Engineering topics
Publications and source records attributed to Saini, Subhash.
An approximation to the Maxwell-Semiconductor Bloch equations is solved for the transverse modes of vertical-cavity surface-emitting lasers. A finite-difference algorithm will be used to calculate the detailed time evolution of the transverse mode profiles of VCSELs. Effects of pumping current and refractive index profiles will be investigated.
Vertical-Cavity Surface-Emitting Lasers (VCSELs) have become increasingly important for many applications, such as optical interconnects and optical data storages. In many applications, critical requirements are very often high beam quality and/or high output powers. To help the design and optimization of VCSELs for such applications, it is essential to have a simulation model that includes the time evolution of the transverse space dependence of the beam. Such models allow arbitrary transverse profiles of laser intensity or carrier density to develop in time without any a priori assumption about the number and types of transverse modes. In addition, extensive research in the past has shown that many-body effects are important in influencing the optical gain and the refractive index of semiconductor quantum wells. It is thus important to include the key many-body effects in simulation models for semiconductor lasers. A computational model, called the Maxwell-Effective Bloch Equations (MEBE) was developed recently that simulates the above features and effects. This model combines the space-time resolution with microscopic many-body effects in a way that is both accurate enough and computationally manageable. This model has been used to simulate edge-emitting high power lasers recently. A new finite-difference algorithm has been developed for solving these equations and will be shown in the presentation. That algorithm will be used to calculate the detailed time evolution of the transverse mode profiles of VCSELs. The VCSELs are based on InGaAs/GaAs quantum well structures. Effects of pumping current and refractive index profiles will be investigated in detail. The simulation results will be compared with other results the use the rate-equation model.
The Charon toolkit for piecemeal development of high-efficiency parallel programs for scientific computing is described. The portable toolkit, callable from C and Fortran, provides flexible domain decompositions and high-level distributed constructs for easy translation of serial legacy code or design to distributed environments. Gradual tuning can subsequently be applied to obtain high performance, possibly by using explicit message passing. Charon also features general structured communications that support stencil-based computations with complex recurrences. Through the separation of partitioning and distribution, the toolkit can also be used for blocking of uni-processor code, and for debugging of parallel algorithms on serial machines. An elaborate review of recent parallelization aids is presented to highlight the need for a toolkit like Charon. Some performance results of parallelizing the NAS Parallel Benchmark SP program using Charon are given, showing good scalability. Some performance results of parallelizing the NAS Parallel Benchmark SP program using Charon are given, showing good scalability.
A variety of complex carbon nanotube (CNT) structures have been observed experimentally. These include sharp bends, branches, tori, and helices. They are believed to be formed by using topological defects such as pentagons and heptagons to connect different CNT. The effects of type, number, and arrangement (separation and orientation) of defects on atomic structures and energetics of complex CNT are investigated using topology, quantum mechanics and molecular mechanics calculations. Energetically stable models are derived for identification of observed complex CNT structures.
A substrate for future atomic chain electronics, where adatoms are placed at designated positions and form atomically precise device components, is studied theoretically. The substrate has to serve as a two-dimensional template for adatom mounting with a reasonable confinement barrier and also provide electronic isolation, preventing unwanted coupling between independent adatom structures. For excellent structural stability, we demand chemical bonding between the adatoms and substrate atoms, but then good electronic isolation may not be guaranteed. Conditions are clarified for good isolation. Because of the chemical bonding, fundamental adatom properties are strongly influenced: a chain with group IV adatoms having two chemical bonds, or a chain with group III adatoms having one chemical bond is semiconducting. Charge transfer from or to the substrate atoms brings about unintentional doping, and the electronic properties have to be considered for the entire combination of the adatom and substrate systems even if the adatom modes are well localized at the surface.
Adatom chains, precise structures artificially created on an atomically regulated surface, are the smallest possible candidates for future nanoelectronics. Since all the devices are created by combining adatom chains precisely prepared with atomic precision, device characteristics are predictable, and free from deviations due to accidental structural defects. In this atomic dimension, however, an analogy to the current semiconductor devices may not work. For example, Si structures are not always semiconducting. Adatom states do not always localize at the substrate surface when adatoms form chemical bonds to the substrate atoms. Transport properties are often determined for the entire system of the chain and electrodes, and not for chains only. These fundamental issues are discussed, which will be useful for future device considerations.
