Acknowledgements

We are grateful to customers and Intel engineers who have helped us in our quest to make this book happen.

Thank you, Scott McMillan, for writing the chapter on MPI. Your insights and writing skills are greatly appreciated.

Thank you to Robert Harrison, Director, Institute for Advanced Computational Science at Stony Brook University, for his support, expertise, and passion, and for sharing his unique insights in the preface.

We wish to thank several people for the documentation or white papers that they had written which we used as source material. The quality and quantity of material we had to start with was a big help in writing this book in a timely manner: Ron Te and Mike Casscles for the System Software Developers Guide; Clayton Craft for the Instruction Set Architecture document; Robert Reed and Shannon Cepeda for the stencil code in Chapter 4; Shannon Cepeda and Wendy Doerner for vectorization training and papers; Martyn Corden for his articles about vectorization; Shannon Cepeda, Larry Meadows, and David Mackay for tuning papers; Johnnie Peters with Alexandra Yates on Linux configuration information in Chapter 10; Rakesh Krishnaiyer for advice on prefetching; Taylor Kidd for countless training documents that highlighted the key points and many items worth special mention; Raymundo Vazquez for information regarding timers; information regarding cluster configurations from Michael Hebenstreit, Romain Dolbeau, Jeremy C. Siadal, Andreas Hirstius, Andres More, Clem Cole, Matias Cabral, and Werner Krotz-Vogel.

We want to extend our gratitude to the great work done by the Intel SSG Information Development team, who has written outstanding product documentation that was both educational for us and useful source material. Our apologies if we fail to mention everyone who contributed; we thank the whole team, which included efforts from Mitch Bodart, Ivan De Cesaris, Kevin Davis, Rajiv Deodhar, Karen Dickinson, Wendy Doerner, Stephen Goodman, Ronald Green, Elena Indrupskaya, Irina Kosareva, Rakesh Krishnaiyer, Lorri Menard, Sabrina Mesters-Woell, Ravi Narayanaswamy, Sunil Pandey, Christopher Raleigh, Premanand M Rao, Amanda Sharp, Michael Sturm, Michael Toguchi, Miwako Tokugawa, Stan Whitlock, and Anatoly Zvezdin.

Many people associated with the Knights Corner project gave input and feedback that was especially valuable to the first chapter, including Mani Anandan, Andrew Brownsword, George Chrysos, William Clifford, Kim Colosimo, Charles Congdon, Joe Curley, Ervin Dehmlow, Pradeep Dubey, Gus Esponosa, Robert Geva, Milind Girkar, Ron Green, Michael Greenfield, John Hengeveld, Scott Huck, Jim Jeffers, Mike Julier, Rakesh Krishnaiyer, Belinda Liverio, David Mackay, Bill Magro, Larry Meadows, Chris Newburn, Brian Nickerson, Beth Reinders, Frances Roth, Bill Savage, Elizabeth Schneider, Gregg Skinner, Sanjiv Shah, Lance Shuler, Dancil Strickland, Brock Taylor, Philippe Thierry, Xinmin Tian, Matt Walsh, and Joe Wolf.

Many of people named thus far helped review parts of the book. We are grateful to them and others who gave their time to help improve this book, including Rob Fowler of RENCI/University of North Carolina, Neil Stringfellow and Jeff Poznanovic of CSCS/Swiss National Supercomputing Centre, Kevin Thomas and James L. Schwarzmeier of Cray Inc. and Clayton Craft, Jeanette Feldhousen, Michael Hebenstreit, Elena Indrupskaya, Hans Pabst, Steve Lionel, Susan Meredith, Xinmin Tian, Zhang Zhang, and Georg Zitzlsberger of Intel, and James A. Ang of Sandia National Laboratories.

Heinrich Bockhorst, Kathy Carver, Jim Cownie, Loc Nguyen, Klaus-Dieter Oertel, Dmitry Sivkov, Alexander Supalov, and James Tullos provided valuable feedback on Chapter 12.

Andrey Vladimirov of Colfax Corporation is an inspiration with his insights and training materials and has provided valued feedback for this book including some code examples used in Chapter 7.

To the team at the Texas Advance Computing Center (TACC) for valuable feedback on the book and the Intel Manycore program, plus elements of the stencil algorithm used in Chapter 3, including Kent Milfeld, Dan Stanzione, Bill Barth, Karl Schultz, Tommy Minyard, Jim Brown, John McCalpin, Lars Koesterke, and TACC Director, Jay Boisseau.

Glenn Brook, Ryan Braby, Ryan Hulguin, Vince Betro, and many others at the National Institute of Computational Science (NICS) for their feedback on the book and their excellent work and feedback using coprocessors for science.

Special thanks from James Reinders to his wife Elizabeth and daughter Katie who were supportive of the long hours writing. Beth Reinders was the notable talent behind guiding Chapter 1 to make sense and her insights into the critical points that needed to be crisp for a broader audience especially for challenges that we had with Chapters 1 and 8. Finally, my coauthor and friend, Jim Jeffers, who was the perfect partner to getting this book out with his talents for teaching, understanding, questioning, and hard work always with a sense of humor even the midst of customer crisis and a major product launch (why did we agree to write this book in our spare time?).

Special thanks from Jim Jeffers to his wife Laura and his children Tim, Patrick, Colleen, Sarah, and Jon for their support, encouragement, and consideration over weekends, evenings, and absences while traveling and working on the book. Also, to my longtime friend and colleague, Joe Curley, for bringing me back to Intel to contribute to this ambitious and exciting program. To Nathan Schultz for his support and encouragement, especially during the final push to complete this book. Finally, to James Reinders for his guidance, expertise, focus, and friendship in driving this book to completion.

For their time, support, and expertise we would like to thank George Chrysos, Joe Curley, Stuart Douglas, Herb Hinstorff, Michael Julier, Michael McCool, Chris Newburn, Elizabeth Reinders, Nathan Schultz, and Arch Robison.

The Intel Press team, in particular David Clark and Stuart Douglas, did a tremendous job helping us complete this book. Dan Mandish for his fine artwork. We appreciate the hard work by the entire Morgan Kaufmann team including the three people we worked with directly: Todd Green, Lindsay Lawrence, and Mohanambal Natarajan.

Numerous colleagues offered information, advice, and vision. Many students took classes from us and their feedback improved the quality of our teaching material that in turn fed into this book. There are certainly many people who have helped directly and indirectly and we failed to mention. We thank all those who helped and we apologize for any who helped us and we failed to mention.

Thank you all,