Graham Pullan

Acceleration of a 3D Euler Solver Using Commodity Graphics Hardware

The porting of twoand three-dimensional Euler solvers from a conventional CPU implementation to the novel target platform of the Graphics Processing Unit (GPU) is described. The motivation for such an effort is the impressive performance that GPUs offer: typically 10 times more floating point operations per second than a modern CPU, with over 100 processing cores and all at a very modest financial cost. Both codes were found to generate the same results on the GPU as the FORTRAN versions did on the CPU. The 2D solver ran up to 29 times quicker on the GPU than on the CPU; the 3D solver 16 times faster. Nomenclature cv Specific heat capacity at constant volume e Specific total energy = cvT + 1 2 V 2 h0 Specific stagnation enthalpy p Pressure t Time u,v Cartesian components of velocity V Velocity (magnitude) T Temperature Yp Stagnation pressure loss coefficient = p01−p0 p01−p2 tb302@cam.ac.uk gp10006@cam.ac.uk

BarraCUDA - a fast short read sequence aligner using graphics processing units

BackgroundWith the maturation of next-generation DNA sequencing (NGS) technologies, the throughput of DNA sequencing reads has soared to over 600 gigabases from a single instrument run. General purpose computing on graphics processing units (GPGPU), extracts the computing power from hundreds of parallel stream processors within graphics processing cores and provides a cost-effective and energy efficient alternative to traditional high-performance computing (HPC) clusters. In this article, we describe the implementation of BarraCUDA, a GPGPU sequence alignment software that is based on BWA, to accelerate the alignment of sequencing reads generated by these instruments to a reference DNA sequence.FindingsUsing the NVIDIA Compute Unified Device Architecture (CUDA) software development environment, we ported the most computational-intensive alignment component of BWA to GPU to take advantage of the massive parallelism. As a result, BarraCUDA offers a magnitude of performance boost in alignment throughput when compared to a CPU core while delivering the same level of alignment fidelity. The software is also capable of supporting multiple CUDA devices in parallel to further accelerate the alignment throughput.ConclusionsBarraCUDA is designed to take advantage of the parallelism of GPU to accelerate the alignment of millions of sequencing reads generated by NGS instruments. By doing this, we could, at least in part streamline the current bioinformatics pipeline such that the wider scientific community could benefit from the sequencing technology.BarraCUDA is currently available from http://seqbarracuda.sf.net

An Accelerated 3D Navier-Stokes Solver for Flows in Turbomachines

A new three-dimensional Navier-Stokes solver for flows in turbomachines has been developed. The new solver is based on the latest version of the Denton codes, but has been implemented to run on Graphics Processing Units (GPUs) instead of the traditional Central Processing Unit (CPU). The change in processor enables an order-of-magnitude reduction in run-time due to the higher performance of the GPU. Scaling results for a 16 node GPU cluster are also presented, showing almost linear scaling for typical turbomachinery cases. For validation purposes, a test case consisting of a three-stage turbine with complete hub and casing leakage paths is described. Good agreement is obtained with previously published experimental results. The simulation runs in less than 10 minutes on a cluster with four GPUs.Copyright

Acceleration of a two-dimensional Euler flow solver using commodity graphics hardware

Abstract The implementation of a two-dimensional Euler solver on graphics hardware is described. The graphics processing unit is highly parallelized and uses a programming model that is well suited to flow computation. Results for a transonic turbine cascade test-case are presented. For large grids (106 nodes) a 40 times speed-up compared with a Fortran implementation on a contemporary CPU is observed.

Secondary Flows and Loss Caused by Blade Row Interaction in a Turbine Stage

A study of the three-dimensional stator-rotor interaction in a turbine stage is presented. Experimental data reveal vortices downstream of the rotor which are stationary in the absolute frame - indicating that they are caused by the stator exit flowfield. Evidence of the rotor hub passage vortices is seen, but additional vortical structures away from the endwalls, which would not be present if the rotor were tested in isolation, are also identified. An unsteady computation of the rotor row is performed using the measured stator exit flowfield as the inlet boundary condition. The strength and location of the vortices at rotor exit are predicted. A formation mechanism is proposed whereby stator wake fluid with steep spanwise gradients of absolute total pressure is responsible for all but one of the rotor exit vortices. This mechanism is then verified computationally using a passive-scalar tracking technique. The predicted loss generation through the rotor row is then presented and a comparison made with a steady calculation where the inlet flow has been mixed out to pitchwise uniformity. The loss produced in the steady simulation, even allowing for the mixing loss at inlet, is 10% less than that produced in the unsteady simulation. This difference highlights the importance of the time-accurate calculation as a tool of the turbomachine designer.

