ALBERT

Description: In this paper, we propose a distributed approach for mapping a single large application to a heterogeneous grid environment. To minimize the execution time of the parallel application, we distribute the mapping overhead to the available nodes of the grid. This approach not only provides a fast mapping of tasks to resources but is also scalable. We adopt a hierarchical grid model and accomplish the job of mapping tasks to this topology using a scheduler tree. Results show that our three-phase algorithm provides high quality mappings, and is fast and scalable.

Description: The ability of computers to solve hitherto intractable problems and simulate complex processes using mathematical models makes them an indispensable part of modern science and engineering. Computer simulations of large-scale realistic applications usually require solving a set of non-linear partial differential equations (PDES) over a finite region. For example, one thrust area in the DOE Grand Challenge projects is to design future accelerators such as the SpaHation Neutron Source (SNS). Our colleagues at SLAC need to model complex RFQ cavities with large aspect ratios. Unstructured grids are currently used to resolve the small features in a large computational domain; dynamic mesh adaptation will be added in the future for additional efficiency. The PDEs for electromagnetics are discretized by the FEM method, which leads to a generalized eigenvalue problem Kx = AMx, where K and M are the stiffness and mass matrices, and are very sparse. In a typical cavity model, the number of degrees of freedom is about one million. For such large eigenproblems, direct solution techniques quickly reach the memory limits. Instead, the most widely-used methods are Krylov subspace methods, such as Lanczos or Jacobi-Davidson. In all the Krylov-based algorithms, sparse matrix-vector multiplication (SPMV) must be performed repeatedly. Therefore, the efficiency of SPMV usually determines the eigensolver speed. SPMV is also one of the most heavily used kernels in large-scale numerical simulations.

Description: Classical mesh partitioning algorithms were designed for rather static situations, and their straightforward application in a dynamical framework may lead to unsatisfactory results, e.g., excessive data migration among processors. Furthermore, special attention should be paid to their amenability to parallelization. In this paper, a novel parallel method for the dynamic partitioning of adaptive unstructured meshes is described. It is based on a linear representation of the mesh using self-avoiding walks.

Description: In this paper we study the performance of the Lustre file system using five scientific and engineering applications representative of NASA workload on large-scale supercomputing systems such as NASA s Pleiades. In order to facilitate the collection of Lustre performance metrics, we have developed a software tool that exports a wide variety of client and server-side metrics using SGI's Performance Co-Pilot (PCP), and generates a human readable report on key metrics at the end of a batch job. These performance metrics are (a) amount of data read and written, (b) number of files opened and closed, and (c) remote procedure call (RPC) size distribution (4 KB to 1024 KB, in powers of 2) for I/O operations. RPC size distribution measures the efficiency of the Lustre client and can pinpoint problems such as small write sizes, disk fragmentation, etc. These extracted statistics are useful in determining the I/O pattern of the application and can assist in identifying possible improvements for users applications. Information on the number of file operations enables a scientist to optimize the I/O performance of their applications. Amount of I/O data helps users choose the optimal stripe size and stripe count to enhance I/O performance. In this paper, we demonstrate the usefulness of this tool on Pleiades for five production quality NASA scientific and engineering applications. We compare the latency of read and write operations under Lustre to that with NFS by tracing system calls and signals. We also investigate the read and write policies and study the effect of page cache size on I/O operations. We examine the performance impact of Lustre stripe size and stripe count along with performance evaluation of file per process and single shared file accessed by all the processes for NASA workload using parameterized IOR benchmark.

Description: The high performance computing (HPC) community has shown tremendous interest in exploring cloud computing as it promises high potential. In this paper, we examine the feasibility, performance, and scalability of production quality scientific and engineering applications of interest to NASA on NASA's cloud computing platform, called Nebula, hosted at Ames Research Center. This work represents the comprehensive evaluation of Nebula using NUTTCP, HPCC, NPB, I/O, and MPI function benchmarks as well as four applications representative of the NASA HPC workload. Specifically, we compare Nebula performance on some of these benchmarks and applications to that of NASA s Pleiades supercomputer, a traditional HPC system. We also investigate the impact of virtIO and jumbo frames on interconnect performance. Overall results indicate that on Nebula (i) virtIO and jumbo frames improve network bandwidth by a factor of 5x, (ii) there is a significant virtualization layer overhead of about 10% to 25%, (iii) write performance is lower by a factor of 25x, (iv) latency for short MPI messages is very high, and (v) overall performance is 15% to 48% lower than that on Pleiades for NASA HPC applications. We also comment on the usability of the cloud platform.

