We describe chipStar, an open source software stack that enables building unmodified CUDA and HIP programs into binaries that rely solely on open cross-vendor compute standards OpenCL and SPIR-V. The relevant technical aspects of chipStar and the feature mismatches between the CUDA/HIP APIs and OpenCL are discussed along with a set of standard extension proposals to bridge the essential gaps in the future. The key benefit of the software stack is its portability, which is demonstrated by providing performance evaluations on a diversity of less common CPU/GPU platforms including RISC-V/PowerVR and ARM Mali. A comparison against the original AMD HIP platform provides a geometric mean of 0.75, a reasonable price to pay for the enhanced portability. chipStar is now considered mature enough for wider testing and even production use, which is demonstrated by successful porting and competitive performance of GAMESS-GPU-HF, a complex HPC application.
AlpayAHeuvelineV (2023) One pass to bind them: the first single-pass sycl compiler with unified code representation across backends. In: Proceedings of the 2023 International Workshop on Opencl, IWOCL ’23. Association for Computing Machinery. https://doi.org/10.1145/3585341.3585351
2.
AlpayASoproniBWünscheH, et al. (2022) Exploring the possibility of a hipsycl-based implementation of oneapi. In: International Workshop on Opencl, IWOCL’22. Association for Computing Machinery. https://doi.org/10.1145/3529538.3530005
ApplencourtTSadawarteAMuralidharanS, et al. (2024) Ponte vecchio across the Atlantic: Single-node benchmarking of two intel gpu systems. In: IEEE International Workshop on Performance Modeling, Benchmarking and Simulation of High Performance Computer Systems. IEEE.
7.
BabejMJääskeläinenP (2020) HIPCL: tool for porting CUDA applications to advanced OpenCL platforms through HIP. In: Proceedings of the International Workshop on OpenCL, IWOCL ’20. Association for Computing Machinery. https://doi.org/10.1145/3388333.3388641
8.
BarcaGMJPooleDLVallejoJLG, et al. (2020) Scaling the hartree-fock matrix build on summit. In: SC20: International Conference for High Performance Computing, Networking, Storage and Analysis. IEEE, 1–14. https://doi.org/10.1109/SC41405.2020.00085
9.
BarcaGJSnowdonCVallejoJ, et al. (2022) Scaling correlated fragment molecular orbital calculations on summit. In: SC22: International Conference for High Performance Computing, Networking, Storage and Analysis. IEEE Computer Society, 1–14. https://doi.org/10.1109/SC41404.2022.00012
10.
CytronRFerranteJRosenBK, et al. (1991) Efficiently computing static single assignment form and the control dependence graph. ACM Transactions on Programming Languages and Systems13(4): 451–490. https://doi.org/10.1145/115372.115320
11.
DoerfertJJasperMHuberJ, et al. (2023) Breaking the vendor lock: performance portable programming through openmp as target independent runtime layer. In: Proceedings of the International Conference on Parallel Architectures and Compilation Techniques, PACT ’22. Association for Computing Machinery, 494–504. https://doi.org/10.1145/3559009.3569687
12.
GardnerMSathrePFengW, et al. (2013) Characterizing the challenges and evaluating the efficacy of a CUDA-to-OpenCL translator. Parallel Computing39(12): 769–786.
13.
GordonMSSchmidtMW (2005) Chapter 41 - advances in electronic structure theory: gamess a decade later. In: DykstraCEFrenkingGKimKS, et al. (eds) Theory and Applications of Computational Chemistry. Elsevier, 1167–1189.
JääskeläinenPde La LamaCSSchnetterE, et al. (2015) Pocl: a performance-portable OpenCL implementation. International Journal of Parallel Programming43(5): 752–785. https://doi.org/10.10007/s10766-014-0320-y
JinZVetterJS (2023) A benchmark suite for improving performance portability of the sycl programming model. In: 2023 IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS). IEEE, 325–327. https://doi.org/10.11009/ISPASS57527.2023.00041
KwackJTrammJBertoniC, et al. (2021) Evaluation of performance portability of applications and mini-apps across amd, intel and nvidia gpus. In: 2021 International Workshop on Performance, Portability and Productivity in HPC (P3HPC). IEEE, 45–56. https://doi.org/10.1109/P3HPC54578.2021.00008
32.
LattnerCAdveV (2004) LLVM: a compilation framework for lifelong program analysis & transformation. In: Proceedings of the 2004 International Symposium on Code Generation and Optimization (CGO’04). California.
PerkinsH (2017) Cuda-on-cl: a compiler and runtime for running nvidia® cudaTM c++11 applications on openclTM 1.2 devices. In: Proceedings of the 5th International Workshop on Opencl, IWOCL 2017. Association for Computing Machinery. https://doi.org/10.1145/3078155.3078156
39.
RydahlAHuberJMcDonoughEL, et al. (2023) Precision and performance analysis of C standard math library functions on GPUs. Proceedings of the ACM on Programming Languages7(OOPSLA2): 1–26. https://doi.org/10.1145/3624062.3624166
40.
SchmidtMWBaldridgeKKBoatzJA, et al. (1993) General atomic and molecular electronic structure system. Journal of Computational Chemistry14(11): 1347–1363.
41.
StrattonJStoneSHwuW (2008) Mcuda: an efficient implementation of cuda kernels for multi-core cpus. In: Languages and Compilers for Parallel Computing - 21st International Workshop, LCPC 2008, Revised Selected Papers, Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). International Workshop on Languages and Compilers for Parallel Computing, 16–30. https://doi.org/10.1007/978-3-540-89740-8_2
ZhaoJBertoniCYoungJ, et al. (2022) HIPLZ: enabling performance portability for exascale systems. In: Proceedings of the Heteropar 2022: 20Th International Workshop on Algorithms, Models and Tools for Parallel Computing on Heterogeneous Platforms. Springer.
