Multi-Node Multi-GPU Diffeomorphic Image Registration for Large-Scale Imaging Problems

Brunn, Malte; Himthani, Naveen; Biros, George; Mehl, Miriam; Mang, Andreas

doi:10.1109/sc41405.2020.00042

Cited by 12 publications

(55 citation statements)

References 50 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We provide technology that allows solving registration problems for clinical datasets in seconds. In addition, we have made available to the public a software that works on multi-node, multi-GPU architectures ( Brunn et al, 2020 , 2021 ) that allows the registration of large-scale microscopic imaging data such as CLARITY imaging ( Kutten et al, 2017 ; Tomer et al, 2014 ).…”

Section: Statement Of Needmentioning

confidence: 99%

“…CLAIRE can be used to register images of 2048 3 (25 B unknowns) on 64 nodes with 256 GPUs on TACC’s Longhorn system ( Brunn et al, 2020 ). CLAIRE has been used for the registration of high resolution CLARITY imaging data ( Brunn et al, 2020 ).…”

Section: Highlightsmentioning

confidence: 99%

“… Summary CLAIRE ( Mang & Biros, 2019 ) is a computational framework for C onstrained LA rge deformation diffeomorphic I mage RE gistration ( Mang et al, 2019 ). It supports highly-optimized, parallel computational kernels for (multi-node) CPU ( Gholami et al, 2017 ; Mang et al, 2019 ; Mang & Biros, 2016 ) and (multi-node multi-)GPU architectures ( Brunn et al, 2020 , 2021 ). CLAIRE uses MPI for distributed-memory parallelism and can be scaled up to thousands of cores ( Mang et al, 2019 ; Mang & Biros, 2016 ) and GPU devices ( Brunn et al, 2020 ).…”

mentioning

confidence: 99%

“…It supports highly-optimized, parallel computational kernels for (multi-node) CPU ( Gholami et al, 2017 ; Mang et al, 2019 ; Mang & Biros, 2016 ) and (multi-node multi-)GPU architectures ( Brunn et al, 2020 , 2021 ). CLAIRE uses MPI for distributed-memory parallelism and can be scaled up to thousands of cores ( Mang et al, 2019 ; Mang & Biros, 2016 ) and GPU devices ( Brunn et al, 2020 ). The multi-GPU implementation uses device direct communication.…”

mentioning

confidence: 99%

“…It features various schemes for regularization of the control problem ( Mang & Biros, 2016 ) and different similarity measures. CLAIRE implements different preconditioners for the reduced space Hessian ( Brunn et al, 2020 ; Mang et al, 2019 ) to optimize computational throughput and enable fast convergence. It uses PETSc ( Balay et al, n.d. ) for scalable and efficient linear algebra operations and solvers and TAO ( Balay et al, n.d. ; Munson et al, 2015 ) for numerical optimization.…”

mentioning

confidence: 99%

See 4 more Smart Citations

CLAIRE: Constrained Large Deformation Diffeomorphic Image Registration on Parallel Computing Architectures

Brunn¹,

Himthani²,

Biros³

et al. 2021

JOSS

Self Cite

View full text Add to dashboard Cite

CLAIRE (Mang & Biros, 2019) is a computational framework for Constrained LArge deformation diffeomorphic Image REgistration . It supports highly-optimized, parallel computational kernels for (multi-node) CPU (Gholami et al., 2017;Mang & Biros, 2016) and (multi-node multi-)GPU architectures (Brunn et al., 2020(Brunn et al., , 2021. CLAIRE uses MPI for distributed-memory parallelism and can be scaled up to thousands of cores Mang & Biros, 2016) and GPU devices (Brunn et al., 2020). The multi-GPU implementation uses device direct communication. The computational kernels are interpolation for semi-Lagrangian time integration, and a mixture of high-order finite difference operators and Fast-Fourier-Transforms (FFTs) for differentiation. CLAIRE uses a Newton-Krylov solver for numerical optimization (Mang & Biros, 2015. It features various schemes for regularization of the control problem (Mang & Biros, 2016) and different similarity measures. CLAIRE implements different preconditioners for the reduced space Hessian (Brunn et al., 2020; to optimize computational throughput and enable fast convergence. It uses PETSc (Balay et al., n.d.) for scalable and efficient linear algebra operations and solvers and TAO (Balay et al., n.d.;Munson et al., 2015) for numerical optimization. CLAIRE can be downloaded at https://github.com/andreasmang/claire.

show abstract