Recently, supercomputers have entered the era of exascale beyond petascale. Overcoming the exascale barrier, which is the biggest problem with a large number of computing nodes and energy efficiency, was possible by using GPUs for computation in artificial intelligence, machine learning, and various high-performance computations. As evidenced by this, seven out of the ten world’s highest performing supercomputers are heterogeneous computing systems using GPUs. Accordingly, we developed a tool that can easily and conveniently monitor multiple vendors GPUs to support heterogeneous computing systems that are widely used in artificial intelligence and high-performance computations.