Acceleration of MRI analysis using multicore and manycore paradigms

Pantoja, Maria; Weyrich, Maxence; Fernández–Escribano, Gerardo

doi:10.1007/s11227-020-03154-9

Cited by 7 publications

(11 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Pantoja et al [18] proposed a hybrid approach to accelerate FreeSurfer with both a GPU and a multicore CPU. In their study, CPU-based code runs with OpenMP threads, which is the same as the original implementation of FreeSurfer.…”

Section: Previous Workmentioning

confidence: 99%

“…Giuliano et al's work [21] showed that running FreeSurfer on a hybrid CPU-to-GPU infrastructure can lead to a reduced execution time compared to using a CPU-only infrastructure. Nevertheless, same as in [18] and [20], an obsolete version of FreeSurfer was used.…”

Section: Previous Workmentioning

confidence: 99%

“…Given the limitations of previous works, in the present study, we accelerate three kernels including one kernel accelerated in [18]. We utilize OpenACC directives to generate GPU-based code, which significantly reduces programming costs caused by manually rewriting kernels with CUDA.…”

Section: Previous Workmentioning

confidence: 99%

“…Although previous work [18] has rewritten the gcamSmoothnessEnergy() kernel with CUDA, for minimized programming costs, we accelerate the same kernel with OpenACC directives. In addition to gcamSmoothnessEnergy(), we also implement GPU-accelerated versions of two other kernels: gcamLabelEnergy() and gcamJacobianEnergy().…”

Section: Target Kernels For Accelerationmentioning

confidence: 99%

“…Given that FreeSurfer is a large-scale project with more than 100 million lines of source code and is being updated frequently to add new features and bug fixes, to create a CUDA-based GPU acceleration of FreeSurfer, developers must rewrite the computation kernels with CUDA to allow compatibility with code changes and version updates of FreeSurfer. As a result of the huge programming costs, the core development team of FreeSurfer stopped integrating the CUDA-based GPU acceleration into FreeSurfer after the release of FreeSurfer 5.0 in 2012 [18].…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Integrating GPU support for FreeSurfer with OpenACC

Shen

Mei

Walldén

et al. 2020

Preprint

View full text Add to dashboard Cite

FreeSurfer is among the most widely used suites of software for the study of cortical and subcortical brain anatomy. However, analysis using FreeSurfer can be time-consuming and it lacks support for the graphics processing units (GPUs) after the core development team stopped maintaining GPU-accelerated versions due to significant programming cost. As FreeSurfer is a large project with millions of source lines, in this work, we introduce and examine the use of a directive-based framework, OpenACC, in GPU acceleration of FreeSurfer, and we found the OpenACC-based approach significantly reduces programming costs. Moreover, because the overhead incurred by CPU-to-GPU data transfer is the major challenge in delivering GPU-based codes of high performance, we compare two schemes, copy-and-transfer and overlapped-fully-transfer, to reduce such data transfer overhead. Experimental results show that the target function we accelerated with overlapped-fully-transfer scheme ran 2.3x as fast as the original CPU-based function, and the GPU-accelerated program achieved an average speedup of 1.2x compared to the original CPU-based program. These results demonstrate the usefulness and potential of utilizing the proposed OpenACC-based approach to integrate GPU support for FreeSurfer which can be easily extended to other computationally expensive functions and modules of FreeSurfer to achieve further speedup.

show abstract

Section: Previous Workmentioning

confidence: 99%

Section: Previous Workmentioning

confidence: 99%

Section: Previous Workmentioning

confidence: 99%

Section: Target Kernels For Accelerationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations