Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Vectorization”

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 181 records · Page 10

Matrix converter operation in current control mode with regulation of output voltage vector

A matrix converter system having a current control mode operation is provided. The matrix converter system includes a matrix converter that includes a plurality of switches. A unity current reference vector having an amplitude of one is determined that defines an angle and frequency of an output current vector as a function of a reference phase angle, which is a function of the output current vector and an output voltage vector representing feedback signals of a multiphase signal output to a load. The unity current reference vector is used to control the matrix converter, to cause the output current vector to be aligned with the unity current reference vector. A unity voltage reference vector having an amplitude of one is provided that defines a reference angle and frequency as a function of a zero phase angle. The output voltage vector and the unity voltage reference vector are aligned to determine the reference phase angle as an angle between the aligned output voltage vector and the output current vector.

42 ENGINEERING↗

Puromycin-resistant lentiviral control shRNA vector, pLKO.1 induces unexpected cellular differentiation of P19 embryonic stem cells

RNA silencing is used as a common method for investigating loss-of-function effects of genes of interest. In mammalian cells, RNA interference (RNAi) or RNA silencing can be achieved by transient siRNA (small or short interfering RNA) transfection or by stable shRNA (short hairpin RNA) systems. Various vectors are used for efficient delivery of shRNA. Lentiviral vectors offer an efficient delivery system for stable and long-term expression of the shRNA in mammalian cells. The widely used lentiviral pLKO.1 plasmid vector is very popular in RNAi studies. A large RNAi database, a TRC (the RNAi Consortium) library, was established based on the pLKO.1-TRC plasmid vector. This plasmid (also called pLKO.1-puro) has a puromycin-resistant gene for selection in mammalian cells along with designs for generating lentiviral particles as well for RNA silencing. While using the pLKO.1-puro TRC control shRNA plasmid for transfection in murine P19 embryonic stem (ES) cells, it was unexpectedly discovered that this plasmid vector induced robust endodermal differentiation. Since P19 ES cells are pluripotent and respond to external stimuli that have the potential to alter the phenotype and thus its stemness, other cell types used in RNA silencing studies do not display the obvious effect and therefore, may affect experiments in subtle ways that would go undetected. This study for the first time provides evidence that raises concern and warrants extreme caution while using the pLKO.1-puro control shRNA vector because of its unexpected non-specific effects on cellular integrity. - Highlights: • In P19 ES cells the pLKO.1-puro lentiviral control shRNA vector induced endodermal differentiation. • P19 ES cells harboring the pCDNA3 plasmid vector retained their stem-ness as opposed to those harboring the pLKO.1-puro vector. • P19 ES cells can serve as a sensor to determine vector safety. • Extreme caution is warranted while using the widely used pLKO.1-puro lentiviral vector for experimental and therapeutic designs.

60 APPLIED LIFE SCIENCES↗

Constraining primordial vector mode from B-mode polarization

The B-mode polarization spectrum of the Cosmic Microwave Background (CMB) may be the smoking gun of not only the primordial tensor mode but also of the primordial vector mode. If there exist nonzero vector-mode metric perturbations in the early Universe, they are known to be supported by anisotropic stress fluctuations of free-streaming particles such as neutrinos, and to create characteristic signatures on both the CMB temperature, E-mode, and B-mode polarization anisotropies. We place constraints on the properties of the primordial vector mode characterized by the vector-to-scalar ratio r{sub v} and the spectral index n{sub v} of the vector-shear power spectrum, from the Planck and BICEP2 B-mode data. We find that, for scale-invariant initial spectra, the ΛCDM model including the vector mode fits the data better than the model including the tensor mode. The difference in χ{sup 2} between the vector and tensor models is Δχ{sup 2} = 3.294, because, on large scales the vector mode generates smaller temperature fluctuations than the tensor mode, which is preferred for the data. In contrast, the tensor mode can fit the data set equally well if we allow a significantly blue-tilted spectrum. We find that the best-fitting tensor mode has a large blue tilt and leads to an indistinct reionization bump on larger angular scales. The slightly red-tilted vector mode supported by the current data set can also create O(10{sup -22})-Gauss magnetic fields at cosmological recombination. Our constraints should motivate research that considers models of the early Universe that involve the vector mode.

