Two broad classes of memory models are available today: models with hardware cache coherence, used in conventional chip multiprocessors, and models that rely upon software to mana...
John H. Kelm, Daniel R. Johnson, William Tuohy, St...
Writing reliable software is difficult. It becomes even more difficult when writing scientific software involving floating-point numbers. Computers provide numbers with limite...
Instruction set customization accelerates the performance of applications by compressing the length of critical dependence paths and reducing the demands on processor resources. W...
Sami Yehia, Nathan Clark, Scott A. Mahlke, Kriszti...
An important reality when integrating scientific data is the fact that data may often be "missing", partially specified, or conflicting. Therefore, in this paper, we pre...