Issue 2 - The Parallel Universe - May 2010

Editorial Intro - James Reinders

It is no surprise that Intel® Parallel Studio is consistently cited in articles and analyst reports about software that helps with parallel programming. Of course, Intel has a long history with high performance computing (HPC) through efforts with OpenMP* and MPI*. Recently, Intel led non-HPC efforts on projects like Intel® Threading Building Blocks (Intel® TBB) and Intel® Parallel Studio.

Offering real assistance for multicore programming has meant addressing modification of existing applications, addressing ease-of-use issues, and supplying tools to aid in the entire process of designing a program. To these ends, Intel started with Intel TBB to address the many challenges of using C++ - designed in an age of singleprocessor systems - in a multiprocessor and multicore processor world.

Encouraged by the success of Intel TBB, Intel addressed the whole design cycle with Intel Parallel Studio. By not just focusing on the alluring topic of language extensions for parallelism, Intel has made significant progress on debugging and tuning issues that have long perplexed parallel programming. Beyond helping debug the most common parallel programming issues, deadlocks, and data races, Intel found key advances to address memory leak errors, which have proven harder to debug in a parallel program.

Serious advances here have proven immensely helpful for development teams striving to have predictable schedules and results. It is not a stretch to say that without these new tools from Intel, a foray into parallel programming is much less likely to succeed.

Legacy, education, and tools were the three key needs identified in a recent joint Intel-Microsoft* customer roundtable on parallelism that I was fortunate enough to attend. Intel's alignment with our activities and these needs are the strongest that I see in the industry.