79 ASTRONOMY AND ASTROPHYSICS↗

An Efficient Wait-Free Vector

The vector is a fundamental data structure, which provides constant-time access to a dynamically-resizable range of elements. Currently, there exist no wait-free vectors. The only non-blocking version supports only a subset of the sequential vector API and exhibits significant synchronization overhead caused by supporting opposing operations. Since many applications operate in phases of execution, wherein each phase only a subset of operations are used, this overhead is unnecessary for the majority of the application. To address the limitations of the non-blocking version, we present a new design that is wait-free, supports more of the operations provided by the sequential vector, and provides alternative implementations of key operations. These alternatives allow the developer to balance the performance and functionality of the vector as requirements change throughout execution. Compared to the known non-blocking version and the concurrent vector found in Intel’s TBB library, our design outperforms or provides comparable performance in the majority of tested scenarios. Over all tested scenarios, the presented design performs an average of 4.97 times more operations per second than the non-blocking vector and 1.54 more than the TBB vector. In a scenario designed to simulate the filling of a vector, performance improvement increases to 13.38 and 1.16 times. This work presents the first ABA-free non-blocking vector. Finally, unlike the other non-blocking approach, all operations are wait-free and bounds-checked and elements are stored contiguously in memory.

97 MATHEMATICS AND COMPUTING↗

Index Sets and Vectorization

Vectorization is data parallelism (SIMD, SIMT, etc.) - extension of ISA enabling the same instruction to be performed on multiple data items simultaeously. Many/most CPUs support vectorization in some form. Vectorization is difficult to enable, but can yield large efficiency gains. Extra programmer effort is required because: (1) not all algorithms can be vectorized (regular algorithm structure and fine-grain parallelism must be used); (2) most CPUs have data alignment restrictions for load/store operations (obey or risk incorrect code); (3) special directives are often needed to enable vectorization; and (4) vector instructions are architecture-specific. Vectorization is the best way to optimize for power and performance due to reduced clock cycles. When data is organized properly, a vector load instruction (i.e. movaps) can replace 'normal' load instructions (i.e. movsd). Vector operations can potentially have a smaller footprint in the instruction cache when fewer instructions need to be executed. Hybrid index sets insulate users from architecture specific details. We have applied hybrid index sets to achieve optimal vectorization. We can extend this concept to handle other programming models.

97 MATHEMATICS AND COMPUTING↗

Optimum instantaneous impulsive orbital injection to attain a specified asymptotic velocity vector.

A nalysis of the necessary conditions of Battin for instantaneous orbital injection, with consideration of the uniqueness of his solution, and of the further problem which arises in the degenerate case when radius vector and asymptotic vector are separated by 180 deg. It is shown that when the angular separation between radius vector and asymptotic velocity vector satisfies theta not equal to 180 deg, there are precisely two insertion-velocity vectors which permit attainment of the target asymptotic velocity vector, one yielding posigrade, the other retrograde motion. When theta equals to 180 deg, there is a family of insertion-velocity vectors which permit attainment of a specified asymptotic velocity vector with a unique insertion-velocity vector for every arbitrary orientation of a target unit angular momentum vector.

Bean, W. C.↗

Static performance of nonaxisymmetric nozzles with yaw thrust-vectoring vanes

A static test was conducted in the static test facility of the Langley 16 ft Transonic Tunnel to evaluate the effects of post exit vane vectoring on nonaxisymmetric nozzles. Three baseline nozzles were tested: an unvectored two dimensional convergent nozzle, an unvectored two dimensional convergent-divergent nozzle, and a pitch vectored two dimensional convergent-divergent nozzle. Each nozzle geometry was tested with 3 exit aspect ratios (exit width divided by exit height) of 1.5, 2.5 and 4.0. Two post exit yaw vanes were externally mounted on the nozzle sidewalls at the nozzle exit to generate yaw thrust vectoring. Vane deflection angle (0, -20 and -30 deg), vane planform and vane curvature were varied during the test. Results indicate that the post exit vane concept produced resultant yaw vector angles which were always smaller than the geometric yaw vector angle. Losses in resultant thrust ratio increased with the magnitude of resultant yaw vector angle. The widest post exit vane produced the largest degree of flow turning, but vane curvature had little effect on thrust vectoring. Pitch vectoring was independent of yaw vectoring, indicating that multiaxis thrust vectoring is feasible for the nozzle concepts tested.

