the CDS affect its performance. In particular, it is shown that the coefficient of performance of the thermoelectric heat pump used to provide heating and cooling for the CDS is the largest factor in determining CDS efficiency. Intrastage heat transfer affects CDS performance indirectly through effects on the coefficient of performance.

Perry, Bruce A.↗

Radiometric Compensation and Calibration for Radarsat ScanSAR

Due to lack of a standard for modeling the radar echo signal in terms of signal unit and coordinates as well as lack of a standard in designing the gain factors in each stage of a processor, absolute radiometric calibration of a SAR system is usually performed by treating the sensor and processor as one inseparable unit. This often makes the calibration procedure complicated and requiring the involvement of both radar system engineers and processor engineers in the whole process. This paper introduces a standard for modeling the radar echo signal and a standard in designing the gain factor of a ScanSAR processor. In this paper, the radar equation is derived based on the amount of energy instead of the power received from a backscatterer. These efforts lead to simple and easy-to-understand equations for radiometric compensation and calibration.

Jin, Michael Y.↗

Partial Overhaul and Initial Parallel Optimization of KINETICS, a Coupled Dynamics and Chemistry Atmosphere Model

KINETICS is a coupled dynamics and chemistry atmosphere model that is data intensive and computationally demanding. The potential performance gain from using a supercomputer motivates the adaptation from a serial version to a parallelized one. Although the initial parallelization had been done, bottlenecks caused by an abundance of communication calls between processors led to an unfavorable drop in performance. Before starting on the parallel optimization process, a partial overhaul was required because a large emphasis was placed on streamlining the code for user convenience and revising the program to accommodate the new supercomputers at Caltech and JPL. After the first round of optimizations, the partial runtime was reduced by a factor of 23; however, performance gains are dependent on the size of the data, the number of processors requested, and the computer used.

KINETICS↗

Optimizing the inner loop of the gravitational force interaction on modern processors

We have achieved superior performance on multiple generations of the fastest supercomputers in the world with our hashed oct-tree N-body code (HOT), spanning almost two decades and garnering multiple Gordon Bell Prizes for significant achievement in parallel processing. Execution time for our N-body code is largely influenced by the force calculation in the inner loop. Improvements to the inner loop using SSE3 instructions has enabled the calculation of over 200 million gravitational interactions per second per processor on a 2.6 GHz Opteron, for a computational rate of over 7 Gflops in single precision (700/0 of peak). We obtain optimal performance some processors (including the Cell) by decomposing the reciprocal square root function required for a gravitational interaction into a table lookup, Chebychev polynomial interpolation, and Newton-Raphson iteration, using the algorithm of Karp. By unrolling the loop by a factor of six, and using SPU intrinsics to compute on vectors, we obtain performance of over 16 Gflops on a single Cell SPE. Aggregated over the 8 SPEs on a Cell processor, the overall performance is roughly 130 Gflops. In comparison, the ordinary C version of our inner loop only obtains 1.6 Gflops per SPE with the spuxlc compiler.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

HPC-NMF: A High-Performance Parallel Algorithm for Nonnegative Matrix Factorization

NMF is a useful tool for many applications in different domains such as topic modeling in text mining, background separation in video analysis, and community detection in social networks. Despite its popularity in the data mining community, there is a lack of efficient distributed algorithms to solve the problem for big data sets. We propose a high-performance distributed-memory parallel algorithm that computes the factorization by iteratively solving alternating non-negative least squares (NLS) subproblems for $\WW$ and $\HH$. It maintains the data and factor matrices in memory (distributed across processors), uses MPI for interprocessor communication, and, in the dense case, provably minimizes communication costs (under mild assumptions). As opposed to previous implementation, our algorithm is also flexible: It performs well for both dense and sparse matrices, and allows the user to choose any one of the multiple algorithms for solving the updates to low rank factors $\WW$ and $\HH$ within the alternating iterations.

