The use of graphics processing units (GPUs) in high-performance parallel computing continues to steadily become more prevalent, often as part of a heterogeneous system. For years, CUDA has been the de facto programming environment for nearly all general-purpose GPU (GPGPU) applications. In spite of this, the framework is available only on NVIDIA GPUs, traditionally requiring reimplementation in other frameworks in order to utilize additional multi-or many-core devices. On the other hand, OpenCL provides an open and vendorneutral programming environment and run-time system. With implementations available for CPUs, GPUs, and other types of accelerators, OpenCL therefore holds the promise of a "write once, run anywhere" ecosystem for heterogeneous computing.Given the many similarities between CUDA and OpenCL, manually porting a CUDA application to OpenCL is almost straightforward, albeit tedious and error-prone. In response to this issue, we created CU2CL, an automated CUDA-to-OpenCL source-to-source translator that possesses a novel design and clever reuse of the Clang compiler framework. Currently, the CU2CL translator covers the primary constructs found in the CUDA Runtime API, and we have successfully translated several applications from the CUDA SDK and Rodinia benchmark suite. CU2CL's translation times are reasonable, allowing for many applications to be translated at once. The number of manual changes required after executing our translator on CUDA source is minimal, with some compiling and working with no changes at all. The performance of our automatically translated applications via CU2CL is on par with their manually ported counterparts.

Traditionally, real-time systems require that the deadlines of all jobs be met. For many applications, however, this is an overly stringent requirement. An occasional missed deadline may cause decreased performance but is nevertheless acceptable. We present an analysis technique by which a lower bound on the percentage of deadlines that a periodic task meets is determined and compare the lower bound with simulation results for an example system. We have implemented the technique in the PERTS real-time system prototyping environment [6,7].

No abstract

Abstract

The GridPP Collaboration is building a UK computing Grid for particle physics, as part of the international effort towards computing for the Large Hadron Collider. The project, funded by the UK Particle Physics and Astronomy Research Council (PPARC), began in September 2001 and completed its first phase 3 years later. GridPP is a collaboration of approximately 100 researchers in 19 UK university particle physics groups, the Council for the Central Laboratory of the Research Councils and CERN, reflecting the strategic importance of the project. In collaboration with other European and US efforts, the first phase of the project demonstrated the feasibility of developing, deploying and operating a Grid-based computing system to meet the UK needs of the Large Hadron Collider experiments. This note describes the work undertaken to achieve this goal. S Supplementary documentation is available from stacks.iop.org/JPhysG/32/N1. References to sections S1, S2.1, etc are to sections within this online supplement.