|
 |
Collaboration
Robert Allen and Virginia Tech (IBM Almaden)
Collaboration with Prof. Tim Long, Polymer Chemistry, Virginia Tech involving the investigation of the polymerization kinetics and mechanism of cyclic olefin/maleic anhydride copolymers. These are materials of high importance to 193nm lithography, and the polymer synthesis has been shrouded in mystery. Tim's group has employed a variety of in-situ characterization techniques to gain a stronger fundamental grasp of the polymerization process used to make these important materials.
Robert Allen and Cornell University (IBM Almaden)
Collaboration with Prof. Chris Ober, Material Science, involving the study of the morphology of polymer/onium salt mixtures. A key component in today's photoresists are ionic photoacid generators (onium salts). Depending on the structure of the polymer and the salt, the mixture can be quite inhomogeneous and exhibit significantly non-uniform depth dependence. RBS and other complimentary techniques are being used to understand these complex systems. Additionally, we are working together on the use of carbon dioxide (liquid or supercritical) as an alternative developer of photoresists.
More on their work
Zaphiris Christidis and Florida State University (IBM Watson)
Christidis has worked closely with FSU and Prof. Krishnamurti on Tropical Meteorology, and Hurricane prediction. The past work has resulted in the establishment of the Real-time Hurricane Center (RHC) at Florida State University, which has provided Hurricane predictions for the Atlantic Ocean for three years. They have done major breakthroughs in this area, by co-developing a hurricane prediction model, which is the centerpiece of the RHC. They have recently moved their collaboration in the area of the Superensemble Weather Prediction (which was widely publicized in the media last year) with more work to do this year.
Andrew Conn, Katya Scheinberg and University of Coimbra, Portugal (IBM Watson)
Dr. Conn and Dr. Scheinberg are working with Prof. Luis Vicente on the properties of the sample sets in derivative free optimization -a branch of optimization that studies methods that do not use the derivatives of the functions. In particular, they derive conditions for multivariate polynomial interpolation and regression on the set of sample points to bound the error between the model and the function that is being approximated. They have also proposed methods of improving the quality of the interpolation set to reduce such error bound. They are currently working on a book on derivative free optimization. Prof. Vicente spent the 2002-2003 academic year as a visitor in the mathematical sciences department.
Barbara Jones and Georgetown University (IBM Almaden)
Jones collaborates with Georgetown's Jim Freericks and Amy Liu on projects involving first-principles theories of multilayered nanostructures, especially those with dissimilar materials. The three have collaborated on research that which shows the mix of spin density into an insulator from the proximity of a Co layer in magnetic tunnel junction structures. These results help to understand the spin polarized current and voltage characteristics of MTJ's, which have been theoretically controversial. Magnetic tunnel junctions are prime components of spintronics, in which the spin of the electron is transported rather than, or in addition to, the charge. The three have collaborated as PI's on several previous NSF grants, which have sent students to Almaden for one-year internships. Jones is also on the Advisory Board of Georgetown's graduate physics department. Recently the three have discussed extending their work to the area of molecular electronics with Almaden's Campbell Scott, and there is an NSF grant proposal with all four submitted pending approval.
More on this work from Freericks.
Barbara Jones and Boston University (IBM Almaden)
Jones collaborates with Boston U.'s Antonio Castro Neto and his previous student Chiung-Yuan Lin, now a postdoc at Almaden, on projects involving many-body theory of strongly correlated electron systems. In particular, they are interested in calculating the response of a Scanning Tunneling Microscope (STM) to the presence of a magnetic atom on a metallic surface, especially when a many-body Kondo effect (low-temperature screening of the magnetic atom) occurs. Using first-principles calculations of the surface and bulk states of the metallic host and atomic theories of the magnetic atom, combined with realistic potentials for the STM tip, they are able to calculate detailed properties of the STM/atom/surface structure.These have been shown to compare very well with experiment, and are hoped to be able to guide future nanostructures constructed by STM atom moving. The three are collaborating with Almaden's STM group, headed by Andreas Heinrich, to further expand their theory. Jones and Castro Neto have also collaborated on so-called Griffiths phases in metallic systems with a high density of magnetic impurities. When Kondo effects and magnetic interactions combine, a state can form in which rare large islands of magnetic atoms dominate the low-temperature behavior. Tunneling effects in these systems also help us learn about tunneling in systems of STM.
More on this work from Castro Neto.
Barbara Jones and Stanford University (IBM Almaden)
Jones collaborates with Stanford's Shoucheng Zhang, shared graduate student Tzen Ong, and previous student Yaroslaw Bazaliy on projects involving theories of quantum interactions in nanoscale magnetic systems. The four have collaborated on research that predicts the effects of spin polarized current through a magnetic material (which has recently received increased experimental and theoretical attention). This work is especially applicable for small nanoscale magnetic systems, such as will likely be used for devices in the future. Barbara recently became a member of the newly formed Center for Probes on the Nanoscale, an NSF-funded center at Stanford. Barbara has a new postoctoral researcher through this center, Chiung-Yuan Lin. In two separate projects, Barbara and Chiung-Yuan, and Barbara and Tzen are collaborating to calculate and understand in depth the effects of magnetic atoms, in clusters or nanolattices, on a metallic surface. This work aims to understand and make predictions about experiments done by Andreas Heinrich's group at IBM, and Hari Manoharan's group at Stanford.
More on this work from Zhang and Manoharan.