Kannan, Ramakrishnan↗

Internship Abstract and Final Reflection

The primary objective for this internship is the evaluation of an embedded natural language processor (NLP) as a way to introduce voice control into future space suits. An embedded natural language processor would provide an astronaut hands-free control for making adjustments to the environment of the space suit and checking status of consumables procedures and navigation. Additionally, the use of an embedded NLP could potentially reduce crew fatigue, increase the crewmember's situational awareness during extravehicular activity (EVA) and improve the ability to focus on mission critical details. The use of an embedded NLP may be valuable for other human spaceflight applications desiring hands-free control as well. An embedded NLP is unique because it is a small device that performs language tasks, including speech recognition, which normally require powerful processors. The dedicated device could perform speech recognition locally with a smaller form-factor and lower power consumption than traditional methods.

Sandor, Edward↗

Hierarchical power distribution in large scale computing systems

Techniques for managing power distribution amongst processors in a massively parallel computer architecture are disclosed. The techniques utilize a hierarchy that organizes the various processors of the massively parallel computer architecture. The hierarchy groups numbers of the processors at the lowest level. When processors complete tasks, the power assigned to those processors is distributed to other processors in the same group so that the performance of those processors can be increased. Hierarchical organization simplifies the calculations required for determining how and when to distribute power, because when tasks are complete and power is available for distribution, a relatively small number of processors are available for consideration to receive that power. The number of processors that are grouped together can be adjusted in real time based on performance factors to improve the trade-off between calculation speed and power distribution efficacy.

42 ENGINEERING↗

A high-performance parallel algorithm for nonnegative matrix factorization

Non-negative matrix factorization (NMF) is the problem of determining two non-negative low rank factors W and H, for the given input matrix A, such that A ≈ WH. NMF is a useful tool for many applications in different domains such as topic modeling in text mining, background separation in video analysis, and community detection in social networks. Despite its popularity in the data mining community, there is a lack of efficient distributed algorithms to solve the problem for big data sets. We propose a high-performance distributed-memory parallel algorithm that computes the factorization by iteratively solving alternating non-negative least squares (NLS) subproblems for W and H. It maintains the data and factor matrices in memory (distributed across processors), uses MPI for interprocessor communication, and, in the dense case, provably minimizes communication costs (under mild assumptions). As opposed to previous implementations, our algorithm is also flexible: (1) it performs well for both dense and sparse matrices, and (2) it allows the user to choose any one of the multiple algorithms for solving the updates to low rank factors W and H within the alternating iterations. We demonstrate the scalability of our algorithm and compare it with baseline implementations, showing significant performance improvements.

Kannan, Ramakrishnan↗

Advanced technology for a satellite multichannel demultiplexer/demodulator

Satellite on-board processing is needed to efficiently service multiple users while at the same time minimizing earth station complexity. The processing satellite receives a wideband uplink at 30 GHz and down-converts it to a suitable intermediate frequency. A multichannel demultiplexer then separates the composite signal into discrete channels. Each channel is then demodulated by bulk demodulators, with the baseband signals routed to the downlink processor for retransmission to the receiving earth stations. This type of processing circumvents many of the difficulties associated with traditional bent-pipe repeater satellites. Uplink signal distortion and interference are not retransmitted on the downlink. Downlink power can be allocated in accordance with user needs, independent of uplink transmissions. This allows the uplink users to employ different data rates as well as different modulation and coding schemes. In addition, all downlink users have a common frequency standard and symbol clock on the satellite, which is useful for network synchronization in time division multiple access schemes. The purpose of this program is to demonstrate the concept of an optically implemented multichannel demultiplexer (MCD). A proof-of-concept (POC) model has been developed which has the ability to receive a 40 MHz wide composite signal consisting of up to 1000 40 kHz QPSK modulated channels and perform the demultiplexing process. In addition a set of special test equipment (STE) has been configured to evaluate the performance of the POC model. The optical MCD is realized as an acousto-optic spectrum analyzer utilizing the capability of Bragg cells to perform the required channelization. These Bragg cells receive an optical input from a laser source and an RF input (the signal). The Bragg interaction causes optical output diffractions at angles proportional to the RF input frequency. These discrete diffractions are optically detected and output to individual demodulators for baseband conversion. Optimization of the MCD design was conducted in order to achieve a compromise between two opposing sources of signal degradation: adjacent channel interference and intersymbol interference. The system was also optimized to allow simple, inexpensive ground stations communications with the MCD. These design goals led to the realization of a POC MCD which demonstrates the demultiplexing function with minimal signal degradation. Performance evaluation results using the STE equipment indicate that the dynamic range of the demultiplexer in the presence of adjacent and multiple channel loading is 40 - 50 dB. Measured bit error rate (BER) probabilities varied from the predicted theoretical results by one dB or less. The performance of the proof-of-concept model indicate that the development of a space qualified optically implemented MCD are feasible. The advantages to such an implementation include reduced size, weight and power and increased reliability when compared with electronic approaches. All of these factors are critical to on-board satellite processors. Further optimization can be conducted which trade ground station complexity and MCD performance to achieve desired system results.

