The natural stone cutting process is carried out by cutting tools called diamond socket (DS) consisting of two parts diamond and matrix. DS performance in rock cutting highly depends on the production cost of the socket, cutting precision, and energy costs. However, DS performance can be demonstrated after the circular saw with DS is connected to the cutting machine and performs the cutting operation. This trial and error method causes negative effects such as time loss, an increase in labor costs, and disruption in production for natural stone processing plants. For this reason, ultrasonic velocity measurement (UVM) processes are performed on DS to predict DS performance. Within the scope of the study, eight groups of boron carbide (B4C)-doped DS are produced. In the one group of these DSs, no-doped is made, and therefore it is considered as a reference socket. The other seven groups are B4C-doped at different rates. Metallographic parameters (MPs) (unit volume weight, porosity, Knoop hardness, weight wear loss) of DSs produced before cutting experiments are determined. Then cutting performance parameters (CPPs) (specific cutting energy, power consumption, specific abrasion) of circular saws prepared with DS groups are obtained by performing cutting experiments on granite samples. After ultrasonic velocity (UV) in diamond sockets is measured experimentally, the correlations between UV and MPs and CPPs are analyzed. As a result of the experimental analyses, it has been determined that there is a significant correlation between these parameters and UV. Based on these correlations, it has been determined that the UVM method can be used both in the quality control of DSs and in the prediction of cutting performance.