Accurately modeling and predicting performance for large-scale applications becomes increasingly difficult as system complexity scales dramatically. Analytic predictive models are useful, but are difficult to construct, usually limited in scope, and often fail to capture subtle interactions between architecture and software. In contrast, we employ multilayer neural networks trained on input data from executions on the target platform. This approach is useful for predicting many aspects of performance, and it captures full system complexity. Our models are developed automatically from the training input set, avoiding the difficult and potentially error-prone process required to develop analytic models. This study focuses on the high-performance, parallel application SMG2000, a much studied code whose variations in execution times are still not well understood. Our model predicts performance on two large-scale parallel platforms within 5%-7% error across a large, multi-dimensional parameter space.

The NLC Test Accelerator (NLCTA) at SLAC was built to address various beam dynamics issues for the Next Linear Collider. An S-Band RF gun is being installed together with a large-angle extraction line at 60 MeV followed by a matching section, buncher and final focus for the laser acceleration experiment, E163. The laser-electron interaction area is followed by a broad range, high resolution spectrometer (HES) for electron bunch analysis. Another spectrometer at 6 MeV will be used for analysis of bunch charges up to 1 nC. Emittance compensating solenoids and the low energy spectrometer (LES) will be used to tune for best operating point and match to the linac. Optical symmetries in the design of the 25.5{sup o} extraction line provide 1:1 phase space transfer without use of sextupoles for a large, 6D phase space volume and range of input conditions. Design techniques, tolerances, tuning sensitivities and orthogonal knobs are discussed.

The Next and Global Linear Collider (NLC/GLC) designs require precision alignment of the beam in the accelerator structures to reduce short range wakefields. The moderately damped and detuned structures themselves provide suitable higher order mode (HOM) signals to measure this alignment. The modes in the lowest dipole band, whose frequencies range from 14-16 GHz, provide the strongest signals. To determine the position resolution they provide, an NLC/GLC prototype structure that was installed in the ASSET facility of the SLAC Linac was instrumented to downmix and digitize these signals. The beam position within the structure was determined by simultaneously measuring the signals at three frequencies (14.3, 15, 15.7 GHz) corresponding to modes localized at the beginning, the middle and the end of the 60 cm long structure. A resolution of 1 micron was achieved even with 28 dB signal attenuation, which is better than the 5 micron resolution required for the NLC/GLC.

We have used the unique spatial sensitivity of polarized neutron and soft x-ray beams in reflection geometry to measure the depth dependence of magnetization across the interface between a ferromagnet and antiferromagnet. The new uncompensated magnetization near the interface responds to applied field, while the uncompensated spins in the antiferromagnetic bulk are pinned, thus providing a means to establish exchange bias.

We report the observation of Debye-Scherrer diffraction using electron pulses emitted from a fs-laser plasma. Titanium sapphire laser pulses with 1.6 mJ/45 fs at 1 kHz are focused on a moving steel tape at close to normal incidence. The laser plasma generated ejects a large number of electrons in the direction of polarization, with a continuous energy spectrum extending up to 100 keV. Selecting an energy range of these electrons and scattering them on a thin aluminium sample generates a ''streaked'' diffraction pattern with unique features.

We present a three-dimensional modeling study of gas flow inthe unsaturated fractured rock of Yucca Mountain. Our objective is toestimate large-scale fracture permeability, using the changes insubsurface pneumatic pressure in response to barometric pressure changesat the land surface. We incorporate the field-measured pneumatic datainto a multiphase flow model for describing the coupled processes ofliquid and gas flow under ambient geothermal conditions. Comparison offield-measured pneumatic data with model-predicted gas pressures is foundto be a powerful technique for estimating the fracture permeability ofthe unsaturated fractured rock, which is otherwise extremely difficult todetermine on the large scales of interest. In addition, this studydemonstrates that the multi-dimensional-flow effect on estimatedpermeability values is significant and should be included whendetermining fracture permeability in heterogeneous fracturedmedia.

Many materials and electronics need to be tested for the radiation environment expected at linear colliders (LC) since both accelerator and detectors will be subjected to large fluences of hadrons, leptons and {gamma}'s over the life of the facility [1]. While the linacs will be superconducting, there are still many uses for NdFeB in the damping rings, injection and extraction lines and final focus. Our understanding of the situation for rare earth, permanent magnet materials was presented at PAC03 [2]. Our first measurements of fast neutron, stepped doses at the UC Davis McClellan Nuclear Reactor Center (UCD MNRC) were presented at EPAC04 [3]. We have extended the doses, included other manufacturer's samples, and measured induced radioactivities which are discussed in detail.