Description: A compact, low-cost laser communications transceiver was prototyped for downlinking data at 10 Gb/s from Earth-orbiting spacecraft. The design can be implemented using flight-grade parts. With emphasis on simplicity, compactness, and light weight of the flight transceiver, the reduced-complexity design and development approach involves: 1. A high-bandwidth coarse wavelength division multiplexed (CWDM) (4 2.5 or 10-Gb/s data-rate) downlink transmitter. To simplify the system, emphasis is on the downlink. Optical uplink data rate is modest (due to existing and adequate RF uplink capability). 2. Highly simplified and compact 5-cm diameter clear aperture optics assembly is configured to single transmit and receive aperture laser signals. About 2 W of 4-channel multiplexed (1,540 to 1,555 nm) optically amplified laser power is coupled to the optical assembly through a fiber optic cable. It contains a highly compact, precision-pointing capability two-axis gimbal assembly to coarse point the optics assembly. A fast steering mirror, built into the optical path of the optical assembly, is used to remove residual pointing disturbances from the gimbal. Acquisition, pointing, and tracking are assisted by a beacon laser transmitted from the ground and received by the optical assembly, which will allow transmission of a laser beam. 3. Shifting the link burden to the ground by relying on direct detection optical receivers retrofitted to 1-m-diameter ground telescopes. 4. Favored mass and volume reduction over power-consumption reduction. The two major variables that are available include laser transmit power at either end of the link, and telescope aperture diameter at each end of the link. Increased laser power is traded for smaller-aperture diameters. 5. Use of commercially available spacequalified or qualifiable components with traceability to flight qualification (i.e., a flight-qualified version is commercially available). An example is use of Telecordia-qualified fiber optic communication components including active components (lasers, amplifiers, photodetectors) that, except for vacuum and radiation, meet most of the qualifications required for space. 6. Use of CWDM technique at the flight transmitter for operation at four channels (each at 2.5 Gb/s or a total of 10 Gb/s data rate). Applying this technique allows utilization of larger active area photodetectors at the ground station. This minimizes atmospheric scintillation/turbulence induced losses on the received beam at the ground terminal. 7. Use of forward-error-correction and deep-interleaver codes to minimize atmospheric turbulence effects on the downlink beam. Target mass and power consumption for the flight data transmitter system is less than 10 kg and approximately 60 W for the 400-km orbit (900-km slant range), and 12 kg and 120 W for the 2,000-km orbit (6,000-km slant range). The higher mass and power for the latter are the result of employing a higher-power laser only.

Description: From its bold start nearly 30 years ago and continuing today, the NASA Advanced Supercomputing (NAS) facility at Ames Research Center has enabled remarkable breakthroughs in the space agency s science and engineering missions. Throughout this time, NAS experts have influenced the state-of-the-art in high-performance computing (HPC) and related technologies such as scientific visualization, system benchmarking, batch scheduling, and grid environments. We highlight the pioneering achievements and innovations originating from and made possible by NAS resources and know-how, from early supercomputing environment design and software development, to long-term simulation and analyses critical to design safe Space Shuttle operations and associated spinoff technologies, to the highly successful Kepler Mission s discovery of new planets now capturing the world s imagination.

Description: An optical link from Earth to an aircraft demonstrates the ability to establish a link from a ground platform to a transceiver moving overhead. An airplane has a challenging disturbance environment including airframe vibrations and occasional abrupt changes in attitude during flight. These disturbances make it difficult to maintain pointing lock in an optical transceiver in an airplane. Acquisition can also be challenging. In the case of the aircraft link, the ground station initially has no precise knowledge of the aircraft s location. An airborne pointing system has been designed, built, and demonstrated using direct-drive brushless DC motors for passive isolation of pointing disturbances and for high-bandwidth control feedback. The airborne transceiver uses a GPS-INS system to determine the aircraft s position and attitude, and to then illuminate the ground station initially for acquisition. The ground transceiver participates in link-pointing acquisition by first using a wide-field camera to detect initial illumination from the airborne beacon, and to perform coarse pointing. It then transfers control to a high-precision pointing detector. Using this scheme, live video was successfully streamed from the ground to the aircraft at 270 Mb/s while simultaneously downlinking a 50 kb/s data stream from the aircraft to the ground.

Description: High-End Computing (HEC) has always played a major role in meeting the modeling and simulation needs of various NASA missions. With NASA's newest 62 teraflops Columbia supercomputer, HEC is having an even greater impact within the Agency and beyond. Significant cutting-edge science and engineering simulations in the areas of space exploration, Shuttle operations, Earth sciences, and aeronautics research, are already occurring on Columbia, demonstrating its ability to accelerate NASA s exploration vision. The talk will describe how the integrated supercomputing production environment is being used to reduce design cycle time, accelerate scientific discovery, conduct parametric analysis of multiple scenarios, and enhance safety during the life cycle of NASA missions.

Description: Understanding the interplay between machines and problems is key to obtaining high performance on parallel machines. This paper investigates the interplay between programming paradigms and communication capabilities of parallel machines. In particular, we explicate the communication capabilities of the IBM SP-2 distributed-memory multiprocessor and the SGI PowerCHALLENGEarray symmetric multiprocessor. Two benchmark problems of bitonic sorting and Fast Fourier Transform are selected for experiments. Communication-efficient algorithms are developed to exploit the overlapping capabilities of the machines. Programs are written in Message-Passing Interface for portability and identical codes are used for both machines. Various data sizes and message sizes are used to test the machines' communication capabilities. Experimental results indicate that the communication performance of the multiprocessors are consistent with the size of messages. The SP-2 is sensitive to message size but yields a much higher communication overlapping because of the communication co-processor. The PowerCHALLENGEarray is not highly sensitive to message size and yields a low communication overlapping. Bitonic sorting yields lower performance compared to FFT due to a smaller computation-to-communication ratio.