To provide the student with the basic tools for the design, implementation and management of complex calculation systems for the processing of large amounts of data.
teacher profile teaching materials
• Computer Architecture:
- Logical and physical organization
- CPU architecture (parallelism, pipeline, superscalar architecture, registers, operations, buffers and internal cache)
- system bus and peripheral bus, main memory, disks
- parallel multicore, multiprocessor and GPU architectures
• Operating systems:
- general functions
- kernel, processes and memory organization, scheduling algorithms
- file system
• Virtual systems and containers
- Virtual machine architecture
- Conainer architecture
• Communication networks:
- Network architectures, topologies of local and geographic networks
- standard TCP / IP routing and communication protocols
- Networks for HPC computing
• Storage systems:
- physical structuring
- RAID systems
- high performance file system
• HPC systems:
- intensive computing, algorithm parallelism, computer farm and job scheduling systems
- MPI libraries for running parallel programs
- Scheduling systems
- new frontiers of scientific computing and GRID.
- Cloud systems
• Algorithms, codes and programs on HPC architectures
Examples of algorithm development and execution of parallel architectures
Development examples using MPI
Development examples on GPU cards
- A. S. Tanenbaum, H. Bos, B. Crispo, C. Palazzi, I moderni sistemi operativi
- A. S. Tanenbaum, T.Austin, Architettura dei calcolatori. Un approccio strutturale
Programme
Premise: the course is delivered in the academic year 2023/2024 still with the old name of "Data Management Laboratory", which will change in 2024/2025 to "High Performance Computing Laboratory". The content of the course is as follows:• Computer Architecture:
- Logical and physical organization
- CPU architecture (parallelism, pipeline, superscalar architecture, registers, operations, buffers and internal cache)
- system bus and peripheral bus, main memory, disks
- parallel multicore, multiprocessor and GPU architectures
• Operating systems:
- general functions
- kernel, processes and memory organization, scheduling algorithms
- file system
• Virtual systems and containers
- Virtual machine architecture
- Conainer architecture
• Communication networks:
- Network architectures, topologies of local and geographic networks
- standard TCP / IP routing and communication protocols
- Networks for HPC computing
• Storage systems:
- physical structuring
- RAID systems
- high performance file system
• HPC systems:
- intensive computing, algorithm parallelism, computer farm and job scheduling systems
- MPI libraries for running parallel programs
- Scheduling systems
- new frontiers of scientific computing and GRID.
- Cloud systems
• Algorithms, codes and programs on HPC architectures
Examples of algorithm development and execution of parallel architectures
Development examples using MPI
Development examples on GPU cards
Core Documentation
- J. F. Kurose, K. W. Ross , Reti di calcolatori e internet. Un approccio top-down- A. S. Tanenbaum, H. Bos, B. Crispo, C. Palazzi, I moderni sistemi operativi
- A. S. Tanenbaum, T.Austin, Architettura dei calcolatori. Un approccio strutturale
Type of delivery of the course
Theoretical lessons in presence and online, Laboratory in presenceType of evaluation
Test on the topics covered during the laboratory hours Oral examination