A substrate for future atomic chain electronics, where adatoms are placed at designated positions and form atomically precise device components, is studied theoretically. The substrate has to serve as a two-dimensional template for adatom mounting with a reasonable confinement barrier and also provide electronic isolation, preventing unwanted coupling between independent adatom structures. However, the two requirements conflict. For excellent electronic isolation, we may seek adatom confinement via van der Waals interaction without chemical bonding to the substrate atoms, but the confinement turns out to be very weak and hence unsatisfactory. An alternative chemical bonding scheme with excellent structural strength is examined, but even fundamental adatom chain properties such as whether chains are semiconducting or metallic are strongly influenced by the nature of the chemical bonding, and electronic isolation is not always achieved. Conditions for obtaining semiconducting chains with well-localized surface-modes, leading to good isolation, are clarified and discussed.
When the device size is reduced down to 0.07 micrometers, the number of dopant atoms in the channel will no longer be macroscopic, typically less than a hundred. A spatial distribution of these dopant atoms will fluctuate statistically from device to device even in identically designed devices, and this places a serious limitation for integration. It is, however, impractical to control dopant positions within atomic dimension. One fundamental solution to this problem is to create electronics with atomically precise, but preferably simple structures. Atomic chains, precise structures of adatoms created on an atomically regulated surface, are candidates for constituent components in future electronics. All the adatoms will be placed at designated positions on the substrate, and all the device structures will be precise, free from any deviations. It was predicted using the tight-binding calculation with universal parameters that silicon chains were metallic and magnesium chains were semiconducting regardless of the lattice spacing, and a possible doping method was also proposed. In these treatments, the substrate was assumed to serve as a non-interacting template holding the adatoms without a formation of chemical bonding with substrate atoms. However, this scheme may not be easy to implement experimentally. Adatoms will have to be fixed with a van der Waals force on the substrate, but the force is generally weak and an extremely low temperature environment has to be prepared to suppress their unwanted thermal displacement. It may be logical to seek a scheme to allow the adatoms to form chemical bonding with the substrate atoms and secure their positions. The substrate effects are studied in detail.
Over the past decade, high performance computing has evolved rapidly; systems based on commodity microprocessors have been introduced in quick succession from at least seven vendors/families. Porting codes to every new architecture is a difficult problem; in particular, here at NASA, there are many large CFD applications that are very costly to port to new machines by hand. The LCM ("Legacy Code Modernization") Project is the development of an integrated parallelization environment (IPE) which performs the automated mapping of legacy CFD (Fortran) applications to state-of-the-art high performance computers. While most projects to port codes focus on the parallelization of the code, we consider porting to be an iterative process consisting of several steps: 1) code cleanup, 2) serial optimization,3) parallelization, 4) performance monitoring and visualization, 5) intelligent tools for automated tuning using performance prediction and 6) machine specific optimization. The approach for building this parallelization environment is to build the components for each of the steps simultaneously and then integrate them together. The demonstration will exhibit our latest research in building this environment: 1. Parallelizing tools and compiler evaluation. 2. Code cleanup and serial optimization using automated scripts 3. Development of a code generator for performance prediction 4. Automated partitioning 5. Automated insertion of directives. These demonstrations will exhibit the effectiveness of an automated approach for all the steps involved with porting and tuning a legacy code application for a new architecture.
With the resurgence of distributed shared memory (DSM) systems based on cache-coherent Non Uniform Memory Access (ccNUMA) architectures and increasing disparity between memory and processors speeds, data locality overheads are becoming the greatest bottlenecks in the way of realizing potential high performance of these systems. While parallelization tools and compilers facilitate the users in porting their sequential applications to a DSM system, a lot of time and effort is needed to tune the memory performance of these applications to achieve reasonable speedup. In this paper, we show that integrating cache performance modeling and tuning support within a parallelization environment can alleviate this problem. The Cache Performance Modeling and Prediction Tool (CPMP), employs trace-driven simulation techniques without the overhead of generating and managing detailed address traces. CPMP predicts the cache performance impact of source code level "what-if" modifications in a program to assist a user in the tuning process. CPMP is built on top of a customized version of the Computer Aided Parallelization Tools (CAPTools) environment. Finally, we demonstrate how CPMP can be applied to tune a real Computational Fluid Dynamics (CFD) application.
