Abstract. This paper presents an application of GPU accelerators in Earth system modeling. We focus on atmospheric chemical kinetics, one the most computationally intensive tasks climate–chemistry model simulations. developed a software package that automatically generates CUDA kernels to numerically integrate kinetics global climate ECHAM/MESSy Atmospheric Chemistry (EMAC), used study change and air quality scenarios. A source-to-source compiler outputs CUDA-compatible kernel by parsing...

The goal of Partitioned Global Address Space (PGAS) languages is to improve programmer productivity in large scale parallel machines. However, PGAS programs may have many fine-grained shared accesses that lead performance degradation. Manual code transformations or compiler optimizations are required the with accesses. downside manual increased program complexity hinders productivity. On other hand, most fine-grain require knowledge physical data mapping and use loop constructs.

The study of atmospheric chemistry-climate interactions is one today's great computational challenges. Advances in the architecture Graphics Processing Units (GPUs) both raw power and memory bandwidth sparked interest for General-Purpose computing on graphics accelerators scientific applications. However, introduction GPUs High Performance Computing (HPC) landscape increased complexity software development, due to inherent heterogeneity requirements programming models design approaches,...

Future multi-core processors will necessitate exploitation of fine-grain, architecture-independent parallelism from applications to utilize many cores with relatively small local memories. We use c264, an end-to-end H.264 video encoder for the Cell processor based on x264, show that exploiting fine-grain remains challenging and requires significant advancement in runtime support. Our implementation c264 achieves speedup between 4.7× 8.6× six synergistic processing elements (SPEs), compared...

No abstract available.

The global climate model ECHAM/MESSy Atmospheric Chemistry (EMAC) is used to study change and air quality scenarios. EMAC constituted by a nonlocal dynamical part with low scalability, local physical/chemical processes high scalability. chemistry-climate does not benefit from the support of accelerators which are nowadays installed in many HPC systems. We strategies offload calculation atmospheric chemistry accelerator technologies (GPU Intel MIC), as typical configurations this most...

Single Instruction Multiple Data (SIMD) Extensions become popular in computer architectures as a simple and efficient way to exploit the data parallelism hidden applications. The compiler research community has proposed automatic vectorization answer complexity of low-level programming vector units. Despite recent advances compilation techniques, modern compilers miss opportunities automatically vectorize code. One biggest challenges is evaluate changes against best hand-written This paper...

The global climate model ECHAM/MESSy Atmospheric Chemistry (EMAC) is a modular that simulates change and air quality scenarios. application includes different sub-models for the calculation of chemical species concentrations, their interaction with land sea, human interaction. paper presents source-to-source parser enables support Graphics Processing Units (GPU) by Kinetic Pre-Processor (KPP) general purpose open-source software tool. requirements host system are also described. source code...

Programs written in Partitioned Global Address Space (PGAS) languages can access any location of the entire address space via standard read/write operations. However, compiler have to create communication mechanisms and runtime system use synchronization primitives ensure correct execution programs. PGAS programs may fine-grained shared accesses that lead performance degradation. One solution is inspector-executor technique determine which are indeed remote be coalesced larger A...

Significant progress has been made in the development of programming languages and tools that are suitable for hybrid computer architectures group several shared-memory multicores interconnected through a network. This paper addresses important limitations code generation partitioned global address space (PGAS) languages. These allow fine-grained communication lead to programs perform many accesses data. When data is distributed remote computing nodes, transformations required prevent...

Abstract. This paper presents an application of GPU accelerators in Earth system modelling. We focus on atmospheric chemical kinetics, one the most computationally intensive tasks climate-chemistry model simulations. developed a software package that automatically generates CUDA kernels to numerically integrate kinetics global climate ECHAM/MESSy Atmospheric Chemistry (EMAC), used study change and air quality scenarios. A source-to-source compiler outputs compatible kernel, by parsing...