Abramovitz, Irwin J.↗

Effective Software Design and Development for the New Graph Architecture HPC Machines

Software applications need to change and adapt as modern architectures evolve. Nowadays advancement in chip design translates to increased parallelism. Exploiting such parallelism is a major challenge in modern software engineering. Multicore processors are about to introduce a significant change in the way we design and use fundamental data structures. In this work we describe the design and programming principles of a software library of highly concurrent scalable and nonblocking data containers. In this project we have created algorithms and data structures for handling fundamental computations in massively multithreaded contexts, and we have incorporated these into a usable library with familiar look and feel. In this work we demonstrate the first design and implementation of a wait-free hash table. Our multiprocessor data structure design allows a large number of threads to concurrently insert, remove, and retrieve information. Non-blocking designs alleviate the problems traditionally associated with the use of mutual exclusion, such as bottlenecks and thread-safety. Lock-freedom provides the ability to share data without some of the drawbacks associated with locks, however, these designs remain susceptible to starvation. Furthermore, wait-freedom provides all of the benefits of lock-free synchronization with the added assurance that every thread makes progress in a finite number of steps. This implies deadlock-freedom, livelock-freedom, starvation-freedom, freedom from priority inversion, and thread-safety. The challenges of providing the desirable progress and correctness guarantees of wait-free objects makes their design and implementation difficult. There are few wait-free data structures described in the literature. Using only standard atomic operations provided by the hardware, our design is portable; therefore, it is applicable to a variety of data-intensive applications including the domains of embedded systems and supercomputers.Our experimental evaluation shows that our hash table design outperforms the most advanced locking solution, provided by Intel's TBB library, by 22%. When compared to more traditional locking designs we show a performance improvement by a factor of 7.92. When compared to alternative non-blocking designs, our hash table demonstrates solid performance gains in a large majority of cases, typically by a factor of 3.44.

99 GENERAL AND MISCELLANEOUS↗

Efficient On-demand Connection Management Mechanisms with PGAS Models on InfiniBand

In the last decade or so, clusters have observed a tremendous rise in popularity due to the excellent price to performance ratio. A variety of Interconnects have been proposed during this period, with InfiniBand leading the way due to its high performance and open standard. At the same time, multiple programming models have emerged in order to meet the requirements of various applications and their programming models. To support requirements of multiple programming models, InfiniBand provides multiple transport semantics, ranging from unreliable connectionless to reliable connected characteristics. Among them, the reliable connection (RC) semantics is being widely due to its high performance and support for novel features like Remote Direct Memory Acesss (RDMA), hardware atomics and Network Fault Tolerance. However, the pairwise connection oriented nature of the RC transport semantics limits its scalability and usage at the increasing processor counts. In this paper, we design and implement on-demand connection management approaches in the context of Partitioned Global Address Space (PGAS) programming models, which provided shared memory abstraction and one-sided communication semantics, leading to the development of multiple languages (UPC, X10, Chapel) and libraries (Global Arrays, MPI-RMA). Using Global Arrays as the research vehicle, we implement this approach with Aggregate Remote Memory Copy Interface (ARMCI), the runtime system of Global Arrays. We evaluate our approach, ARMCI-On Demand Connection Management (ARMCI-ODCM) using various microbenchmarks and benchmarks (LU Factorization, RandomAccess and Lennard Jones simulation) and application (Subsurface transport over multiple phases (STOMP)). With the performance evaluation for up to 4096 processors, we are able to have a multi-fold reduction in connection memory with a negligible degradation in performance. Using STOMP at 4096 processors, reduces the overall connection memory by 66 times with no performance degradation. To the best of our knowledge, this is the first design, implementation and evaluation of on-demand connection management with InfiniBand using PGAS models.

