Lecture Utz Haus (Cray)

The N&O group would like to invite anyone who is interested to attend the lecture of Utz Haus, entiteld: Challenges at the Interface of Optimization and High Performance Computing and Analytics Abstract:For decades we have been able to rely on Moore’s law and Denard scaling to offer faster computing resources transparently with each new
  • Lecture Utz Haus (Cray)
  • 2016-11-21T15:00:00+01:00
  • 2016-11-21T16:00:00+01:00
  • The N&O group would like to invite anyone who is interested to attend the lecture of Utz Haus, entiteld: Challenges at the Interface of Optimization and High Performance Computing and Analytics Abstract:For decades we have been able to rely on Moore’s law and Denard scaling to offer faster computing resources transparently with each new
  • When 21-11-2016 from 15:00 to 16:00 (Europe/Amsterdam / UTC100)
  • Where Room L120, first floor CWI, Science Park 123 Amsterdam
  • Web Visit external website
  • Add event to calendar iCal

The N&O group would like to invite anyone who is interested to attend the lecture of Utz Haus, entiteld: 
Challenges at the Interface of Optimization and High Performance Computing and Analytics

Abstract:

For decades we have been able to rely on Moore’s law and Denard scaling to offer faster computing resources transparently with each new hardware generation. In recent years this has changed, and pervasive parallelism is the only avenue towards the exascale computational devices (10^18 FLOP/s) expected in the next 5 years in a non-quantum paradigm. At the same time we are seeing that data movement is now more expensive than floating point operations — its energy consumption will be the limiting factor. While from a view of computational complexity this does not change algorithmic complexity, we argue that practically it makes all the difference.
In this talk we will discuss challenges and opportunities arising when using high-performance computing systems to solve optimization problems, in particular in the context of (graph) analytics and big data, stochastic optimization, and large-scale mixed-integer programming.