From: BPB Publications firstname.lastname@example.org
To: "'DR. N. B. Venkat Nagasuri'"
Date: Wednesday, December 8, 2010, 4:14 PM
We believe we had a chance to go through the book and any views about the same.
Comments on the book “Mutli-core Programming Increasing Performance through Software Multi-threading”, by Shameem Akhter and Jason Roberts, Intel Press, Via BPB Publishers, New Delhi, India.
Prof NB Venkateswarlu email@example.com
AITAM, Tekkali, India
This is the first book which I read on Multi-core processors. As such I feel it is very nicely written book. Authors gave nice introduction to general computers (single core) and then switched to multi-core computers. I found explanation regarding Gustafsan’s law is very illuminating which explains the essence of multi-core systems.
A separate chapter on threads is very attractive for a new student who is interested to switch over to the multi-thread based programming. It gives nice introduction different threading models with examples from practical implementations. Also, it introduces the reader to virtual machines, the art of today and tomorrow.
Main bottlenecks in thread based parallel programming in conjunction with synchronization is emphasised very nicely. It is nice to see that authors have discussed both with respect to pthreads, Windows and Java threads.
Also, authors introduced to practical multi-core programming using OpenMP and MPI. No doubt that the authors have done full justice to reach a new intern to parallel programming. However, it would have nice if they have given more hints about real problems what they have solved in their respective organizations.
Separate chapters on debugging multi-core programs and Intel Software development tools is illuminating but requires some more technical tint.
Over all, this book can be a handy one for a student who wants to venture into multi-core programming. Some tit bits about what should be the expected background of the reader along with some reading assignments would have made this as more attractive.