Clare-Marie Karat and University of Maryland (IBM Watson)
Dr. Karat is working with Dr. Andrew Sears of the University of Maryland - Baltimore County as principal investigators on a project funded under the Universal Access Program of the Computer and Information Science and Engineering Division of the National Science Foundation. "Using Speech Recognition to Enhance Communication Capabilities for Individuals with Disabilities" will model the interaction between users and CSR systems to develop a usable and efficient speech recognition user interface for individuals with disabilities.
Hillel Kolodner and Tel-Aviv University (IBM Haifa)
Hillel Kolodner collaborates with Mooly Sagiv at TAU and his students. Their research aims to apply static analysis and compilation techniques in order to improve the performance of automatic memory management (a.k.a garbage collection). Joint work on static and dynamic heap liveness has led to techniques and tools that identify potential space savings on top of what's obtained by a garbage collector. Such information can be used by software developers to perform simple source rewritings that allow savings. In addition, they have developed static algorithms that identify program points at which memory can be deallocated and when heap references will not be used further. Such information can be used by a garbage collector to automatically collect more space. This work has been the basis for Ran Shaham's PhD thesis and joint papers that have appeared in ISMM 2001, PLDI 2001, ISMM 2002 and SAS 2003. Currently Mooly and Hillel jointly supervise Ron Morad, a Masters student.
More on their work
Isidore Rigoutsos and Princeton University (IBM Watson)
Since the fall of 1999, Dr. Rigoutsos and Prof. Thomas Shenk (from the Department of Molecular Biology) have been collaborating on the application of computational methods to the study and analysis of the Human Cytomegalovirus (a.k.a Human Herpesvirus 5). They have already jointly published three journal papers.
Isidore Rigoutsos and Massachusetts Institute of Technology (IBM Watson)
Dr. Rigoutsos and Prof. Gregory Stephanopoulos (from the Department of Chemical Engineering) have been collaborating on applications of pattern discovery to the analysis of biological data. They co-supervise several PhD candidates, and have been teaching a spring semester and a summer professional course, both on Bioinformatics, since January of 2000. Rigoutsos and Stephanopoulos recently published a joint paper and are in the process of finishing and submitting several more for publication. The two have recently accepted an invitation by Springer-Verlag to edit a book as part of their "Advances in Biochemical Engineering/Biotechnology" series.
Katya Scheinberg and Columbia University (IBM Watson)
Dr. Scheinberg and Prof. Goldfarb are working on properties of the linear systems arising in the interior point methods for continuous convex optimization. A new property of Cholesky factorization of the matrices in these systems was proved to hold. They proposed and analyzed new methods of handling systems of linear equations that are dense due to the presence of dense columns in the original data or due to the structure of the problem. Prof. Goldfarb was an academic visitor to the mathematical sciences department in 2003 and 2005.
Campbell Scott and University of California, Santa Cruz (IBM Almaden)
Scott and Prof. Sue Carter in the Physics Department are determining the transport properties of conjugated polymers. The results provide the input parameters for numerical simulation of the operating characteristics of organic light emitting diodes, photovoltaic cells and memory elements.
More on their work
Frank Tip and Rutgers University (IBM Watson)
Dr. Tip is collaborating with Barbara Ryder and several students on a collaborative project between Rutgers and IBM Research on "Change Impact Analysis of Object-Oriented Systems". The goal of this project is to provide programmers with feedback on the semantic impact of changes made to applications that they are developing, and help them isolate those changes that are responsible for test failures. This project is supported by NSF grant CCR-0204410 with Barbara Ryder (PI), and Frank Tip (co-PI).
Andreas Waechter and Carnegie Mellon University (IBM Watson)
Joint research with Prof. Lorenz T. Biegler, Department of Chemical Engineering at Carnegie Mellon University. The topic of this research is the development of an interior point algorithm for large-scale nonlinear optimization, and its implementation (under the name IPOPT) as open source software, available at the IBM-initiated repository www.coin-or.org. The code can be used as a general-purpose solver, for example for optimization problems formulated with the modelling language AMPL. Further research will include the efficient exploitation of problem structure occurring in certain engineering applications, such as nonlinear optimal control.
Andrew Conn, Jon Lee and Andreas Waechter and Carnegie Mellon University (IBM Watson)
Joint research with Prof Lorenz T. Biegler and Ignacio E. Grossman, Department of Chemical Engineering and Gerard Cornuejols and Francois Margot, Tepper School of Business at Carnegie Mellon University. The topic of this research is the development of algorithms and software and their application to business, logistics and engineering optimization problems that have discrete and nonlinear aspects. The general idea is to leverage the outstanding, and in many ways complementary, expertise at CMU and Watson, to make significant progress in this difficult and important area. Our goals are (i) to make fundamental advances, (ii) to solve difficult applied problems, and (ii) to develop reusable software.
Israel Wagner and Technion - Israel University of Technology (IBM Haifa)
Wagner has a joint project with Prof. Freddy Bruckstein on "Ant Algorithms for Efficient Traversal of Faulty Networks" and other ant-inspired algorithms for robotics and search.
More on their work
Greg Wallraff and University of Colorado (IBM Almaden)
A collaboration with Prof. J. Michl studying the acid catalyzed reactions that occur in polymer films of a new class of silicon based photoresist developed by the lithography group at Almaden. This resist is being evaluated for use in advanced DRAM and Logic (microprocessor) applications employing KrF (248 nm) lithography.
|
|
|
|
|
