The talk will report about the publication “The baryon density of the Universe from an improved rate of deuterium burning”, Nature 587, 210–213(2020).
Data from high energy physics experiments are expensive. It takes many person-hours and multiple millions of euros/dollars/pounds/swiss francs to produce and collect data from the LHC.
Deep Neural Networks (DNN) represent a performance-hungry application. Floating-Point (FP) and custom floating-point-like arithmetic satisfies this hunger.
Seminars and colloquia
These present the latest particle physics results, with one section (LHC seminars) dedicated to the LHC experiments, and the other (non-LHC seminars) open to all other experimental activities, at CERN and elsewhere.
The seminars will also include presentations of new ideas, proposals or status reports from projects in progress.
Data science is a growing interdisciplinary field, consisting in the development of new methods to process and analyse large datasets, in order to extract insight on specific problems. Many of these technique are part of HEP data processing and analysis workflows since decades. The advent of modern computational technologies has boosted research in Data Science, with development of novel approaches to data analysis, specific software tools and specific computing architectures. These new technologies offer new opportunities and solutions to problems encountered in HEP research. This seminar series is an opportunity for the CERN community to get in contact with people from HEP as well as from other science domains and industry, developing new data science solutions.
These are non-technical talks of general interest covering a broad range, not limited only to to particle physcis and related fields.
They are intended for a general academic audience and open to all people at CERN from all departments.
Organised by the Detector Seminar Panel
The Detector Seminar Panel has the mandate to organise regular Detector Seminars and occasional Detector Forums. These two events aim to make state-of-the-art presentations for scientist and engineers in the following fields:
· Particle detectors and their performances
· Detector engineering and infrastructure
· Detector electronics
· Detector applications in HEP and other fields
The aim of the Detector Seminars is to spread information on new developments and to stimulate in depth discussions within the detector community present at CERN. The seminars are held in regular intervals (typically monthly) and have specific topics or summarize highlights from conferences. The speakers are experts in the field and are selected by the panel.
The Detector Forums intend to organise half- or full-day events around one theme that is of interest for the detector community at CERN. The Forum events treat a topic more profoundly and include presentations from different groups.
EP Software Seminars
Organised by Andrea Bocci (EP-CMG), Dirk Düllmann (IT-ST-AD), Peter Hristov (EP-AIP), Axel Naumann (EP-SFT), Niko Neufeld (EP-LBC), and Andreas Salzburger (EP-ADP)
Scientific software development expertise is crucial for the mission of CERN, and warrants a visible platform for exchange.
The EP Software Seminar provides this platform, following the tradition set by the Detector Seminar, the Electronics Seminar, and the EP-IT Data-Science Seminar.
The EP Software Seminar's role is the dissemination of results from software activities in the context of EP, as well as external high-profile presentations for relevant technologies. Its topics include software algorithms; hardware-related code aspects; and software engineering topics, from any area of data processing applications such as reconstruction, simulation, online software and triggering, or data analysis and modeling. Examples include tracking algorithms, compression libraries, programming models for heterogeneous hardware and their schedulers, data acquisition software, tailored installations, firmware, software development lifecycle, and code optimization.
The seminar will take place not more than once a month; its content and timing will be coordinated with the EP-IT Data-Science Seminar and the Computing Seminar to avoid overlaps.
To nominate a speaker for the seminar please respond to this survey: HERE.