Mason, Mary L.↗

The Design of a Templated C++ Small Vector Class for Numerical Computing

We describe the design and implementation of a templated C++ class for vectors. The vector class is templated both for vector length and vector component type; the vector length is fixed at template instantiation time. The vector implementation is such that for a vector of N components of type T, the total number of bytes required by the vector is equal to N * size of (T), where size of is the built-in C operator. The property of having a size no bigger than that required by the components themselves is key in many numerical computing applications, where one may allocate very large arrays of small, fixed-length vectors. In addition to the design trade-offs motivating our fixed-length vector design choice, we review some of the C++ template features essential to an efficient, succinct implementation. In particular, we highlight some of the standard C++ features, such as partial template specialization, that are not supported by all compilers currently. This report provides an inventory listing the relevant support currently provided by some key compilers, as well as test code one can use to verify compiler capabilities.

Moran, Patrick J.↗

Experimental Study of an Axisymmetric Dual Throat Fluidic Thrust Vectoring Nozzle for Supersonic Aircraft Application

An axisymmetric version of the Dual Throat Nozzle concept with a variable expansion ratio has been studied to determine the impacts on thrust vectoring and nozzle performance. The nozzle design, applicable to a supersonic aircraft, was guided using the unsteady Reynolds-averaged Navier-Stokes computational fluid dynamics code, PAB3D. The axisymmetric Dual Throat Nozzle concept was tested statically in the Jet Exit Test Facility at the NASA Langley Research Center. The nozzle geometric design variables included circumferential span of injection, cavity length, cavity convergence angle, and nozzle expansion ratio for conditions corresponding to take-off and landing, mid climb and cruise. Internal nozzle performance and thrust vectoring performance was determined for nozzle pressure ratios up to 10 with secondary injection rates up to 10 percent of the primary flow rate. The 60 degree span of injection generally performed better than the 90 degree span of injection using an equivalent injection area and number of holes, in agreement with computational results. For injection rates less than 7 percent, thrust vector angle for the 60 degree span of injection was 1.5 to 2 degrees higher than the 90 degree span of injection. Decreasing cavity length improved thrust ratio and discharge coefficient, but decreased thrust vector angle and thrust vectoring efficiency. Increasing cavity convergence angle from 20 to 30 degrees increased thrust vector angle by 1 degree over the range of injection rates tested, but adversely affected system thrust ratio and discharge coefficient. The dual throat nozzle concept generated the best thrust vectoring performance with an expansion ratio of 1.0 (a cavity in between two equal minimum areas). The variable expansion ratio geometry did not provide the expected improvements in discharge coefficient and system thrust ratio throughout the flight envelope of typical a supersonic aircraft. At mid-climb and cruise conditions, the variable geometry design compromised thrust vector angle achieved, but some thrust vector control would be available, potentially for aircraft trim. The fixed area, expansion ratio of 1.0, Dual Throat Nozzle provided the best overall compromise for thrust vectoring and nozzle internal performance over the range of NPR tested compared to the variable geometry Dual Throat Nozzle.

Flamm, Jeffrey D.↗

Cosmological evolution in vector-tensor theories of gravity

We present a detailed study of the cosmological evolution in general vector-tensor theories of gravity without potential terms. We consider the evolution of the vector field throughout the expansion history of the Universe and carry out a classification of models according to the behavior of the vector field in each cosmological epoch. We also analyze the case in which the Universe is dominated by the vector field, performing a complete analysis of the system phase map and identifying those attracting solutions which give rise to accelerated expansion. Moreover, we consider the evolution in a universe filled with a pressureless fluid in addition to the vector field and study the existence of attractors in which we can have a transition from matter domination to vector domination with accelerated expansion so that the vector field may play the role of dark energy. We find that the existence of solutions with late-time accelerated expansion is a generic prediction of vector-tensor theories and that such solutions typically lead to the presence of future singularities. Finally, limits from local gravity tests are used to get constraints on the value of the vector field at small (Solar System) scales.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Adenovirus vector infection of non-small-cell lung cancer cells is a trigger for multi-drug resistance mediated by P-glycoprotein

