Multicore Processors & Cloud Computing

SHARE:
  • EECS faculty members Hari Balakrishnan and Devavrat Shah with EECS graduate students Jonathan Perry, and Amy Ousterhout, and Hans Fugal of Facebook have devised a new system to reduce delay time in data center queues. Using Fastpass, the name given to the new system, the group has experimentally reduced the average queue length of routers by as much as 99.6 percent in a Facebook data center. Read more.
  • The potential for multicore computing on a chip has gained new traction with the work by MIT Electrical Engineering and Computer Science (EECS) Department faculty member Li-Shiuan Peh and EECS graduate student Bhavya Daya as they present a new 36-core chip on which each core acts as a mini Internet using a router to complete a communication network for data transport, while keeping local data up to date. Read more.
  • MIT's Computer Science and Artificial Intelligence Lab (CSAIL) held a two day conference celebrating 50 years of computer science looking forward to the future with solutions for today's obstacles and tomorrow's solutions. Read more.
  • Today, March 12, 2014, marks the 25th anniversary of Tim Berners-Lee's proposal for managing general information about accelerators and experiments at CERN, the European Organization for Nuclear Research based in Geneva where Berners-Lee worked at the time as a software engineer. He proposed building a distributed (global) hypertext system which he initially called "Mesh" updating it a year later to the "World Wide Web" as he wrote the code. Read more.
  • CSAIL Principal Investigator Srini Devadas and three former students have been selected as the 2014 winners of the Most Influential Paper Award at a prestigious systems research conference. Read more.
  • MIT EECS faculty members in the Computer Science and Artificial Intelligence Lab CSAIL Frans Kaashoek, Nickolai Zeldovich and Armando Solar-Lezama along with EECS graduate student Xi Wang have created a system they call Stack which will automatically scan programmer's code to avoid compilers from tossing bits of code that might not appear essential. Read more.
  • Daniel Sanchez, assistant professor in the MIT Electrical Engineering and Computer Science Department (EECS) and principal investigator in the Computer Science and Artificial Intelligence Lab (CSAIL) and his student Nathan Beckmann have developed a software alternative to manage high-speed on-chip memory (cache). Read more
  • Srini Devadas, the Edwin Sibley Webster Professor of Electrical Engineering and Computer Science, and members of the Computational Structures Group in the Computer Science and Artificial Intelligence Lab (CSAIL) have developed a new system that not only disguises a server's memory-access patterns, but also prevents attacks that rely on how long computations take.
  • Armando Solar-Lezama, EECS faculty member and head of the Computer-Aided Programming Group at CSAIL, has teamed with Sumit Gulwani, a colleague at Microsoft Research and EECS graduate student Rishabh Singh, to develop a new software system than can automatically identify errors in students' programming assignments, recommending corrections. This work could also lead to automated grading -- a big problem for MOOCs. Read more...

Pages

Subscribe to Multicore Processors & Cloud Computing