HEAP Lab Talk
High Performance Computing in Neuroscience – a look at the Human Brain Project
Francesco Leporati
Associate Professor at the Computer Science Department of the University of Pavia
DEIB - Seminar Room (building 20)
December 14th, 2017
12.30 pm - 2.00 pm
Contact:
Francesca Micol Rossi
Research Line:
System architectures
Francesco Leporati
Associate Professor at the Computer Science Department of the University of Pavia
DEIB - Seminar Room (building 20)
December 14th, 2017
12.30 pm - 2.00 pm
Contact:
Francesca Micol Rossi
Research Line:
System architectures
Sommario
On December 14th, 2017 at 12.30 pm, “High Performance Computing in Neuroscience – a look at the Human Brain Project” seminar will be held in DEIB Seminar Room, as new appointment of HEAP Lab Talks.
The correct and complete comprehension of the activity and functionalities of the human brain, together with its pathologies, is one of the most fascinating tasks that scientists face up.
Throughout the world, several projects have been funded to improve the knowledge of the physiology and of the mechanisms that form the basis of our neural activity. This has consequences that not only in medicine but also in other disciplines like biology, psychology, ethics and law.
Looking at current works in neuroscience, it seems very clear that the modelisation and simulation of brain mechanisms but also the investigation technologies adopted to do it, most of which are related to bioimaging, all play an important role.
This often implies strong intersections between electronics and computer science, leading to new research in the reverse engineering of neuronal functions, brain inspired intelligent systems and very sophisticated deep learning approaches.
There is therefore a need for more research on High Performance Computing for (realistic) neuroscience which employs both software and hardware instruments to simulate brain models. Typically these technologies are a cluster of many core processors or supercomputers, GPUs, parallel DSPs architectures for simulation and FPGA or ASIC for neuronal circuits reproduction (i. e. “brain hardware”).
The talk will focus on the problems that scientists face in this field of research and, in particular, the challenges that two EU projects Human Brain Project and Helicoid have encountered.
The correct and complete comprehension of the activity and functionalities of the human brain, together with its pathologies, is one of the most fascinating tasks that scientists face up.
Throughout the world, several projects have been funded to improve the knowledge of the physiology and of the mechanisms that form the basis of our neural activity. This has consequences that not only in medicine but also in other disciplines like biology, psychology, ethics and law.
Looking at current works in neuroscience, it seems very clear that the modelisation and simulation of brain mechanisms but also the investigation technologies adopted to do it, most of which are related to bioimaging, all play an important role.
This often implies strong intersections between electronics and computer science, leading to new research in the reverse engineering of neuronal functions, brain inspired intelligent systems and very sophisticated deep learning approaches.
There is therefore a need for more research on High Performance Computing for (realistic) neuroscience which employs both software and hardware instruments to simulate brain models. Typically these technologies are a cluster of many core processors or supercomputers, GPUs, parallel DSPs architectures for simulation and FPGA or ASIC for neuronal circuits reproduction (i. e. “brain hardware”).
The talk will focus on the problems that scientists face in this field of research and, in particular, the challenges that two EU projects Human Brain Project and Helicoid have encountered.