P-glycoprotein (P-gp) is an ATP-binding cassette protein involved in cancer multi-drug resistance (MDR). It has been reported that infection with some bacteria and viruses induces changes in the activities of various drug-metabolizing enzymes and transporters, including P-gp. Although human adenoviruses (Ad) cause the common cold, the effect of Ad infection on MDR in cancer has not been established. In this study, we investigated whether Ad infection is a cause of MDR in A549, H441 and HCC827 non-small-cell lung cancer (NSCLC) cell lines, using an Ad vector system. We found that Ad vector infection of NSCLC cell lines induced P-gp mRNA expression, and the extent of induction was dependent on the number of Ad vector virus particles and the infection time. Heat-treated Ad vector, which is not infectious, did not alter P-gp mRNA expression. Uptake experiments with doxorubicin (DOX), a P-gp substrate, revealed that DOX accumulation was significantly decreased in Ad vector-infected A549 cells. The decrease of DOX uptake was blocked by verapamil, a P-gp inhibitor. Our results indicated that Ad vector infection of NSCLC cells caused MDR mediated by P-gp overexpression. The Ad vector genome sequence is similar to that of human Ad, and therefore human Ad infection of lung cancer patients may lead to chemoresistance in the clinical environment. -- Highlights: •Adenovirus vector infection induced P-gp mRNA expression in three NSCLC cell lines. •Adenovirus vector infection enhanced P-gp-mediated doxorubicin efflux from the cells. •The increase of P-gp was not mediated by nuclear receptors (PXR, CAR) or COX-2.

60 APPLIED LIFE SCIENCES↗

AAVPG: A vigilant vector where transgene expression is induced by p53

Using p53 to drive transgene expression from viral vectors may provide on demand expression in response to physiologic stress, such as hypoxia or DNA damage. Here we introduce AAVPG, an adeno-associated viral (AAV) vector where a p53-responsive promoter, termed PG, is used to control transgene expression. In vitro assays show that expression from the AAVPG-luc vector was induced specifically in the presence of functional p53 (1038±202 fold increase, p<0.001). The AAVPG-luc vector was an effective biosensor of p53 activation in response to hypoxia (4.48±0.6 fold increase in the presence of 250 µM CoCl{sub 2}, p<0.001) and biomechanical stress (2.53±0.4 fold increase with stretching, p<0.05). In vivo, the vigilant nature of the AAVPG-luc vector was revealed after treatment of tumor-bearing mice with doxorubicin (pre-treatment, 3.4×10{sup 5}±0.43×10{sup 5} photons/s; post-treatment, 6.6×10{sup 5}±2.1×10{sup 5} photons/s, p<0.05). These results indicate that the AAVPG vector is an interesting option for detecting p53 activity both in vitro and in vivo. - Highlights: • AAV vector where transgene expression is controlled by the tumor suppressor p53. • The new vector, AAVPG, shown to function as a biosensor of p53 activity, in vitro and in vivo. • The p53 activity monitored by the AAVPG vector is relevant to cancer and other diseases. • AAVPG reporter gene expression was activated upon DNA damage, hypoxia and mechanical stress.

60 APPLIED LIFE SCIENCES↗

Self-complementary and tyrosine-mutant rAAV vectors enhance transduction in cystic fibrosis bronchial epithelial cells

Recombinant adeno-associated virus (rAAV) vector platforms have shown considerable therapeutic success in gene therapy for inherited disorders. In cystic fibrosis (CF), administration of first-generation rAAV2 was safe, but clinical benefits were not clearly demonstrated. Therefore, next-generation vectors that overcome rate-limiting steps in rAAV transduction are needed to obtain successful gene therapy for this devastating disease. In this study, we evaluated the effects of single-strand or self-complementary (sc) rAAV vectors containing single or multiple tyrosine-to-phenylalanine (Y-F) mutations in capsid surface-exposed residues on serotypes 2, 8 or 9. For this purpose, CF bronchial epithelial (CFBE) cells were transduced with rAAV vectors, and the transgene expression of enhanced green fluorescence protein (eGFP) was analyzed at different time points. The effects of vectors on the cell viability, host cell cycle and in association with co-adjuvant drugs that modulate intracellular vector trafficking were also investigated. Six rAAV vectors demonstrated greater percentage of eGFP{sup +} cells compared to their counterparts at days 4, 7 and 10 post-transduction: rAAV2 Y(272,444,500,730)F, with 1.95-, 3.5- and 3.06-fold increases; rAAV2 Y(252,272,444,500,704,730)F, with 1.65-, 2.12-, and 2-fold increases; scrAAV2 WT, with 1.69-, 2.68-, and 2.32-fold increases; scrAAV8 Y773F, with 57-, 6.06-, and 7-fold increases; scrAAV9 WT, with 7.47-, 4.64-, and 3.66-fold increases; and scrAAV9 Y446F, with 8.39-, 4.62-, and 4.4-fold increases. At days 15, 20, and 30 post-transduction, these vectors still demonstrated higher transgene expression than transfected cells. Although the percentage of eGFP{sup +} cells reduced during the time-course analysis, the delta mean fluorescence intensity increased. These vectors also led to increased percentage of cells in G{sub 1}-phase without eliciting any cytotoxicity. Prior administration of bortezomib or genistein did not increase eGFP expression in cells transduced with either rAAV2 Y(272,444,500,730)F or rAAV2 Y(252,272,444,500,704,730)F. In conclusion, self-complementary and tyrosine capsid mutations on rAAV serotypes 2, 8, and 9 led to more efficient transduction than their counterparts in CFBE cells by overcoming the intracellular trafficking and second-strand DNA synthesis limitations.