54 ENVIRONMENTAL SCIENCES↗

Scalable Out-of-Core Solvers on Xeon Phi Cluster

This paper documents the implementation of a distributive out-of-core (OOC) solver for performing LU and Cholesky factorizations of a large dense matrix on clusters of many-core programmable co-processors. The out-of- core algorithm combines both the left-looking and right-looking schemes aimed to minimize the movement of data between the CPU host and the co-processor, optimizing data locality as well as computing throughput. The OOC solver is built to align with the format of the ScaLAPACK software library, making it readily portable to any existing codes using ScaLAPACK. A runtime analysis conducted on Beacon (an Intel Xeon plus Intel Xeon Phi cluster which composed of 48 nodes of multi-core CPU and MIC) at the Na- tional Institute for Computational Sciences is presented. Comparison of the performance on the Intel Xeon Phi and GPU clusters are also provided.

D'Azevedo, Ed F↗

Data Movement in Data-Intensive High Performance Computing.

The cost of executing a floating point operation has been decreasing for decades at a much higher rate than that of moving data. Bandwidth and latency, two key metrics that determine the cost of moving data, have degraded significantly relative to processor cycle time and execution rate. Despite the limitation of sub-micron processor technology and the end of Dennard scaling, this trend will continue in the short-term making data movement a performance-limiting factor and an energy/power efficiency concern. Even more so in the context of large-scale and data-intensive systems and workloads. This chapter gives an overview of the aspects of moving data across a system, from the storage system to the computing system down to the node and processor level, with case study and contributions from researchers at the San Diego Supercomputer Center, the Oak Ridge National Laboratory, the Pacific Northwest National Laboratory, and the University of Delaware.

Data Movement↗

Data management system performance modeling

This paper discusses analytical techniques that have been used to gain a better understanding of the Space Station Freedom's (SSF's) Data Management System (DMS). The DMS is a complex, distributed, real-time computer system that has been redesigned numerous times. The implications of these redesigns have not been fully analyzed. This paper discusses the advantages and disadvantages for static analytical techniques such as Rate Monotonic Analysis (RMA) and also provides a rationale for dynamic modeling. Factors such as system architecture, processor utilization, bus architecture, queuing, etc. are well suited for analysis with a dynamic model. The significance of performance measures for a real-time system are discussed.

Kiser, Larry M.↗

Dense LU Factorization on Multicore Supercomputer Nodes

Dense LU factorization is a prominent benchmark used to rank the performance of supercomputers. Many implementations, including the reference code HPL, use block-cyclic distributions of matrix blocks onto a two-dimensional process grid. The process grid dimensions drive a trade-off between communication and computation and are architecture- and implementation-sensitive. We show how the critical panel factorization steps can be made less communication-bound by overlapping asynchronous collectives for pivot identification and exchange with the computation of rank-k updates. By shifting this trade-off, a modified block-cyclic distribution can beneficially exploit more available parallelism on the critical path, and reduce panel factorization's memory hierarchy contention on now-ubiquitous multi-core architectures. The missed parallelism in traditional block-cyclic distributions arises because active panel factorization, triangular solves, and subsequent broadcasts are spread over single process columns or rows (respectively) of the process grid. Increasing one dimension of the process grid decreases the number of distinct processes in the other dimension. To increase parallelism in both dimensions, periodic 'rotation' is applied to the process grid to recover the row-parallelism lost by a tall process grid. During active panel factorization, rank-1 updates stream through memory with minimal reuse. In a column-major process grid, the performance of this access pattern degrades as too many streaming processors contend for access to memory. A block-cyclic mapping in the more popular row-major order does not encounter this problem, but consequently sacrifices node and network locality in the critical pivoting steps. We introduce 'striding' to vary between the two extremes of row- and column-major process grids. As a test-bed for further mapping experiments, we describe a dense LU implementation that allows a block distribution to be defined as a general function of block to processor. Other mappings can be tested with only small, local changes to the code.