The interaction of turbine inter-platform leakage flow with the mainstream flow

Individual nozzle guide vanes (NGV???s) in modern aeroengines are often cast as a single piece with integral hub and casing endwalls. When in operation, there is a leakage flow through the chord-wise interplatform gaps. An investigation into the effect of this leakage flow on turbine performance is presented. Efficiency measurements and NGV exit area traverse data from a low-speed research turbine are reported. Tests show that this leakage flow can have a significant impact on turbine performance, but that below a threshold leakage fraction this penalty does not rise with increasing leakage flow rate. The effect of various seal clearances are also investigated. Results from steady-state simulations using a three-dimensional multiblock Reynolds-averaged Navier-Stokes solver are presented with particular emphasis paid to the physics of the mainstream/leakage interaction and the loss generation.

The Importance of Shroud Leakage Modeling in Multistage Turbine Flow Calculations

Three-dimensional steady multistage calculations, using the mixing plane approach, are compared with experimental measurement in a low-speed three-stage model turbine. The comparisons are made with two levels of shroud seal clearance, one representative of a real turbine and one with minimal seal clearance and almost no shroud leakage. Three different calculations are compared. The first computes the main blade path with no modeling of shroud leakage. The second includes a simple model of shroud leakage using sources and sinks on the end-walls, and the third is a multiblock calculation with all leakage paths and cavities computed. It is found that neglect of shroud leakage makes the computed velocity profiles and loss distributions significantly different to those measured. Simple modeling of shroud leakage gives some improvement but full calculation of the leakage flows and cavities is necessary to obtain good agreement between calculation and measurement.

An Experimental and Computational Study of the Formation of a Streamwise Shed Vortex in a Turbine Stage

Shear layers shed by aircraft wings roll up into vortices. A similar, though far less common, phenomenon can occur in the wake of a turbomachine blade. This paper presents experimental data from a new single-stage turbine that has been commissioned at the Whittle Laboratory. Two low-aspect ratio stators have been tested with the same rotor row. Surface flow visualization illustrates the extremely strong secondary flows present in both NGV designs. These secondary flows lead to conventional passage vortices, but also to an intense vortex sheet which is shed from the trailing edge of the blades. Pneumatic probe traverses show how this sheet rolls up into a concentrated vortex in the second stator design, but not in the first. A simple numerical experiment is used to model the shear layer instability and the effects of trailing edge shape and exit yaw angle distribution are investigated. It is found that the latter has a strong influence on shear layer rollup: inhibiting the formation of a vortex downstream of NGV 1 but encouraging it behind NGV 2.

An Accelerated 3D Navier–Stokes Solver for Flows in Turbomachines

A new three-dimensional Navier–Stokes solver for flows in turbomachines has been developed. The new solver is based on the latest version of the Denton codes but has been implemented to run on graphics processing units (GPUs) instead of the traditional central processing unit. The change in processor enables an order-of-magnitude reduction in run-time due to the higher performance of the GPU. The scaling results for a 16 node GPU cluster are also presented, showing almost linear scaling for typical turbomachinery cases. For validation purposes, a test case consisting of a three-stage turbine with complete hub and casing leakage paths is described. Good agreement is obtained with previously published experimental results. The simulation runs in less than 10 min on a cluster with four GPUs.

SBLOCK: A Framework for Efficient Stencil-Based PDE Solvers on Multi-core Platforms

We present a new software framework for the implementation of applications that use stencil computations on block-structured grids to solve partial differential equations. A key feature of the framework is the extensive use of automatic source code generation which is used to achieve high performance on a range of leading multi-core processors. Results are presented for a simple model stencil running on Intel and AMD CPUs as well as the NVIDIA GT200 GPU. The generality of the frame- work is demonstrated through the implementation of a complete application consisting of many different stencil computations, taken from the field of computational fluid dynamics.