The final review meeting of the FP7 REPARA project took place in Bruxelles end of October. REPARA demonstrated that existing sequential standard C++ code may be refactored such that both execution time and power consumption may be reduced to one half, by applying the REPARA assisted, semi-automatic refactoring methodology. Our…
The paper “A reconfiguration Algorithm for Power-Aware Parallel Applications” by Daniele De Sensi, Massimo Torquati and Marco Danelutto has been accepted for publication in the journal ACM Transactions on Architecture and Code Optimization (TACO). We were invited to present the work at HiPEAC 2017 Conference (Link: http://www.hipeac.net/2017/stockholm/).
The REPARA project has ended. Our group was leading the WP6 “Runtime for parallel heterogeneous architectures”. See also UNIPI web site
The paper “Proactive elasticity and energy awareness in data stream processing” by Tiziano De Matteis and Gabriele Mencagli has been accepted for publication in Journal of Systems and Software. Link: http://www.sciencedirect.com/science/article/pii/S0164121216301467
A challenging issue in the development of parallel implementations of Data Stream Processing patters is to be able to reduce the costs of dynamic memory allocation. In this work we study two parallel patterns for sliding-window computations and we discuss different implementation variants related to how dynamic memory is managed. …
We presented a poster about our activity on power aware computing at ACACES 2016 summer school. [Poster] [Abstract]
The paper “Data stream processing via code annotations” by Marco Danelutto, Tiziano De Matteis, Gabriele Mencagli and Massimo Torquati has been accepted for for publication of The Journal of Supercomputing (Springer). Link: http://link.springer.com/article/10.1007/s11227-016-1793-9
A DSL based toolchain supporting the design of parallel applications where parallelism is structured after parallel design pattern compositions has been presented at the TOOLS workshop at the ICCS 2016 conference in San Diego. The proof-of-concept prototype developed so far supports pattern expression editing, refactoring, optimization with the possibility to consult…
The following papers from our group has been accepted at 9th International Symposium on High-Level Parallel Programming and Applications: “A Power-Aware, Self-Adaptive Macro Data Flow Framework” by Marco Danelutto, Daniele De Sensi and Massimo Torquati “SPar: a DSL for High-Level and Productive Stream Parallelism” by Dalvan Griebler, Marco Danelutto, Massimo Torquati and Luiz Fernandes “Analysing…
The paper entitled “Continuous skyline queries on multicore architectures” by Tiziano De Matteis, Salvatore Di Girolamo and Gabriele Mencagli has been accepted for publication on Concurrency and Computation: Practice and Experience.
