Challenges and possible approaches: towards the petaflops computers

Qian, Depei; Zhu, Danfeng

doi:10.1007/s11704-009-0022-6

Cited by 3 publications

(2 citation statements)

References 50 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…GPU (Graphics Processor Unit) has been applied extensively in various scientific and engineering domains. These are the primary components in the latest supercomputers in terms of performance gain and energy efficiency 1 . Due to its high parallelism and efficiency, GPU was utilized to obtain high performance in many of linear algebra problems involving matrices.…”

Section: Introductionmentioning

confidence: 99%

Padding free bank conflict resolution for CUDA-based matrix transpose algorithm

Khan

Al-Mouhamed

Fatayar

et al. 2014

15th IEEE/ACIS International Conference on Software Engineering, Artificial Intelligence, Networking and Parallel/Distributed C

View full text Add to dashboard Cite

The advances of Graphic Processing Units (GPU) technology and the introduction of CUDA programming model facilitates developing new solutions for sparse and dense linear algebra solvers. Matrix Transpose is an important linear algebra procedure that has deep impact in various computational science and engineering applications. Several factors hinder the expected performance of large matrix transpose on GPU devices. The degradation in performance involves the memory access pattern such as coalesced access in the global memory and bank conflict in the shared memory of streaming multiprocessors within the GPU. In this paper, two matrix transpose algorithms are proposed to alleviate the aforementioned issues of ensuring coalesced access and conflict free bank access. The proposed algorithms have comparable execution times with the NVIDIA SDK bank conflict-free matrix transpose implementation. The main advantage of proposed algorithms is that they eliminate bank conflicts while allocating shared memory exactly equal to the tile size (T x T) of the problem space. However, to the best of our knowledge an extra space of Tx(T+1) needs to be allocated in the published research. We have also applied the proposed transpose algorithm to recursive gaussian implementation of NVIDIA SDK and achieved about 6% improvement in performance.

show abstract

Section: Introductionmentioning

confidence: 99%

Padding free bank conflict resolution for CUDA-based matrix transpose algorithm

Khan

Al-Mouhamed

Fatayar

et al. 2014

15th IEEE/ACIS International Conference on Software Engineering, Artificial Intelligence, Networking and Parallel/Distributed C

View full text Add to dashboard Cite

show abstract

“…These are the primary components in the latest supercomputers in terms of performance gain and energy efficiency 1 . Due to its high parallelism and efficiency, GPU was utilized to obtain high performance in many of linear algebra problems involving matrices.…”

Section: Introductionmentioning

confidence: 99%

Padding Free Bank Conflict Resolution for CUDA-Based Matrix Transpose Algorithm

Khan¹,

Al-Mouhamed²,

Fatayer³

et al. 2014

IJNDC

View full text Add to dashboard Cite

The advances of Graphic Processing Units (GPU) technology and the introduction of CUDA programming model facilitates developing new solutions for sparse and dense linear algebra solvers. Matrix Transpose is an important linear algebra procedure that has deep impact in various computational science and engineering applications. Several factors hinder the expected performance of large matrix transpose on GPU devices. The degradation in performance involves the memory access pattern such as coalesced access in the global memory and bank conflict in the shared memory of streaming multiprocessors within the GPU. In this paper, two matrix transpose algorithms are proposed to alleviate the aforementioned issues of ensuring coalesced access and conflict free bank access. The proposed algorithms have comparable execution times with the NVIDIA SDK bank conflict -free matrix transpose implementation. The main advantage of proposed algorithms is that they eliminate bank conflicts while allocating shared memory exactly equal to the tile size (T x T) of the problem space. However, to the best of our knowledge an extra space of Tx(T+1) needs to be allocated in the published research. We have also applied the proposed transpose algorithm to recursive gaussian implementation of NVIDIA SDK and achieved about 6% improvement in performance.

show abstract