The Grid Collector is a system that facilitates the effective analysis and spontaneous exploration of scientific data. It combines an efficient indexing technology with a Grid file management technology to speed up common analysis jobs on high-energy physics data and to enable some previously impractical analysis jobs. To analyze a set of high-energy collision events, one typically specifies the files containing the events of interest, reads all the events in the files, and filters out unwanted ones. Since most analysis jobs filter out significant number of events, a considerable amount of time is wasted by reading the unwanted events. The Grid Collector removes this inefficiency by allowing users to specify more precisely what events are of interest and to read only the selected events. This speeds up most analysis jobs. In existing analysis frameworks, the responsibility of bringing files from tertiary storages or remote sites to local disks falls on the users. This forces most of analysis jobs to be performed at centralized computer facilities where commonly used files are kept on large shared file systems. The Grid Collector automates file management tasks and eliminates the labor-intensive manual file transfers. This makes it much easier to perform analyses that require …

During the past five years, there has been a concerted program at SLAC and KEK to develop accelerator structures that meet the high gradient (65 MV/m) performance requirements for the Next Linear Collider (NLC) and Global Linear Collider (GLC) initiatives. The design that resulted is a 60-cm-long, traveling-wave structure with low group velocity and 150 degree per cell phase advance. It has an average iris size that produces an acceptable short-range wakefield, and dipole mode damping and detuning that adequately suppresses the long-range wakefield. More than eight such structures have operated at a 60 Hz repetition rate over 1000 hours at 65 MV/m with 400 ns long pulses, and have reached breakdown rate levels below the limit for the linear collider. Moreover, the structures are robust in that the rates continue to decrease over time, and if the structures are briefly exposed to air, the rates recover to their low levels within a few days. This paper presents a summary of the results from this program, which effectively ended last August with the selection of ''cold'' technology for an International Linear Collider (ILC).

None

With the planned installation of a superconducting rf system, the new operation mode of TLS, the electron storage ring at NSRRC, is expected to double the beam intensity. Several accelerator physics topics need to be examined. Beam instability of single-bunch longitudinal microwave instability is one of these topics. We consider two approaches to measure the effective broad band impedance. We compare these measurement results with each other and to old data [Ref.1]. We calculate the threshold current of microwave instability with a mode-mixing analysis code written by Dr. K. Oide of KEK [Ref.2]. We also develop a multi-particle tracking code to simulate the instability. The results of simulation and measurement are compared and discussed. We conclude that doubling of beam current from 200 mA (1.5 mA/bunch) to 400 mA (3 mA/bunch) will not trigger the microwave instability even without a Landau cavity to lengthen the bunch. The benefit of Landau cavity is mainly for beam life time.

We present measurements of the effective vertical emittance and IP {beta}-function in the PEP-II asymmetric B Factory at SLAC. These beam parameters are extracted from fits to the longitudinal dependence of the luminosity and the vertical luminous size, measured using e{sup +}e{sup -} {yields} {mu}{sup +}{mu}{sup -} events recorded in the BABAR detector. The results are compared, for different sets of machine conditions, to accelerator-based measurements of the optical functions of the two beams.

Background studies during the design, construction, commissioning, operation and improvement of BaBar and PEP-II have been greatly influenced by results from a program referred to as LPTURTLE (Lost Particle TURTLE) which was originally conceived for the purpose of studying gas background for SLC. This venerable program is still in use today. We describe its use, capabilities and improvements and refer to current results now being applied to BaBar.

The analysis of orbit response matrices has been used very successfully to measure and correct the gradient and skew gradient distribution in many accelerators. It allows determination of an accurately calibrated model of the coupled machine lattice, which then can be used to calculate the corrections necessary to improve coupling, dynamic aperture and ultimately luminosity. At PEP-II, the Matlab version of LOCO has been used to analyze coupled response matrices for both the LER and the HER. The large number of elements in PEP-II and the very complicated interaction region present unique challenges to the data analysis. All necessary tools to make the analysis method useable at PEP-II have been implemented and LOCO can now be used as a routine tool for lattice diagnostic.

Seven large power supplies (LGPS) with output ratings from 72kW to 270kW power PEP-II quad magnets in the electron-positron collider region. These supplies have posed serious maintenance and reliability problems since they were installed in 1997, resulting in loss of accelerator availability. A redesign/rebuild program was undertaken by the SLAC Power Conversion Department. During the 2004 summer shutdown all the control circuits in these supplies were redesigned and replaced. A new PWM control board, programmable logic controller, and touch panel have been installed to improve LGPS reliability, and to make troubleshooting easier. In this paper we present the details of this rebuilding program and results.

None

We describe an experiment for production of high harmonic x-ray radiation from Thomson backscattering of an ultra-short high power density laser by a relativistic electron beam at the PLEIADES facility at LLNL. In this scenario, electrons execute a ''figure-8'' motion under the influence of the high-intensity laser field, where the constant characterizing the field strength is expected to exceed unity: a{sub L} = eE{sub L}/m{sub e}cw{sub L} {ge} 1. With large a{sub L} this motion produces high harmonic x-ray radiation and significant broadening of the spectral peaks. This paper is intended to give a layout of the PLEIADES experiment, along with progress towards experimental goals.

None

None

The measurement of the top quark pair production crosssection inproton-antiproton collisions at 1.96 TeV is a test ofquantumchromodynamics and could potentially be sensitive to newphysics beyondthe standard model. I report on the latest t-tbarcross section resultsfrom the CDF and DZero experiments in various finalstate topologies whicharise from decays of top quark pairs.