Intel Tech Talks :: 23Oct07
Moore’s Law is going strong. Transistor densities will keep climbing well into the next decade (or longer), but a number of issues (power density and leakage current to name two) make it impractical to use those higher densities to speed up the chip’s clock. How is Intel going to turn those transistors into increased performance? By using parallelism and putting multiple cores on a single chip.
Saying you’re going to put many cores on a chip is easy, but how you actually do it is another matter altogether. Once you have the cores on the chip, how are you going to write software for the many core CPU? To explore a host of many core issues, Intel created an 80 core test chip that is running a real application kernel ran at over one trillion single precision floating point operations per second. In this talk, two developers from this effort will talk about the project and the lessons it holds for the future of many core chips; Nitin Borkar (the overall leader of the project) and Tim Mattson (lead the software development team).
Dr. Tim Mattson is a Principal Engineer at Intel’s Microprocessor Technology Labs. He joined Intel in 1993. Among his many roles at Intel, he was applications manager for the including the world's first TeraFLOP supercomputer (ASCI Red). He helped create OpenMP, founded the Open Cluster Group (OSCAR), and launched Intel’s programs in computing for the Life Sciences. Dr. Mattson earned a PhD. for his work on quantum molecular scattering theory (UCSC, 1985). This was followed by a Post-doc at Caltech where he worked on the Caltech/JPL hypercubes. Currently, Dr. Mattson is conducting research on performance modeling for future multi-core microprocessors and how different programming models map onto these systems. He is a co-author of the book “Patterns for Parallel Programming.”
You can read more of Tim's thoughts on his Blog .
Nitin Borkar is engineering manager of Intel's Microprocessor Technology Lab.