VecMeter: Measuring Vectorization on the Xeon Phi

Wide vector units in Intel's Xeon Phi accelerator cards can significantly boost application performance when used effectively. However, there is a lack of performance tools that provide programmers accurate information about the level of vectorization in their codes. This paper presents VecMeter, an easy-to-use tool to measure vectorization on the Xeon Phi. VecMeter utilizes binary instrumentation and therefore does not require source code modifications. This paper describes the design of VecMeter, demonstrates its accuracy, defines a metric for quantifying vectorization, and provides an example where the tool can guide code optimization to improve performance by up to 33%.