99 GENERAL AND MISCELLANEOUS↗

Suitability of Gray Water for Hydroponic Crop Production Following Biological and Physical Chemical and Biological Subsystems

The water present in waste streams from a human habitat must be recycled in Controlled Ecological Life Support Systems (CELSS) to limit resupply needs and attain self-sufficiency. Plants play an important role in providing food, regenerating air, and producing purified water via transpiration. However, we have shown that the surfactants present in hygiene waste water have acute toxic effects on plant growth (Bubenheim et al. 1994; Greene et al., 1994). These phytotoxic affects can be mitigated by allowing the microbial population on the root surface to degrade the surfactant, however, a significant suppression (several days) in crop performance is experienced prior to reaching sub-toxic surfactant levels and plant recovery. An effective alternative is to stabilize the microbial population responsible for degradation of the surfactant on an aerobic bioreactor and process the waste water prior to utilization in the hydroponic solution (Wisniewski and Bubenheim, 1993). A sensitive bioassay indicates that the surfactant phytotoxicity is suppressed by more than 90% within 5 hours of introduction of the gray water to the bioreactor; processing for more than 12 hours degrades more than 99% of the phytotoxin. Vapor Compression Distillation (VCD) is a physical / chemical method for water purification which employees sequential distillation steps to separate water from solids and to volatilize contaminants. The solids from the waste water are concentrated in a brine and the pure product water (70 - 90% of the total waste water volume depending on operating conditions) retains non of the phytotoxic effects. Results of the bioassay were used to guide evaluations of the suitability of recovered gray water following biological and VCD processing for hydroponic lettuce production in controlled environments. Lettuce crops were grown for 28 days with 100% of the input water supplied with recovered water from the biological processor or VCD. When compared with the growth of plants in control hydroponic solution containing pure deionized water, no growth difference could be measured resulting from any of the recovered water treatments. Both biological treatment and VCD offer alternative technology approaches to recovering water from waste streams appropriate for input into a crop production system. A high level of crop performance (food, air, and water production) can be maintained with either processor; selection decisions can be based on other factors regarding system integration.

Bubenheim, David L.↗

Mapping Dense LU Factorization on Multicore Supercomputer Nodes

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

Computer Science↗

Separable projection integrals for higher-order correlators of the cosmic microwave sky: Acceleration by factors exceeding 100

We present a case study describing efforts to optimise and modernise “Modal”, the simulation and analysis pipeline used by the Planck satellite experiment for constraining general non-Gaussian models of the early universe via the bispectrum (or three-point correlator) of the cosmic microwave background radiation. We focus on one particular element of the code: the projection of bispectra from the end of inflation to the spherical shell at decoupling, which defines the CMB we observe today. This code involves a three-dimensional inner product between two functions, one of which requires an integral, on a non-rectangular domain containing a sparse grid. We show that by employing separable methods this calculation can be reduced to a one-dimensional summation plus two integrations, reducing the overall dimensionality from four to three. The introduction of separable functions also solves the issue of the non-rectangular sparse grid. This separable method can become unstable in certain scenarios and so the slower non-separable integral must be calculated instead. We present a discussion of the optimisation of both approaches. We demonstrate significant speed-ups of ≈100×, arising from a combination of algorithmic improvements and architecture-aware optimisations targeted at improving thread and vectorisation behaviour. The resulting MPI/OpenMP hybrid code is capable of executing on clusters containing processors and/or coprocessors, with strong-scaling efficiency of 98.6% on up to 16 nodes. We find that a single coprocessor outperforms two processor sockets by a factor of 1.3× and that running the same code across a combination of both microarchitectures improves performance-per-node by a factor of 3.38×. By making bispectrum calculations competitive with those for the power spectrum (or two-point correlator) we are now able to consider joint analysis for cosmological science exploitation of new data.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