ALBERT

Description: This paper provides an overview of the Mars Laser Communications Demonstration Project, a joint project between NASA s Goddard Space Flight Center (GSFC), the Jet Propulsion Laboratory, California Institute of Technology (JPL), and the Massachusetts Institute of Technology Lincoln Laboratory (MIT/LL). It reviews the strawman designs for the flight and ground segments, the critical technologies required, and the concept of operations. It reports preliminary conclusions from the Mars Lasercom Study conducted at MIT/LL and on additional work done at JPL and GSFC. The lasercom flight terminal will be flown on the Mars Telecom Orbiter (MTO) to be launched by NASA in 2009, and will demonstrate a technology which has the potential of vastly improving NASA s ability to communicate throughout the solar system.

Description: The Conjugate Gradient (CG) algorithm is perhaps the best-known iterative technique to solve sparse linear systems that are symmetric and positive definite. A sparse matrix-vector multiply (SPMV) usually accounts for most of the floating-point operations within a CG iteration. In this paper, we investigate the effects of various ordering and partitioning strategies on the performance of parallel CG and SPMV using different programming paradigms and architectures. Results show that for this class of applications, ordering significantly improves overall performance, that cache reuse may be more important than reducing communication, and that it is possible to achieve message passing performance using shared memory constructs through careful data ordering and distribution. However, a multi-threaded implementation of CG on the Tera MTA does not require special ordering or partitioning to obtain high efficiency and scalability.

Description: Tlis paper presents an overview of the preliminary design of both the flight and ground systems of the Optical Communication Demonstration and High-Rate Link Facility which will demonstrate optical communication from the International Space Station to ground after its deployment in October 2002. The overview of the preliminary design of the Flight System proceeds by contrasting it with the design of the laboratory-model unit, emphasizing key changes and the rationale behind the design choices. After presenting the preliminary design of the Ground System, the timetable for the construction and deployment of the flight and ground systems is outlined.

Description: Mountain-top to mountain-top optical link experiments have been initiated at JPL, in order to perform a systems level evaluation of optical communications. Progress made so far is reported. ne NASA, JPL developed optical communications demonstrator (OCD) is used to transmit a laser signal from Strawberry Peak (SP), located in the San Bernadino mountains of California. This laser beam is received by a 0.6 m aperture telescope at JPL's Table Mountain Facility (TMF), located in Wrightwood, California. The optical link is bi-directional with the TMF telescope transmitting a continuous 4-wave (cw) 780 run beacon and the OCD sending back a 840 nm, 100 - 500 Mbps pseudo noise (PN) modulated, laser beam. The optical link path is at an average altitude of 2 km above sea level, covers a range of 46.8 km and provides an atmospheric channel equivalent to approx. 4 air masses. Average received power measured at either end fall well within the uncertainties predicted by link analysis. The reduction in normalized intensity variance (sigma(sup 2, sub I)) for the 4-beam beacon, compared to each individual beam, at SP, was from approx. 0.68 to 0.22. With some allowance for intra-beam mis-alignment, this is consistent with incoherent averaging. The sigma(sup2, sub I) measured at TMF approx. 0.43 +/- 0.22 exceeded the expected aperture averaged value of less than 0.1, probably because of beam wander. The focused spot sizes of approx. 162 +/- 6 microns at the TMF Coude and approx. 64 +/- 3 microns on the OCD compare to the predicted size range of 52 - 172 microns and 57 - 93 microns, respectively. This is consistent with 4 - 5 arcsec of atmospheric "seeing". The preliminary evaluation of OCD's fine tracking indicates that the uncompensated tracking error is approx. 3.3 micro rad compared to approx. 1.7 micro rad observed in the laboratory. Fine tracking performance was intermittent, primarily due to beacon fades on the OCD tracking sensor. The best bit error rates observed while tracking worked were 1E-5 to 1E-6.

Description: The growing gap between sustained and peak performance for scientific applications has become a well-known problem in high performance computing. The recent development of parallel vector systems offers the potential to bridge this gap for a significant number of computational science codes and deliver a substantial increase in computing capabilities. This paper examines the intranode performance of the NEC SX6 vector processor and the cache-based IBM Power3/4 superscalar architectures across a number of key scientific computing areas. First, we present the performance of a microbenchmark suite that examines a full spectrum of low-level machine characteristics. Next, we study the behavior of the NAS Parallel Benchmarks using some simple optimizations. Finally, we evaluate the perfor- mance of several numerical codes from key scientific computing domains. Overall results demonstrate that the SX6 achieves high performance on a large fraction of our application suite and in many cases significantly outperforms the RISC-based architectures. However, certain classes of applications are not easily amenable to vectorization and would likely require extensive reengineering of both algorithm and implementation to utilize the SX6 effectively.

Description: The creation of parameter study suites has recently become a more challenging problem as the parameter studies have become multi-tiered and the computational environment has become a supercomputer grid. The parameter spaces are vast, the individual problem sizes are getting larger, and researchers are seeking to combine several successive stages of parameterization and computation. Simultaneously, grid-based computing offers immense resource opportunities but at the expense of great difficulty of use. We present ILab, an advanced graphical user interface approach to this problem. Our novel strategy stresses intuitive visual design tools for parameter study creation and complex process specification, and also offers programming-free access to grid-based supercomputer resources and process automation.

Description: An important characteristic of distributed grids is that they allow geographically separated multicomputers to be tied together in a transparent virtual environment to solve large-scale computational problems. However, many of these applications require effective runtime load balancing for the resulting solutions to be viable. Recently, we developed a latency tolerant partitioner, called MinEX, specifically for use in distributed grid environments. This paper compares the performance of MinEX to that of METIS, a popular multilevel family of partitioners, using simulated heterogeneous grid configurations. A solver for the classical N-body problem is implemented to provide a framework for the comparisons. Experimental results show that MinEX provides superior quality partitions while being competitive to METIS in speed of execution.