We present a cache performance modeling methodology that facilitates the tuning of uniprocessor cache performance for applications executing on shared memory multiprocessors by accurately predicting the effects of source code level modifications. Measurements on a single processor are initially used for identifying parts of code where cache utilization improvements may significantly impact the overall performance. Cache simulation based on trace-driven techniques can be carried out without gathering detailed address traces. Minimal runtime information for modeling cache performance of a selected code block includes: base virtual addresses of arrays, virtual addresses of variables, and loop bounds for that code block. Rest of the information is obtained from the source code. We show that the cache performance predictions are as reliable as those obtained through trace-driven simulations. This technique is particularly helpful to the exploration of various "what-if' scenarios regarding the cache performance impact for alternative code structures. We explain and validate this methodology using a simple matrix-matrix multiplication program. We then apply this methodology to predict and tune the cache performance of two realistic scientific applications taken from the Computational Fluid Dynamics (CFD) domain.
We describe an electronic transport model and an implementation approach that respond to the challenges of device modeling for gigascale integration. We use the density-gradient (DG) transport model, which adds tunneling and quantum smoothing of carrier density profiles to the drift-diffusion model. We present the current implementation of the DG model in PROPHET, a partial differential equation solver developed by Lucent Technologies. This implementation approach permits rapid development and enhancement of models, as well as run-time modifications and model switching. We show that even in typical bulk transport devices such as P-N diodes and BJTs, DG quantum effects can significantly modify the I-V characteristics. Quantum effects are shown to be even more significant in small, surface transport devices, such as sub-0.1 micron MOSFETs. In thin-oxide MOS capacitors, we find that quantum effects may reduce gate capacitance by 25% or more. The inclusion of quantum effects in simulations dramatically improves the match between C-V simulations and measurements. Significant quantum corrections also occur in the I-V characteristics of short-channel MOSFETs due to the gate capacitance correction.
We present an HPF implementation of BT, SP, LU, FT, CG and MG of NP132.3-serial benchmark set. The implementation is based on HPF performance model of the benchmark specific primitive operations with distributed arrays. We present profiling and performance data on SGI Origin 2000 and compare the results with NPB2.3. We discuss an HPF limitation in organizing pipelined computations which results in memory and communication overhead in BT and SP.
Carbon nanotubes (CNT) enable nanoelectromechanical systems (NEMS) because of their inherent nanostructure, intrinsic electric conductivity and mechanical resilience. The collaborative work between Stanford (experiment) and NASA Ames (theory and simulation) has made progress in two types of CNT based NEMS for nanoelectronics and sensor applications. The CNT tipped scanning probe microscopy (SPM) is a NEMS in which CNT tips are used for nanoscale probing, imaging and manipulating. It showed great improvement in probing surfaces and biological systems over conventional tips. We have recently applied it to write (lithography) and read (image) uniform SiO2 lines on large Si surface area at speed up to 0.5 mm per s. Preliminary work using approximately 10 nm multiwall nanotube tips produced approximately 10 nm structures and showed that the CNT tips didn't wear down when crashed as conventional tips often do. This presents a solution to the long standing tip-wear problem in SPM nanolithography. We have also explored potential of CNT tips in imaging DNA in water. Preliminary experiment using 10 nm CNT tips reached 5 nm resolution. The 1 nm nanolithography and 1 nm DNA imaging can be expected by using approximately 1 nm CNT tips. In contrast to CNT tipped SPM, we also fabricated CNT devices on silicon wafer in which CNTs connect patterned metallic lines on SiO2/Si by a simple chemical vapor deposition process. Using conventional lithography for silicon wafer, we have been able to obtain CNT based transistors and sensors. Investigations of the CNT NEMS as physical, biological and chemical sensors are in progress and will be discussed.
