The summary (shown in the following figure) provides an overview of the RAM primitives used in the design and how many of them are inferred (or instantiated using XPMs).
The Summary is broken down by primitive type and the overall utilization of these primitives is also shown. For distributed RAM, it is reported as used Memory LUTs (LUTMs), not the total number of Distributed RAM primitives. This can help identify if a lot of inefficient DRAM primitives are in the design (typically older designs or designs with a lot of instantiation may exhibit this). For example, the cost per bit on RAM32X1D can be as high as 2 LUTs per bit (or 1 LUT per bit if LUT combined), but a RAM32M16 can be around ½ LUT per bit if all bits are used.