Customer Reviews

Parallel Programming in OpenMP

5 star:		(0)
4 star:		(5)
3 star:		(0)
2 star:		(0)
1 star:		(0)

 

 

Chandra et al. have put together a readable, helpful introduction to parallel programming with OpenMP. Unlike its major competitor, MPI, OpenMP assumes a shared memory model, in which every processor has the same view of the same address space as every other. At least as a start, this cuts the intellectual load way down. The programmer adds just one concept to the problem, parallelism, without adding buffering, communication networks, and lots of other stuff as well.

After two introductory chapters, the authors introduce OpenMP in three stages: loop parallelism, general parallelism, and synchronization, roughly in order of increasing complexity. The authors present the necessary OpenMP pragmas and APIs at each step, showing how they address the immediate problems. An appendix summarizes the pragmas and APIs, in both their C/C++ and Fortran forms. OO C++ programmers may be dismayed by the amount of attention paid to an un-cool language like Fortran, but need to realize that it's still the lingua franca of performance programming. And, in fairness, the authors spend equal time on C++ idiosyncrasies, such as constructor invocations for variables that are silently replicated in each of the parallel threads.

If you've ever done performance programming, you're groaningly aware that getting the parallelism right is actually the easy part. The tricky parts come in breaking dependencies, in scheduling, in ensuring spatial and temporal locality, and in dealing with cache coherency issues of multiprocessors. The authors give great introductions to all of the basics. This includes a patient description of how caches actually work, since there's a new crop of beginners every day. The authors describe performance analysis tools, but only briefly. The tools differ so much between vendors and between one rev and the next, that any detailed description would be useless to most readers immediately and obsolete for all readers very soon.

This won't turn a beginner into the guru of performance computing. It will, however, establish a working competence in one popular parallelization tool, OpenMP, and in the computing technologies that affect parallel performance.

//wiredweird

This book has many examples of how to parallelize Fortran 77 programs with loops using OpenMP directives, but coverage of how to parallelize Fortran 95 code using array operations is sparse. For this, one should read the tutorial "Parallel Programming in Fortran 95 using OpenMP", by Miguel Hermanns, available at the OpenMP web site.

I found this book to be a well written ground up how to on OpenMP. It is approachable by someone not well versed in parallel programming. I believe it was written before the wide scale advent of multi core architectures and in those pre multi core days most users of OpenMP would have been in the scientific community and have been interested largely in speeding up fortran codes. So the focus on the fortran constructs is understandable. However in todays world with every desktop equipped with a multi core cpu the book would be better with a stronger focus on c++. Despite the heavy focus on fortran examples the book does include information on using OpenMP from c++. I rate the book highly, because of its clarity, approachability, and style and hope future editions have a stronger showing of c++ examples.

Hoping for just information on OpenMP, I was pleased to find much information about issues with parallelizing algorithms. In fact, OpenMP itself is actually very tiny, easily fitting on a few quick reference cards. Applying OpenMP, or any multithreading for that matter, is what actually determines success. I was particularly pleased with the section on cache lines and their impact on design.

this is probably the first book about OMP. The author has decribed the uses of many functions and directives of OMP. The examples (in Fortran and C) given are also useful. Generally this is a good book to get you started off with OMP.