GPUs-RRTMG_LW: high-efficient and scalable computing for a longwave radiative transfer model on multiple GPUs

Wang, Yuzhu; Guo, Mingxin; Zhao, Yuan; Jiang, Junguang

doi:10.1007/s11227-020-03451-3

Cited by 10 publications

(5 citation statements)

References 43 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…His RRTMG longwave radiation scheme (RRTMG_LW) achieves an acceleration speed of 69× when implemented on a NVIDIA GTX 680 GPU [20] and 127× on a single K40 GPU [21]. Y. Wang, et al developed a GPU version of RRTMG_LW that was implemented in CUDA Fortran and integrated into CAS-ESM [22]- [24].…”

Section: Related Workmentioning

confidence: 99%

GPU-RRTMG_SW: Accelerating a Shortwave Radiative Transfer Scheme on GPU

Wang

et al. 2021

IEEE Access

Self Cite

View full text Add to dashboard Cite

As high performance computing technology continues to develop rapidly, graphics processing units (GPUs) are more widely used for daily computing tasks. However, some applications, such as weather forecasting, require large-scale computation. Thus, due to the computationally intensive characteristics of the rapid radiative transfer model for general circulation models (RRTMG) in the Earth system model, this study uses GPU-related technology to accelerate computation of the model. First, two kinds of algorithms using GPU technology to accelerate the RRTMG shortwave radiation scheme (RRTMG_SW) are proposed. Then, an optimization method for data transmission between host and device is proposed. Finally, after using CUDA Fortran and CUDA C to implement these algorithms, two GPU versions of RRTMG_SW, namely CF-RRTMG_SW (CUDA Fortran version) and CC-RRTMG_SW (CUDA C version), were developed. The experimental results demonstrate that the proposed acceleration algorithms are effective. Without I/O transfer, running CC-RRTMG_SW on a NVIDIA GeForce Titan V GPU achieved a 38.88× speedup when compared to a single Intel Xeon E5-2680 CPU core.INDEX TERMS high performance computing, graphics processing unit, compute unified device architecture, shortwave radiative transfer.

show abstract

Section: Related Workmentioning

confidence: 99%

GPU-RRTMG_SW: Accelerating a Shortwave Radiative Transfer Scheme on GPU

Wang

et al. 2021

IEEE Access

Self Cite

View full text Add to dashboard Cite

show abstract

“…CC BY 4.0 License. Mielikainen et al, 2012b;Mielikainen et al, 2013a ;Mielikainen et al, 2013b;Price et al, 2014;Huang et al, 2015), ocean models such as LASG/IAP Climate System Ocean Model (LICOM; Jiang et al, 2019;Wang et al, 2021a) and Princeton Ocean Model (POM; Xu et al, 2015), and the Earth System Model of Chinese Academy of Sciences (CAS-EMS; Wang et al, 2021b ;Wang et al, 2021c). Govett et al, (2017) used Open Accelerator (OpenACC) directives to port the dynamics of NIM to the GPU and achieved 2.5x acceleration.…”

Section: Introductionmentioning

confidence: 99%

“…For the Princeton Ocean Model, Xu et al, (2015) use CUDA C to carry out heterogeneous porting and optimization, the performance of gpu-POM v1.0 on four GPUs is comparable to that on 408 standard Intel Xeon X5670 CPU cores. In terms of climate system model, Wang et al, (2021c) and Wang et al, (2021b) porting scheme is the most complex, but its computational performance is the highest (Mielikainen et al, 2012b;Wahib and Maruyama, 2013;Xu et al, 2015).…”

Section: Introductionmentioning

confidence: 99%

GPU-HADVPPM V1.0: high-efficient parallel GPU design of the Piecewise Parabolic Method (PPM) for horizontal advection in air quality model (CAMx V6.10)

Cao

Wang

et al. 2023

Preprint

View full text Add to dashboard Cite

Abstract. With semiconductor technology gradually approaching its physical and thermal limits, Graphics processing unit (GPU) is becoming an attractive solution in many scientific applications due to their high performance. This paper presents an application of GPU accelerators in air quality model. We endeavor to demonstrate an approach that runs a PPM solver of horizontal advection (HADVPPM) for air quality model CAMx on GPU clusters. Specifically, we first convert the HADVPPM to a new Compute Unified Device Architecture C (CUDA C) code to make it computable on the GPU (GPU-HADVPPM). Then, a series of optimization measures are taken, including reducing the CPU-GPU communication frequency, increasing the size of data computation on GPU, optimizing the GPU memory access, and using thread and block indices in order to improve the overall computing performance of CAMx model coupled with GPU-HADVPPM (named as CAMx-CUDA model). Finally, a heterogeneous, hybrid programming paradigm is presented and utilized with the GPU-HADVPPM on GPU clusters with Massage Passing Interface (MPI) and CUDA. Offline experiment results show that running GPU-HADVPPM on one NVIDIA Tesla K40m and NVIDIA Tesla V100 GPU can achieve up to 845.4x and 1113.6x acceleration. By implementing a series of optimization schemes, the CAMx-CUDA model resulted in a 29.0x and 128.4x improvement in computational efficiency using a GPU accelerator card on a K40m and V100 cluster, respectively. In terms of the single-module computational efficiency of GPU-HADVPPM, it can achieve 1.3x and 19.4x speedup on NVIDIA Tesla K40m GPU and NVIDA Tesla V100 GPU respectively. The multi-GPU acceleration algorithm enables 4.5x speedup with 8 CPU cores and 8 GPU accelerators on V100 cluster.

