Search NASA⌕ Search

Engineering topics

Dubois, Andrew J

Publications and source records attributed to Dubois, Andrew J.

Non-preconditioned conjugate gradient on cell and FPCA-based hybrid supercomputer nodes

This work presents a detailed implementation of a double precision, Non-Preconditioned, Conjugate Gradient algorithm on a Roadrunner heterogeneous supercomputer node. These nodes utilize the Cell Broadband Engine Architecture{trademark} in conjunction with x86 Opteron{trademark} processors from AMD. We implement a common Conjugate Gradient algorithm, on a variety of systems, to compare and contrast performance. Implementation results are presented for the Roadrunner hybrid supercomputer, SRC Computers, Inc. MAPStation SRC-6 FPGA enhanced hybrid supercomputer, and AMD Opteron only. In all hybrid implementations wall clock time is measured, including all transfer overhead and compute timings.

97 MATHEMATICS AND COMPUTING↗

Non-preconditioned conjugate gradient on cell and FPGA based hybrid supercomputer nodes

This work presents a detailed implementation of a double precision, non-preconditioned, Conjugate Gradient algorithm on a Roadrunner heterogeneous supercomputer node. These nodes utilize the Cell Broadband Engine Architecture{sup TM} in conjunction with x86 Opteron{sup TM} processors from AMD. We implement a common Conjugate Gradient algorithm, on a variety of systems, to compare and contrast performance. Implementation results are presented for the Roadrunner hybrid supercomputer, SRC Computers, Inc. MAPStation SRC-6 FPGA enhanced hybrid supercomputer, and AMD Opteron only. In all hybrid implementations wall clock time is measured, including all transfer overhead and compute timings.

99 GENERAL AND MISCELLANEOUS↗

Sparse matrix-vector multiplication on a reconfigurable supercomputer

Double precision floating point Sparse Matrix-Vector Multiplication (SMVM) is a critical computational kernel used in iterative solvers for systems of sparse linear equations. The poor data locality exhibited by sparse matrices along with the high memory bandwidth requirements of SMVM result in poor performance on general purpose processors. Field Programmable Gate Arrays (FPGAs) offer a possible alternative with their customizable and application-targeted memory sub-system and processing elements. In this work we investigate two separate implementations of the SMVM on an SRC-6 MAPStation workstation. The first implementation investigates the peak performance capability, while the second implementation balances the amount of instantiated logic with the available sustained bandwidth of the FPGA subsystem. Both implementations yield the same sustained performance with the second producing a much more efficient solution. The metrics of processor and application balance are introduced to help provide some insight into the efficiencies of the FPGA and CPU based solutions explicitly showing the tight coupling of the available bandwidth to peak floating point performance. Due to the FPGA's ability to balance the amount of implemented logic to the available memory bandwidth it can provide a much more efficient solution. Finally, making use of the lessons learned implementing the SMVM, we present an fully implemented nonpreconditioned Conjugate Gradient Algorithm utilizing the second SMVM design.

97 MATHEMATICS AND COMPUTING↗

A complete implementation of the conjugate gradient algorithm on a reconfigurable supercomputer

The conjugate gradient is a prominent iterative method for solving systems of sparse linear equations. Large-scale scientific applications often utilize a conjugate gradient solver at their computational core. In this paper we present a field programmable gate array (FPGA) based implementation of a double precision, non-preconditioned, conjugate gradient solver for fmite-element or finite-difference methods. OUf work utilizes the SRC Computers, Inc. MAPStation hardware platform along with the 'Carte' software programming environment to ease the programming workload when working with the hybrid (CPUIFPGA) environment. The implementation is designed to handle large sparse matrices of up to order N x N where N <= 116,394, with up to 7 non-zero, 64-bit elements per sparse row. This implementation utilizes an optimized sparse matrix-vector multiply operation which is critical for obtaining high performance. Direct parallel implementations of loop unrolling and loop fusion are utilized to extract performance from the various vector/matrix operations. Rather than utilize the FPGA devices as function off-load accelerators, our implementation uses the FPGAs to implement the core conjugate gradient algorithm. Measured run-time performance data is presented comparing the FPGA implementation to a software-only version showing that the FPGA can outperform processors running up to 30x the clock rate. In conclusion we take a look at the new SRC-7 system and estimate the performance of this algorithm on that architecture.

97 MATHEMATICS AND COMPUTING↗