Page 12 of 15

CONCLUSION

In this paper, we provided an overview of the Intel IA-64 compiler. We described the organization of the compiler, as well as the features and functionality of several optimization techniques. The compiler applies region and loop-level control and data transformations, as well as global optimizations, to programs. All the optimization techniques in the compiler are aware of profile information and effectively use interprocedural analysis information. The optimizations effectively target three main goals while compiling an application: i) to minimize the overhead of memory accesses, ii) to minimize the overhead of branches, and iii) to maximize instruction-level parallelism. We described how the optimization techniques in the Intel IA-64 compiler take advantage of the IA-64 architectural features for improved application performance. We illustrated the techniques with example codes, and we highlighted the benefits as a result of specific optimizations. The Intel IA-64 compiler incorporates all the infrastructure and technology necessary to leverage the IA-64 architecture for improved integer and floating-point performance.