60 APPLIED LIFE SCIENCES↗

Real space iterative reconstruction for vector tomography (RESIRE-V)

Tomography has had an important impact on the physical, biological, and medical sciences. To date, most tomographic applications have been focused on 3D scalar reconstructions. However, in some crucial applications, vector tomography is required to reconstruct 3D vector fields such as the electric and magnetic fields. Over the years, several vector tomography methods have been developed. Here, we present the mathematical foundation and algorithmic implementation of REal Space Iterative REconstruction for Vector tomography, termed RESIRE-V. RESIRE-V uses multiple tilt series of projections and iterates between the projections and a 3D reconstruction. Each iteration consists of a forward step using the Radon transform and a backward step using its transpose, then updates the object via gradient descent. Incorporating with a 3D support constraint, the algorithm iteratively minimizes an error metric, defined as the difference between the measured and calculated projections. The algorithm can also be used to refine the tilt angles and further improve the 3D reconstruction. To validate RESIRE-V, we first apply it to a simulated data set of the 3D magnetization vector field, consisting of two orthogonal tilt series, each with a missing wedge. Our quantitative analysis shows that the three components of the reconstructed magnetization vector field agree well with the ground-truth counterparts. We then use RESIRE-V to reconstruct the 3D magnetization vector field of a ferromagnetic meta-lattice consisting of three tilt series. Our 3D vector reconstruction reveals the existence of topological magnetic defects with positive and negative charges. We expect that RESIRE-V can be incorporated into different imaging modalities as a general vector tomography method. To make the algorithm accessible to a broad user community, we have made our RESIRE-V MATLAB source codes and the data freely available at https://github.com/minhpham0309/RESIRE-V.

47 OTHER INSTRUMENTATION↗

Recent progress with the top to bottom approach to vectorization in GeantV

SIMD acceleration can potentially boost by factors the application throughput. Achieving efficient SIMD vectorization for scalar code with complex data flow and branching logic, goes however way beyond breaking some loop dependencies and relying on the compiler. Since the refactoring effort scales with the number of lines of code, it is important to understand what kind of performance gains can be expected in such complex cases. We started to investigate a couple of years ago a top to bottom vectorization approach to particle transport simulation. Percolating vector data to algorithms was mandatory since not all the components can internally vectorize. Vectorizing low-level algorithms is certainly necessary, but not sufficient to achieve relevant SIMD gains. In addition, the overheads for maintaining the concurrent vector data flow and copy data have to be minimized. In the context of a vectorization R&D for simulation we developed a framework to allow different categories of scalar and vectorized components to co-exist, dealing with data flow management and real-time heuristic optimizations. The paper describes our approach on coordinating SIMD vectorization at framework level, making a detailed quantitative analysis of the SIMD gain versus overheads, with a breakdown by components in terms of geometry, physics and magnetic field propagation. We also present the more general context of this R&D work and goals for 2018.

97 MATHEMATICS AND COMPUTING↗

Multiaxis control power from thrust vectoring for a supersonic fighter aircraft model at Mach 0.20 to 2.47

The aeropropulsive characteristics of an advanced twin-engine fighter aircraft designed for supersonic cruise have been studied in the Langley 16-Foot Tansonic Tunnel and the Lewis 10- by 10-Foot Supersonic Tunnel. The objective was to determine multiaxis control-power characteristics from thrust vectoring. A two-dimensional convergent-divergent nozzle was designed to provide yaw vector angles of 0, -10, and -20 deg combined with geometric pitch vector angles of 0 and 15 deg. Yaw thrust vectoring was provided by yaw flaps located in the nozzle sidewalls. Roll control was obtained from differential pitch vectoring. This investigation was conducted at Mach numbers from 0.20 to 2.47. Angle of attack was varied from 0 to about 19 deg, and nozzle pressure ratio was varied from about 1 (jet off) to 28, depending on Mach number. Increments in force or moment coefficient that result from pitch or yaw thrust vectoring remain essentially constant over the entire angle-of-attack range of all Mach numbers tested. There was no effect of pitch vectoring on the lateral aerodynamic forces and moments and only very small effects of yaw vectoring on the longitudinal aerodynamic forces and moments. This result indicates little cross-coupling of control forces and moments for combined pitch-yaw vectoring.

