Abstract
With AMD reinforcing their ambition in the scientific high performance computing ecosystem, we extend the hardware scope of the Ginkgo linear algebra package to feature a HIP backend for AMD GPUs. In this paper, we report and discuss the porting effort from CUDA, the extension of the HIP framework to add missing features such as cooperative groups, the performance price of compiling HIP code for AMD architectures, and the design of a library providing native backends for NVIDIA and AMD GPUs while minimizing code duplication by using a shared code base.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
A complex atomic_add involves separate real and imaginary atomic_add and thus is not strictly an atomic operation, as no ordering between the individual components of multiple complex atomic operations is guaranteed.
- 2.
References
The Top 500 List. https://www.top500.org/
The US Exascale Computing Project (ECP). https://www.exascaleproject.org/
Anzt, H., Dongarra, J., Flegar, G., Higham, N.J., Quintana-Ortí, E.S.: Adaptive precision in block-Jacobi preconditioning for iterative sparse linear system solvers. Concurrency Comput. Pract. Exp. 31(6), e4460 (2019)
Danalis, A., et al.: The scalable heterogeneous computing (SHOC) benchmark suite. In: Proceedings of the 3rd Workshop on General-Purpose Computation on Graphics Processing Units, pp. 63–74 (2010). https://doi.org/10.1145/1735688.1735702. dl.acm.org
Kuznetsov, E., Stegailov, V.: Porting CUDA-based molecular dynamics algorithms to AMD ROCm platform using hip framework: performance analysis. In: Voevodin, V., Sobolev, S. (eds.) RuSCDays 2019. CCIS, vol. 1129, pp. 121–130. Springer, Cham (2019). https://doi.org/10.1007/978-3-030-36592-9_11
NVIDIA Corp.: Whitepaper: NVIDIA TESLA V100 GPU Architecture (2017)
Roth, P.C.: Experiences with the Heterogeneouscompute Interface for Portability (HIP) on OLCF Summit, October 2019. https://www.olcf.ornl.gov/wp-content/uploads/2019/10/Roth-HIP-on-Summit-20191009.pdf
SuiteSparse: Matrix Collection. https://sparse.tamu.edu. Accessed Jan 2020
Sun, Y., et al.: Evaluating performance tradeoffs on the radeon open compute platform. In: 2018 IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS), pp. 209–218, April 2018. https://doi.org/10.1109/ISPASS.2018.00034
Zubair, M., Warner, J., Wagner, D.: Optimization of a solver for computational materials and structures problems on NVIDIA Volta and AMD Instinct GPUs. In: 2019 IEEE/ACM 10th Workshop on Latest Advances in Scalable Algorithms for Large-Scale Systems (ScalA), pp. 9–16, November 2019. https://doi.org/10.1109/ScalA49573.2019.00007
Acknowledgements
This research was supported by the Exascale Computing Project (17-SC-20-SC) and the Helmholtz Impuls und Vernetzungsfond VH-NG-1241.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2021 Springer Nature Switzerland AG
About this paper
Cite this paper
Tsai, Y.M., Cojean, T., Ribizel, T., Anzt, H. (2021). Preparing Ginkgo for AMD GPUs – A Testimonial on Porting CUDA Code to HIP. In: Balis, B., et al. Euro-Par 2020: Parallel Processing Workshops. Euro-Par 2020. Lecture Notes in Computer Science(), vol 12480. Springer, Cham. https://doi.org/10.1007/978-3-030-71593-9_9
Download citation
DOI: https://doi.org/10.1007/978-3-030-71593-9_9
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-71592-2
Online ISBN: 978-3-030-71593-9
eBook Packages: Computer ScienceComputer Science (R0)