NEXT:
4 December 2025 | h 14:30 | @Aula Magistrale 5th floor

Short Talk

Your Optimizing Compiler is Not Optimizing Enough. To Hell With Multiple Recursions!

Federico Bruzzone

Optimizing compilers exhibit limitations when handling multiple recursive functions (e.g., Fibonacci). While single recursion is optimized via Tail-Call Optimization into efficient iteration, multiple recursion often forces the compiler to retain a partial, performance-limiting recursive structure. Analyzing the optimized LLVM IR reveals that even state-of-the-art compilers fail to achieve complete incrementalization for this pattern. We argue that the automatic, general-setting iterative transformation of multiple recursive functions remains a partially unsolved and critical challenge in compiler theory and practice.

Short Talk

Limit Break with GPUs: Massive Crowd Simulation

Vincenzo Lombardo

The ability to simulate realistic crowds is a highly sought-after capability in the fields of entertainment (video games, movies), urban planning and evacuation simulations. Traditional approaches to crowd simulation rely on heavy Central Processing Unit (CPU) computation. This approach has limitations in terms of scalability and performance, which are solvable with the use of Graphics Programming Units (GPUs) and parallel computing techniques. In fact, the development of Compute Shaders on GPU allows the execution of general-purpose operations alongside traditional rendering tasks within real-time applications. This paper aims to contribute to the current literature on crowd simulation methods by developing a real-time simulation model that integrates and expands several techniques from literature, adapted and optimized to exploit GPU computing capabilities. The proposed model incorporates continuous representations for crowds in order to simulate human movement and decision-making. The achieved results demonstrate a high level of scalability and efficiency. The implemented techniques and optimizations allow the model to handle a significant number of agents while maintaining real-time performances to achieve reduced simulation time and good user experience. Stress tests showcase that the proposed model significantly outperforms other macroscopic models, maintaining a stable frame rate of 60 FPS when simulating 20,000 agents even on mid-range systems intended for personal use.

Short Talk

Recognizing Distance-Count Matrices is Difficult

Chiara Prezioso

Axiomatization of centrality measures often involves proving that something cannot hold by providing a counterexample (i.e., a graph for which that specific centrality index fails to have a given property). In the context of geometric centralities, building such counterexamples requires constructing a graph with specific distance counts between nodes, as expressed by its distance-count matrix. We prove that deciding whether a matrix is the distance-count matrix of a graph is strongly NP-complete. This negative result implies that a brute-force approach to building this kind of counterexample is out of question, and cleverer approaches are required.

Presentations

Research Spotlights, What Else?

Several researchers will share insights into their current work and ongoing projects. Alberto Bertoncini will discuss "Service Orchestration in Dynamic Multi-layer Heterogeneous Mobile Networks", focusing on scalable and adaptive methods for managing resources and deploying modular services in fast-evolving scenarios. Alessandro Cagiano will give a presentation on "Model Compression and Applied AI: Knowledge Distillation, Text-to-Speech, and Legal AI Systems", exploring the design of reduced-complexity AI pipelines aimed at expressive audio generation and dependable assistance in juridical domains. Simone Casorerio will present his research on "Boolean Functions, Circuits, and Cryptography", exploring the connections between these fundamental concepts in computer science and their applications in secure communication. Finally, Ali Tabaraei will investigate "The Impact of Domain Generalization on Explainable Deep Learning for Heart Sound Classification", aiming to enhance the interpretability and reliability of deep learning models in medical diagnostics.