InTheLoop | 06.29.2015
Donofrio to Lead Computer Architecture Group
David Donofrio—a veteran Computer Systems Engineer at Berkeley Lab—has been selected to lead the new Computer Architecture Group in the Computational Research Division (CRD). The group’s mission is to research, design and implement new architectural simulation capabilities, novel hardware architecture features, assess design trade-offs, and facilitate the co-design process for leading-edge high performance computing (HPC) systems.
John Shalf, who heads CRD’s Computer Science Department, notes that Donofrio’s long-standing interest in computer architecture and performance, as well as his understanding of the computing problems at Berkeley Lab and the Department of Energy make him an ideal candidate to lead the Computer Architecture group.
“Berkeley and Sandia are the National Labs leading the architecture part of DOE’s exascale research portfolio, and the Computer Architecture Group is central to this work. Dave’s experience and leadership will be a great asset to the team,” says Shalf. »Read more.
ESnet's Monga to Address 'Programmable Cities' Conference
Inder Monga, ESnet's division deputy for technology and chief technology officer, will address a day-long workshop on "Programmable Cities" on Tuesday, July 7 in Bristol, UK. The city is building what it calls the world's first programmable smart city, designed to enable creative interaction between people and things, and not just machine to machine. During the workshop on software defined networking (SDN), Monga will address the gathering about the role that SDN can play in programmable smart cities. ESnet has been a key supporter and developer of this open networking technology.
This Week's CS Seminars
Summer Student Talk: Using Coopetition to Advance Protein Folding
Thursday, July 2, 11:00am-12:00pm, Bldg 50B Rm 4205 (Remote: OSF 254)
Silvia Crivelli, LBNL and UC Davis
Protein folding is one of the biggest challenges in modern biology. It is about predicting the shape of proteins, which determines their function. This problem is highly multidisciplinary and it is also extremely hard. A competition called CASP was created 20 years ago to advance and assess methods for protein structure prediction. CASP takes place every other summer and the results show that progress has slowed down significantly for the past 10 years. Recently, a new approach was started that uses social media and a science gateway to bring together scientists from all over the world to collaborate and compete. Although it is too early to draw any conclusions about the impact of coopetition in the solution of the protein folding problem, it is clear that the experiment has started a truly multidisciplinary community. In this talk I will discuss how coopetition is broadening the community that tackles the field and how it has inspired students to learn about computational science.
A video conference feed will be available to summer students working at NERSC in OSF in conference room 254.