BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:A Fast General-Purpose Clustering Algorithm Based on FPGAs for Hig
 h-Throughput Data Processing
DTSTART;VALUE=DATE-TIME:20120503T180000Z
DTEND;VALUE=DATE-TIME:20120503T190000Z
DTSTAMP;VALUE=DATE-TIME:20130519T090131Z
UID:indico-contribution-34@cern.ch
DESCRIPTION:Speakers: BERETTA\, Matteo Mario (Istituto Nazionale Fisica Nu
 cleare (IT))\nWe present a fast general-purpose algorithm for high-through
 put clustering of data ”with a two dimensional organization”. The\nalg
 orithm is designed to be implemented with FPGAs or custom electronics. The
  key feature is a processing time that scales\nlinearly with the amount of
  data to be processed. This means that clustering can be performed in pipe
 line with the readout\, without\nsuffering from combinatorial delays due t
 o looping multiple times through all the data. This feature makes this alg
 orithm especially\nwell suited for problems where the data has high densit
 y\, e.g. in the case of tracking devices working under high-luminosity\nco
 ndition such as those of LHC or Super-LHC.\nThe algorithm is organized in 
 two steps: the first step (core) clusters the data\; the second step analy
 zes each cluster of data to\nextract the desired information. The current 
 algorithm is developed as a clustering device for modern high-energy physi
 cs pixel\ndetectors. However\, the algorithm has much broader field of app
 lications. In fact\, its core does not specifically rely on the kind of\nd
 ata or detector it is working for\, while the second step can and should b
 e tailored for a given application. For example\, in case of\nspatial meas
 urement with silicon pixel detectors\, the second step performs center of 
 charge calculation. Applications can thus be\nforeseen to other detectors 
 and other scientific fields ranging from HEP calorimeters to medical imagi
 ng.\nAn additional advantage of this two steps approach is that the typica
 l clustering related calculations (second step) are separated\nfrom the co
 mbinatorial complications of clustering. This separation simplifies the de
 sign of the second step and it enables it to\nperform sophisticated calcul
 ations achieving offline-quality in online applications. The algorithm is 
 general purpose in the sense\nthat only minimal assumptions on the kind of
  clustering to be performed are made.\n\nhttp://indico.cern.ch/contributio
 nDisplay.py?contribId=34&sessionId=6&confId=154525
LOCATION:INFN Pisa
URL:http://indico.cern.ch/contributionDisplay.py?contribId=34&sessionId=6&
 confId=154525
END:VEVENT
END:VCALENDAR