Searcharxiv⌕ Search

arXiv subjects

Fabio Luporini

Publications and source records attributed to Fabio Luporini.

21 records · Page 2Linked to original sources

An algorithm for the optimization of finite element integration loops

We present an algorithm for the optimization of a class of finite element integration loop nests. This algorithm, which exploits fundamental mathematical properties of finite element operators, is proven to achieve a locally optimal operation count. In specified circumstances the optimum achieved is global. Extensive numerical experiments demonstrate significant performance improvements over the state of the art in finite element code generation in almost all cases. This validates the effectiveness of the algorithm presented here, and illustrates its limitations.

cs.MS↗

COFFEE: an Optimizing Compiler for Finite Element Local Assembly

The numerical solution of partial differential equations using the finite element method is one of the key applications of high performance computing. Local assembly is its characteristic operation. This entails the execution of a problem-specific kernel to numerically evaluate an integral for each element in the discretized problem domain. Since the domain size can be huge, executing efficient kernels is fundamental. Their op- timization is, however, a challenging issue. Even though affine loop nests are generally present, the short trip counts and the complexity of mathematical expressions make it hard to determine a single or unique sequence of successful transformations. Therefore, we present the design and systematic evaluation of COF- FEE, a domain-specific compiler for local assembly kernels. COFFEE manipulates abstract syntax trees generated from a high-level domain-specific language for PDEs by introducing domain-aware composable optimizations aimed at improving instruction-level parallelism, especially SIMD vectorization, and register locality. It then generates C code including vector intrinsics. Experiments using a range of finite-element forms of increasing complexity show that significant performance improvement is achieved.

cs.MS↗

A Survey on Open Problems for Mobile Robots

Gathering mobile robots is a widely studied problem in robotic research. This survey first introduces the related work, summarizing models and results. Then, the focus shifts on the open problem of gathering fat robots. In this context, "fat" means that the robot is not represented by a point in a bidimensional space, but it has an extent. Moreover, it can be opaque in the sense that other robots cannot "see through" it. All these issues lead to a redefinition of the original problem and an extension of the CORDA model. For at most 4 robots an algorithm is provided in the literature, but is gathering always possible for n>4 fat robots? Another open problem is considered: Boundary Patrolling by mobile robots. A set of mobile robots with constraints only on speed and visibility is working in a polygonal environment having boundary and possibly obstacles. The robots have to perform a perpetual movement (possibly within the environment) so that the maximum timespan in which a point of the boundary is not being watched by any robot is minimized.

cs.RO↗