Krell Institute Inc

Programs: The department of energy computational science graduate fellowship program selects highly qualified students and supports their research and academic study for up to four years of their phd graduate work in a computational science and engineering discipline area that is important to the united states department of energy's office of science and national nuclear security administration. The doe csgf fellows must be either u. S. Citizens or permanent resident aliens, and as part of the program, must complete a program of study at a united states research university, spend at least twelve consecutive weeks at a department of energy laboratory working on a research project of importance to the nation, and attend an annual program review to share their research experience with other students in the doe computational science graduate fellowship program and with department of energy headquarters and laboratory research staff. During fy2017, 20 fellows completed a research practicum and 20 students were selected to receive fellowships, bringing the total number of fellows in the program to 79. Fellows were enrolled at 25 different research universities in the united states. Of the nine fellows who completed the 4th year of their fellowship this year, they have self-reported that all are continuing work on their phd, are post-doctoral students or have graduated and are working. A total of 51 alumni of the program are currently employed at doe laboratories.

the stewardship science graduate fellowship (ssgf) program selects highly qualified students in stewardship science disciplines that are important to the united states department of energy's national nuclear security administration and supports their research and academic study for up to four years of their phd graduate work. Stewardship science includes areas of study relative to high energy density physics, nuclear science, materials under extreme conditions, and hydrodynamics. The fellows complete a program of study at a united states research university, spend at least twelve consecutive weeks at a department of energy national defense laboratory working on a research project of importance to the nation, and attend an annual symposium to share their research experience with other students in the stewardship science graduate fellowship program and with department of energy headquarters and laboratory research staff. The fellows complete their twelve-week research practicum at los alamos national laboratory, lawrence livermore national laboratory or sandia national laboratories prior to the end of their 2nd year as a fellow. To participate in the fellowship, ssgf fellows must be either u. S. Citizens or permanent resident aliens. During fy2017, two fellows completed the phd fellowship program and are now employed in either postdoctoral programs, in an educational field, or at department of energy laboratories. Five new students were selected to receive fellowships, bringing the total number of fellows in the program to 25. Ssgf fellows were enrolled at 18 different research universities in the united states.

open-speedshop (o-ss) is the result of a multiple year partnership between the krell institute, the department or energy (doe) national nuclear security administration, the doe's office of science; nasa, lawrence livermore national laboratory, los alamos national laboratory and sandia national laboratories. O-ss builds on top of a broad list of community infrastructures, most notably dyninst and mrnet from the universities of wisconsin and maryland, libmonitor from rice university, and papi from university of tennessee, knoxville. O-ss is an open source multi-platform linux performance tool which is targeted to support performance analysis of applications running on both single node and large scale intel, amd, arm, intel phi, ppc, power8 gpu processor based systems and on blue gene and cray platforms. O-ss operates on existing application binaries, with no need to recompile, to gather a number of different types of performance information, relates that information back to the application source code and then displays the associated performance information back to the user. O-ss is explicitly designed with usability in mind and is for application developers and computer scientists. The base functionality includes: - program counter sampling - support for callstack analysis - hardware performance counter access and sampling - mpi profiling and tracing - i/o profiling and tracing - floating point exception analysis - posix thread analysis - memory analysis - nvidia cuda tracingin addition, o-ss is designed to be modular and extensible. It supports several levels of plug-ins, which allow users to add their own performance experiments. The key accomplishments in 2017 for the o-ss team were improvements to the scalability of o-ss through the use of the filtering capabilities of the component based tool framework (cbtf) based transport mechanism for the memory experiment, both cbtf and offline instrumentor tool support for power8, intel mic and arm processor based systems, continued build and install improvements through the use the cmake build tool and the spack install tool, and support for cbtf on ibm power platforms, updated nvidia gpu data collectors that support the new cuda and cupti interfaces, release of the new qt4/qt5 graphical user interface (gui) for all of the o-ss experiments, and improved user documentation.