A workshop entitled “What Now in Instruction-Level Parallelism Research?” will be held on September 20-21, 2010 in Seattle, WA.  While we encourage you to submit a position paper to this workshop, you are also encouraged to post your thoughts right here on this blog! Historically, the computing industry has been driven by a set of exponential increases in single-thread performance. The ubiquity of multi-cores and the fact that much of the IT industry is relying on main-streaming parallel processing for survival is a truly seismic event. At the same time, there remains a huge gap between the theoretical limits of instruction-level parallelism (ILP) and what processors actually attain. Novel techniques […]

NSF/CISE has just announced a new program at the Interface between Computer Science and Economics & Social Science (ICES): Computational thinking has the potential to change the types of questions considered by social and economic scientists. For example, Nash (and other) equilibria lie at the heart of theories about the behavior of economic agents. Computational thinking can help characterize the range and robustness of possible equilibria and markets for which the computation of equilibria is intractable. Theories of strategic learning by computational agents, studied both in economics and computer science, can shed light on the dynamics of how agents arrive at equilibria. Theories of the spread of contagion or gossip […]

Congress mandates periodic assessments of the federal Networking and Information Technology Research and Development (NITRD) program — the program under which 14 federal agencies coordinate their investments in order to maintain America’s leadership in information technology. The history of these assessments goes back to the Brooks/Sutherland National Academies study in 1995, and includes several reviews by the President’s Information Technology Advisory Committee (in particular the landmark 1999 PITAC report), and most recently a 2007 assessment by the President’s Council of Advisors on Science and Technology (PCAST).  The impact of these assessments on the direction of the NITRD program has been significant. The  White House Office of Science and Technology Policy […]

In his major address on immigration policy on July 1, President Barack Obama noted: “And while we provide students from around the world visas to get engineering and computer science degrees at our top universities, our laws discourage them from using those skills to start a business or power a new industry right here in the United States. Instead of training entrepreneurs to create jobs on our shores, we train our competition.” Read the full text here.

I notice that CACM Editor-in-Chief Moshe Vardi’s letter in the July 2010 issue of CACM speaks to what he calls “Hypercriticality,” and cites my post of May 4 here on the CCC Blog. (You can find Moshe’s letter in CACM vol. 53, no. 7, p. 5; if you are logged into the CACM website, you can find it here.) Moshe appears to agree that we in the computing research community are often too harsh when reviewing one another’s work. (Contributed by John Leslie King, University of Michigan)

Ken Jennings, the man known for his record-breaking streak of 74 consecutive wins and $2.52 million in earnings on the popular TV quiz show Jeopardy! back in 2004, may have some competition on his hands. This Sunday’s New York Times Magazine contains an incredibly fascinating expose about “Watson,” an advanced “question answering” machine that IBM researchers have been busy developing for the past half-decade.  The story provides a step-by-step account of the challenges and research advances underlying Watson’s development — including a detailed description of how Watson works today.  It chronicles early wins — and, notably, losses — for the supercomputer versus real-life former Jeopardy! contests.  And it describes ways in which natural language processing and data mining advances […]