Capone, Francis J.↗

Static performance investigation of a skewed-throat multiaxis thrust-vectoring nozzle concept

The static performance of a jet exhaust nozzle which achieves multiaxis thrust vectoring by physically skewing the geometric throat has been characterized in the static test facility of the 16-Foot Transonic Tunnel at NASA Langley Research Center. The nozzle has an asymmetric internal geometry defined by four surfaces: a convergent-divergent upper surface with its ridge perpendicular to the nozzle centerline, a convergent-divergent lower surface with its ridge skewed relative to the nozzle centerline, an outwardly deflected sidewall, and a straight sidewall. The primary goal of the concept is to provide efficient yaw thrust vectoring by forcing the sonic plane (nozzle throat) to form at a yaw angle defined by the skewed ridge of the lower surface contour. A secondary goal is to provide multiaxis thrust vectoring by combining the skewed-throat yaw-vectoring concept with upper and lower pitch flap deflections. The geometric parameters varied in this investigation included lower surface ridge skew angle, nozzle expansion ratio (divergence angle), aspect ratio, pitch flap deflection angle, and sidewall deflection angle. Nozzle pressure ratio was varied from 2 to a high of 11.5 for some configurations. The results of the investigation indicate that efficient, substantial multiaxis thrust vectoring was achieved by the skewed-throat nozzle concept. However, certain control surface deflections destabilized the internal flow field, which resulted in substantial shifts in the position and orientation of the sonic plane and had an adverse effect on thrust-vectoring and weight flow characteristics. By increasing the expansion ratio, the location of the sonic plane was stabilized. The asymmetric design resulted in interdependent pitch and yaw thrust vectoring as well as nonzero thrust-vector angles with undeflected control surfaces. By skewing the ridges of both the upper and lower surface contours, the interdependency between pitch and yaw thrust vectoring may be eliminated and the location of the sonic plane may be further stabilized.

Wing, David J.↗

Implementing Scientific Simulation Codes Highly Tailored for Vector Architectures Using Custom Configurable Computing Machines

The motivation for this work comes from an observation that amidst the push for Massively Parallel (MP) solutions to high-end computing problems such as numerical physical simulations, large amounts of legacy code exist that are highly optimized for vector supercomputers. Because re-hosting legacy code often requires a complete re-write of the original code, which can be a very long and expensive effort, this work examines the potential to exploit reconfigurable computing machines in place of a vector supercomputer to implement an essentially unmodified legacy source code. Custom and reconfigurable computing resources could be used to emulate an original application's target platform to the extent required to achieve high performance. To arrive at an architecture that delivers the desired performance subject to limited resources involves solving a multi-variable optimization problem with constraints. Prior research in the area of reconfigurable computing has demonstrated that designing an optimum hardware implementation of a given application under hardware resource constraints is an NP-complete problem. The premise of the approach is that the general issue of applying reconfigurable computing resources to the implementation of an application, maximizing the performance of the computation subject to physical resource constraints, can be made a tractable problem by assuming a computational paradigm, such as vector processing. This research contributes a formulation of the problem and a methodology to design a reconfigurable vector processing implementation of a given application that satisfies a performance metric. A generic, parametric, architectural framework for vector processing implemented in reconfigurable logic is developed as a target for a scheduling/mapping algorithm that maps an input computation to a given instance of the architecture. This algorithm is integrated with an optimization framework to arrive at a specification of the architecture parameters that attempts to minimize execution time, while staying within resource constraints. The flexibility of using a custom reconfigurable implementation is exploited in a unique manner to leverage the lessons learned in vector supercomputer development. The vector processing framework is tailored to the application, with variable parameters that are fixed in traditional vector processing. Benchmark data that demonstrates the functionality and utility of the approach is presented. The benchmark data includes an identified bottleneck in a real case study example vector code, the NASA Langley Terminal Area Simulation System (TASS) application.

Rutishauser, David↗