show abstract

“…SHDOM is almost 2 orders of magnitude more computationally efficient than Monte Carlo on CPU for multi-angle imagery (Pincus and Evans, 2009). Monte Carlo solvers specialized for 3D atmospheric scattering problems have been slow to adopt a GPU-based computation, which is anticipated to give a reduction in the wall time of between 1 and 2 orders of magnitude (Efremenko et al, 2014;Ramon et al, 2019;Wang et al, 2021;Lee et al, 2022), thereby making Monte Carlo competitive against SHDOM in the future.…”

Section: Introductionmentioning

confidence: 99%

Retrieving 3D distributions of atmospheric particles using Atmospheric Tomography with 3D Radiative Transfer – Part 1: Model description and Jacobian calculation

et al. 2023

View full text Add to dashboard Cite

Abstract. Our global understanding of clouds and aerosols relies on the remote sensing of their optical, microphysical, and macrophysical properties using, in part, scattered solar radiation. These retrievals assume that clouds and aerosols form plane-parallel, homogeneous layers and utilize 1D radiative transfer (RT) models, limiting the detail that can be retrieved about the 3D variability in cloud and aerosol fields and inducing biases in the retrieved properties for highly heterogeneous structures such as cumulus clouds and smoke plumes. To overcome these limitations, we introduce and validate an algorithm for retrieving the 3D optical or microphysical properties of atmospheric particles using multi-angle, multi-pixel radiances and a 3D RT model. The retrieval software, which we have made publicly available, is called Atmospheric Tomography with 3D Radiative Transfer (AT3D). It uses an iterative, local optimization technique to solve a generalized least squares problem and thereby find a best-fitting atmospheric state. The iterative retrieval uses a fast, approximate Jacobian calculation, which we have extended from Levis et al. (2020) to accommodate open and periodic horizontal boundary conditions (BCs) and an improved treatment of non-black surfaces. We validated the accuracy of the approximate Jacobian calculation for derivatives with respect to both the 3D volume extinction coefficient and the parameters controlling the open horizontal boundary conditions across media with a range of optical depths and single-scattering properties and find that it is highly accurate for a majority of cloud and aerosol fields over oceanic surfaces. Relative root mean square errors in the approximate Jacobian for a 3D volume extinction coefficient in media with cloud-like single-scattering properties increase from 2 % to 12 % as the maximum optical depths (MODs) of the medium increase from 0.2 to 100.0 over surfaces with Lambertian albedos <0.2. Over surfaces with albedos of 0.7, these errors increase to 20 %. Errors in the approximate Jacobian for the optimization of open horizontal boundary conditions exceed 50 %, unless the plane-parallel media providing the boundary conditions are optically very thin (∼0.1). We use the theory of linear inverse RT to provide insight into the physical processes that control the cloud tomography problem and identify its limitations, supported by numerical experiments. We show that the Jacobian matrix becomes increasing ill-posed as the optical size of the medium increases and the forward-scattering peak of the phase function decreases. This suggests that tomographic retrievals of clouds will become increasingly difficult as clouds become optically thicker. Retrievals of asymptotically thick clouds will likely require other sources of information to be successful. In Loveridge et al. (2023a; hereafter Part 2), we examine how the accuracy of the retrieved 3D volume extinction coefficient varies as the optical size of the target medium increases using synthetic data. We do this to explore how the increasing error in the approximate Jacobian and the increasingly ill-posed nature of the inversion in the optically thick limit affect the retrieval. We also assess the accuracy of retrieved optical depths and compare them to retrievals using 1D radiative transfer.

show abstract

GPUs-RRTMG_LW: high-efficient and scalable computing for a longwave radiative transfer model on multiple GPUs

Cited by 10 publications

References 43 publications

GPU-RRTMG_SW: Accelerating a Shortwave Radiative Transfer Scheme on GPU

GPU-RRTMG_SW: Accelerating a Shortwave Radiative Transfer Scheme on GPU

GPU-HADVPPM V1.0: high-efficient parallel GPU design of the Piecewise Parabolic Method (PPM) for horizontal advection in air quality model (CAMx V6.10)

Retrieving 3D distributions of atmospheric particles using Atmospheric Tomography with 3D Radiative Transfer – Part 1: Model description and Jacobian calculation

Contact Info

Product

Resources

About