this paper proposes a novel hardware compilation approach targeting dataflow designs. this approach is based on aspect-orientedprogramming to decouple design development from design optimisation, thus improving porta...
详细信息
ISBN:
(纸本)9781479904945;9781479904938
this paper proposes a novel hardware compilation approach targeting dataflow designs. this approach is based on aspect-orientedprogramming to decouple design development from design optimisation, thus improving portability and developer productivity while enabling automated exploration of design trade-offs to enhance performance. We introduce FAST, a language for specifying dataflow designs that supports our approach. Optimisation strategies for the generated designs are specified in FAST, making use of facilities in the domain-specific aspect-oriented language, LARA. Our approach is demonstrated by implementing various seismic imaging designs for Reverse-Time Migration (RTM), which have performance comparable to state-of-the-art FPGA implementations while being produced with improved developer productivity.
the proceedings contain 6 papers. the topics discussed include: declarative layer composition in framework-based environments;scoping changes in self-supporting development environments with context- oriented programm...
ISBN:
(纸本)9781450312769
the proceedings contain 6 papers. the topics discussed include: declarative layer composition in framework-based environments;scoping changes in self-supporting development environments with context- orientedprogramming;uniting global and local context behavior with context petri nets;bridging real-world contexts and units of behavioral variations by composite layers;and DynamicSchema: a lightweight persistency framework for context-oriented data management.
We present a novel design methodology for the mapping of nested loops onto programmable hardware accelerators. Key features of our approach are: (1) Design entry in form of a functional programming language and loop p...
详细信息
ISBN:
(纸本)9781479904945;9781479904938
We present a novel design methodology for the mapping of nested loops onto programmable hardware accelerators. Key features of our approach are: (1) Design entry in form of a functional programming language and loop parallelization in the polyhedron model, (2) the underlying accelerator architectures consist of lightweight, tightly-coupled, and programmable processor arrays, which can exploit both loop-level parallelism and instruction-level parallelism, (3) support of zero-overhead looping not only for innermost loops but also for arbitrarily nested loops. We implemented the proposed methodology in a prototype design tool and evaluated selected benchmarks by comparing our code generator withthe Trimaran compilation framework. As the results show, our approach can reduce the size of the generated processor codes up to 64% while at the same time achieving a significant higher throughput.
objectorientedprogramming languages have raised the level of abstraction by supporting the object querying on collections. programming languages can execute first class query constructs, using query optimization tec...
详细信息
ISBN:
(纸本)9780769549484;9781467358330
objectorientedprogramming languages have raised the level of abstraction by supporting the object querying on collections. programming languages can execute first class query constructs, using query optimization techniques from the database field, for run time optimizations. Existing approaches, however, such as Java Query Language (JQL), which executes such query constructs on collections have high run time overhead. therefore, we propose an approach to reduce the burden of run time overhead by performing most of the query optimization for object queries on collections at compile time. this approach both analyzes the source code and obtains the metadata provided through annotations. It relies on building histograms from the metadata information. then, the predicate and join selectivity estimates within a query are computed from these histograms. the selectivity estimates are maintained accurate withthe incremental maintenance of histograms to the data changes in the program at run time. Next, boththe selection and join optimizations are applied on the queries. the optimizations help in skipping and eliminating the execution of some of the predicates and joins based on the collected metadata. Finally, a query plan is generated at the compile time through the proposed selectivity cost heuristic. the query itself is executed at run time according to the determined query plan. But, in cases of inaccurate metadata and significant data changes in the source code, the query plan is modified at run time according to the correct selectivity estimates obtained from the updated histograms. Our experimental results demonstrate that our approach reduces the run time overhead of a program with collections more than the earlier approaches such as JQL.
the proceedings contain 30 papers. the topics discussed include: is it dangerous to use version control histories to study source code evolution?;evaluating the design of the r language: objects and functions for data...
ISBN:
(纸本)9783642310560
the proceedings contain 30 papers. the topics discussed include: is it dangerous to use version control histories to study source code evolution?;evaluating the design of the r language: objects and functions for data analysis;McSAF: a static analysis framework for MATLAB;multiple aggregate entry points for ownership types;inference and checking of object ownership;object initialization in X10;lock inference in the presence of large libraries;an analysis of the Mozilla jetpack extension framework;smaller footprint for java collections;enhancing JavaScript with transactions;JavaScript as an embedded DSL;correlation tracking for points-to analysis of JavaScript;constraint-based refactoring with foresight;magda: a new language for modularity;and Marco: safe, expressive macros for any language.
the proceedings contain 6 papers. the topics discussed include: context-orientedprogramming in highly concurrent systems;designing event-based context transition in context-orientedprogramming;the Cartesian approach...
ISBN:
(纸本)9781450305310
the proceedings contain 6 papers. the topics discussed include: context-orientedprogramming in highly concurrent systems;designing event-based context transition in context-orientedprogramming;the Cartesian approach to context;layered method dispatch with NVOKEDYNAMIC - an implementation study;ContextLua - dynamic behavioral variations in computer games;and a graph-based operational semantics for context-orientedprogramming.
the proceedings contain 9 papers. the topics discussed include: complete coinductive subtyping for abstract compilation of object-oriented languages;towards a semantic model for java wildcards;strong exception-safety ...
ISBN:
(纸本)9781450305402
the proceedings contain 9 papers. the topics discussed include: complete coinductive subtyping for abstract compilation of object-oriented languages;towards a semantic model for java wildcards;strong exception-safety for java-like languages;modular verification of linked lists with views via separation logic;procedure-modular verification of control flow safety properties;prototyping a tool environment for run-time assertion checking in JML with communication histories;aliasing control with view-based typestate;refactoring and representation independence for class hierarchies;and a theory of typestate-orientedprogramming.
Several recent papers on automatic face verification have significantly raised the performance bar by developing novel, specialised representations that outperform standard features such as SIFT for this problem. this...
详细信息
Several recent papers on automatic face verification have significantly raised the performance bar by developing novel, specialised representations that outperform standard features such as SIFT for this problem. this paper makes two contributions: first, and somewhat surprisingly, we show that Fisher vectors on densely sampled SIFT features, i.e. an off-the-shelf object recognition representation, are capable of achieving state-of-the-art face verification performance on the challenging "Labeled Faces in the Wild" benchmark;second, since Fisher vectors are very high dimensional, we show that a compact descriptor can be learnt from them using discriminative metric learning. this compact descriptor has a better recognition accuracy and is very well suited to large scale identification tasks.
We present a technique to make applications resilient to failures. this technique is intended to maintain a faulty application functional in the field while the developers work on permanent and radical fixes. We targe...
详细信息
ISBN:
(纸本)9781467330763
We present a technique to make applications resilient to failures. this technique is intended to maintain a faulty application functional in the field while the developers work on permanent and radical fixes. We target field failures in applications built on reusable components. In particular, the technique exploits the intrinsic redundancy of those components by identifying workarounds consisting of alternative uses of the faulty components that avoid the failure. the technique is currently implemented for Java applications but makes little or no assumptions about the nature of the application, and works without interrupting the execution flow of the application and without restarting its components. We demonstrate and evaluate this technique on four mid-size applications and two popular libraries of reusable components affected by real and seeded faults. In these cases the technique is effective, maintaining the application fully functional with between 19% and 48% of the failure-causing faults, depending on the application. the experiments also show that the technique incurs an acceptable runtime overhead in all cases.
暂无评论