Distinguished Lectures Series

The ECE Distinguished Lectures Series brings world-class researchers to the University to share their research and discoveries.

Pario:  the Next Step Beyond Audio and Video

Todd C. Mowry

While audio and video technologies have had a profound impact on our lives, it is important to remember that we do not live in a world of just sounds and moving images: we live in a physical, 3D world. To enable new classes of applications that are currently unthinkable, we would like to create a new technology (which we call “pario”) that will allow us to physically render moving 3D objects as real artifacts.  Similar to how audio and video technologies allows us to capture and reproduce sound and moving images, respectively, with “pario” we could capture and reproduce the shape, motion, and appearance of arbitrary 3D objects.

At Carnegie Mellon University and Intel Research Pittsburgh, we are exploring hardware and software techniques to make this vision a reality through something that we call “claytronics”.  Claytronics is analogous to modeling clay that can control its own shape.  It is comprised of very large numbers of very tiny robots that can collectively morph into arbitrary shapes under software control.

In his talk, Professor Mowry will describe the technical progress that they have made so far on claytronics, as well as suggesting what this technology might enable.


From Communications to Radar and Back

Robert Calderbank

Today synergies with cooperative wireless communication are creating an opportunity to develop the Shannon Theory of distributed active sensing. This talk will explore how to use the available degrees of freedom of time, space, frequency and polarization to see faster, to see more finely where necessary, and to see with greater sensitivity, by being more discriminating about how we look.


GPU Computing:  Why is it exciting to many application developers?

Wen-Mei Hwu

Modern GPUs such as the NVIDIA GeForce-8 series are increasingly designed as massively parallel programmable processors. New architecture interfaces have alleviated the need for application developers to deal with graphics programming languages and interfaces. For example, CUDA programmers simply treat the GeForce 8800 GTX as a computing processor that consists of 128 processor cores, has a peak performance of 367 single-precision GFLOPS, features 86.4 GB/s memory bandwidth, contains 768MB of main memory, incurs very little cost in creating thousands of threads, and allows efficient data sharing and synchronization across subsets of threads. With millions of units already in use, it has also become arguably the largest installation of massively parallel system in history.


Virtualization Renaissance

Mendel Rosenblum

Computer system hardware-level Virtual Machines (i.e. VMs whose interface matches the physical machine) went from a vibrant research community with significant impact on the computing industry in the 1970s, to near extinction in the late 1980s and early 1990s. Today, there has been a resurgence of interest in this technology, both in the research community with multiple papers in the latest operating systems conferences, and in the commercial marketplace with Fortune 500 companies deploying the technology for their enterprise computing needs.

In this talk, Professor Rosenblum will examine some of the reasons for this resurrection by describing the attractive attributes of hardware-level virtual machines. He will show examples of how a technology best-known for running multiple, simple single-user operating system environments on a mainframe is profoundly changing how computing is done. Rosenblum will cover some recent developments in virtual machine technology that can improve the reliability, manageability, efficiency, and security of modern computing systems.


Quantum-mechanically Entangled Atoms and Raising Schrodinger’s Cat

David Wineland

Already in 1935 Erwin Schrödinger knew that, when extended to the realm of our everyday experience, quantum theory permits rather bizarre situations. To illustrate his point, he introduced his well-known cat that can simultaneously be both dead and alive, correlated with the superposition states of a radioactive particle.

These days we can create situations that have the same attributes of this unfortunate cat, although so far only on the scale of a few atoms – a very small Schrödinger kitten. These experiments might be viewed simply as quantum parlor tricks, but we now see that they might actually be useful for something. For example, the entanglement generated in a system of trapped atomic ions can be used to enhance quantum limited-metrology or might eventually lead to a quantum computer, a mesoscopic realization of Schrödinger’s cat. Progress and future prospects in these areas will be briefly described.


Metamaterial-Enabled Resonant Electrically-Small Scattering and Radiating Systems in the Microwave and Optical Regimes

Richard Ziolkowski

There continues to be a great desire for high performance electrically small radiating and scattering systems from the microwave to the optical regimes whose physical characteristics and electromagnetic responses could be tailored to satisfy a wide range of applications.  Metamaterials, artificial materials whose electromagnetic responses can in principle be engineered to any negative or positive value, have been shown recently to be a potential enabling technology for these radiating and scattering systems.

Traditional electrically small radiating and scattering systems are poor transducers of their input or excitation energy.  Several metamaterial-based configurations have been demonstrated recently that significantly improve the radiating and scattering performance characteristics of these systems.  The resulting systems are resonant despite being significantly sub-wavelength in size.  For instance, an electrically-small epsilon-negative (ENG) or double negative (DNG) spherical shell surrounding an electrically-small dipole antenna can be designed to act as an effective distributed inductor that is properly matched to the capacitive electric dipole element to form a naturally resonant LC structure, as well as to act as a resistive matching element to the source.  Thus, an overall efficiency of 100% can be achieved theoretically in such an electrically-small radiating system.  The reciprocal configuration, plane wave scattering from an electrically-small ENG or DNG metamaterial-coated sphere, has been shown to exhibit unity scattering.  Moreover, by introducing gain media, the effects of losses and dispersion can be controlled.  For instance, lasing has been demonstrated at visible wavelengths in an electrically-small metal coated nano-particle.

A review of the progress to date on all of these resonant metamaterial-based electrically small radiating and scattering systems, and their potential practical microwave and optical realizations will be given.  Practical issues, including losses and dispersion, will be emphasized in the discussion.


The Affinity Propagation Algorithm

Brendan Frey

How would you identify a small number of face images that together accurately represent a data set of face images? How would you identify a small number of sentences that accurately reflect the content of a document? How would you learn a codebook useful for quantizing speech signals? How would you identify a small number of cities that are most easily accessible from all other cities by commercial airline? How would you identify segments of DNA that reflect the expression properties of genes?

Identifying data centers, or exemplars, is an NP-hard problem, but they are traditionally found by randomly choosing an initial subset of data points and then iteratively refining it.

Professor Frey will describe a method called ‘affinity propagation’, which takes as input measures of similarity between pairs of data points. Real-valued messages are exchanged between data points until a high-quality set of exemplars and corresponding clusters gradually emerges. He will discuss aspects of affinity propagation that could impact its efficient implementation in multi-core architectures, FPGA hardware and VLSI hardware.