Recently, with the development of Internet of Things (IoT) technologies, the amount of information processed in the cloud is getting larger. This situation can cause a delay in data transfer and an increase in power consumption. Therefore it may cause negative effects on any information processing, like machine learning. Physical reservoir computing (PRC) is a solution, where we can use physical dynamics as a computational resource to reduce the amount of power consumption. In this report, we propose to utilize a surface acoustic wave (SAW) device for PRC, and for this purpose, we designed and fabricated SAW phase modulators that have a waveguide to modulate a SAW efficiently. We prepared devices with waveguide lengths of 5 mm, 10 mm, and 15 mm and widths of 10 μm and 18 μm and evaluated signal attenuation and modulation effects. As a result, we clarified the effect of these parameters to the properties of SAW modulators.