Modulation and switching of semiconductor lasers are important for laser-based information technology. Typically the speed of modulation and switching is limited by interband processes such as stimulated and spontaneous recombinations which occur on a nanosecond time scale. This is why the diode laser modulation has been restricted to tens of GHz. Modulation at higher speed is highly desirable as the information technology enters into the so-called tera-era. In this paper, we study the possibility of utilizing THz-field-induced plasma heating to modulate quantum-well lasers. This is a timely study since, with the advancement of THz solid-state sources and free-electron lasers, THz physics and related technology is currently coming out of its infancy. The investigation of interplaying THz and optical fields is also of intruiging fundamental interest. First, we introduce theoretical plasma heating results for the quantum-well optical amplifier in the presense of an intense half-cycle THz pulse. The heated carrier distributions are then utilized to calculate the THz-pulse-induced change in refractive index and gain profile. Since the electron-hole-plasma is heated using intraband transitions, we circumvent the usual complications due to an overall change in density, and the nonlinear recovery is governed solely by the carrier-LO-phonon interactions, typically 5 ps for a complete recovery. This procedure implies THz and sub-THz switching and recovery rates, respectively; using either gain modulation or index modulation. Plasma heating via steady-state THz fields is also studied. Finally, numerical simulation of a coupled set of equations to investigate the THz modulation based on a simplified model for quantum-well lasers is presented. Our results show that a semiconductor laser can be modulated at up to 1 THz with little distortion with a THz field amplitude at the order of a few kV/cm. Laser responses to a change in THz frequency will be shown. Constraints, practicalities, and applications will be discussed.
Atomic chains, precise structures of adatoms created on an atomically regulated surface, are candidates for constituent elements in future electronics. It was predicted that Si chains were metallic and Mg chains were semiconducting, and a doping method was also discussed. The substrate was assumed to work as a non-interacting template holding the adatoms. However, this scheme requires a low-temperature environment so that the adatoms will not displace from their ideal positions due to unwanted thermal agitation. For better structural stability, we may seek a scheme to allow the adatoms to form chemical bonding with the substrate atoms and secure their positions. The chemical bonding has two major effects on the chain electronic properties. First, only the remaining s and p orbitals in an adatom not used for the chemical bonding can decide chain band structures, rather than the full set of orbitals previously assumed. Second, because of a possible HOMO energy difference between adatom and substrate atom, semiconducting chains are unintentionally doped. These effects are studied with a self-consistent tight-binding method with universal parameters. With one adatom per unit cell, adatom chains are semiconducting: (1) if adatoms are of group III and form one chemical bond per adatom, or (2) if adatoms are of group IV and form two chemical bonds. 'Me previous result [I I of realizing semiconducting chains by group II adatoms without chemical bonds is consistent with this picture. When the HOMO energy in an adatom is shallower than that in a substrate atom, the entire chain is positively polarized, unintentionally achieving p-type doping. When deeper, the chain is negatively charged, achieving n-type doping.
The possibility of nanolithography of silicon and germanium surfaces with bare carbon nanotube tips of scanning probe microscopy devices is considered with large scale classical molecular dynamics (MD) simulations employing Tersoff's reactive many-body potential for heteroatomic C/Si/Ge system. Lithography plays a key role in semiconductor manufacturing, and it is expected that future molecular and quantum electronic devices will be fabricated with nanolithographic and nanodeposition techniques. Carbon nanotubes, rolled up sheets of graphene made of carbon, are excellent candidates for use in nanolithography because they are extremely strong along axial direction and yet extremely elastic along radial direction. In the simulations, the interaction of a carbon nanotube tip with silicon surfaces is explored in two regimes. In the first scenario, the nanotubes barely touch the surface, while in the second they are pushed into the surface to make "nano holes". The first - gentle scenario mimics the nanotube-surface chemical reaction induced by the vertical mechanical manipulation of the nanotube. The second -digging - scenario intends to study the indentation profiles. The following results are reported in the two cases. In the first regime, depending on the surface impact site, two major outcomes outcomes are the selective removal of either a single surface atom or a surface dimer off the silicon surface. In the second regime, the indentation of a silicon substrate by the nanotube is observed. Upon the nanotube withdrawal, several surface silicon atoms are adsorbed at the tip of the nanotube causing significant rearrangements of atoms comprising the surface layer of the silicon substrate. The results are explained in terms of relative strength of C-C, C-Si, and Si-Si bonds. The proposed method is very robust and does not require applied voltage between the nanotube tips and the surface. The implications of the reported controllable etching and hole-creating for nanolithography on silicon are discussed in detail.
About VI author/originator verification in form 1676: (1) There is no export controlled, confidential commercial information. (2) Regarding the patent, the technical field covered here is related to ARC-14246, 'Doping Method of Semiconducting Atomic Chains." Most results are covered by ARC-14246.