Academic Projects
The Graph Database (GDB) comparison is designed to provide comprehensively compare and better understand the existing graph database system solutions. It has been developed in collaboration with Prof. Y. Velgrakis and M. Lissandrini. The Graph Benchmark made its first appearance in our "Beyond Macrobenchmarks: Microbenchmark-based Graph Database Evaluation" VLDB19 research paper.
"CLIC: What if?" is a framework for the simulation for calls for tenders under the Italian regulations. It has been developed in collaboration with the faculty of law at the Univesity of Trento, within the project "CLIC: Trasparenza e modernizzazione delle procedure di aggiudicazione degli appalti pubblici".
Archetypum is a Tomasulo's protocol visualizer. With it, users can design a simple processor and then simulate the execution of an arbitrary assembly program on it.
The processor comes configured with reasonable settings for a live demo/teaching session. The user may then decide anytime to modify any of its capabilities like registers, functional units, memory latency, cache, and branch predictor.
Once completed the customization process, the user shall provide a program to execute. They may decide between using one of the examples and write a program in our ARM-like assembly. The language supports all the essential instructions, which are also explained in the documentation.
Finally, it comes the time for the simulation.
The execution is governed by a clock, with adjustable speed, that the user can stop anytime.
Once stopped, the user may decide to make it resume or advance cycle by cycle.
The simulation view provides, besides the clock controls,
a series of widgets (tables) that display the
status of each component of the processor.
Archetypum has been developed in collaboration with G. Bortoli as a project for the Advanced Architecture course by Prof. R.Passerone.
RaftScope is the Raft consensus protocol visualizer originally developed by Ongaro to present his work.
A. Nardelli, F.Hoxa and I added support for the missing features, like cluster membership configuration changes, and we updated it to the latest revision. This work was conducted as a project for the Distributed Algorithm course by Prof A. Montresor.
DIP is the result of the Dataset Integration Project: a dataset and process management system for big data. It is a novel software address the needs only partially fulfilled by the (at the time) existing solutions (e.g., CKAN, Oozie). This project will be released as a module of a framework that is being developed by the dbTrento group.