With the advent of the modern subsea observatory, two basic system architectures have emerged enabling the scientific community to study the oceans in an efficient manner for the region of interest, providing cost efficient, reliable, and proven solutions. Ethernet and optical networks dominate subsea communications architectures as utilized on programs such as MARS, VENUS, NEPTUNE, and NEMO due to the already developed and proven telecommunications hardware used in each application. With the two basic hardware configurations these observatories employ, a standard configuration has been generated that may enable experiments to be moved between observatories. This simple function of interconnectivity for experiments between observatories enables the study of oceanographic events on a global scale for the first time in history while creating efficiencies for funding. With currently planned hardware, a single experiment may study at least four different geographic regions with a minimum of three more locations available in the near future.The three main factors that define the architecture requirements of a modern subsea observatory are power, communication rate, and the distance required between regions of interest. Power defines the abilities of the node to support high wattage equipment like lights and other sensor arrays, while also affecting step-out distance due to the decrease in voltage based on resistive loads. Communication rates are vital as they define the systems ability to support instruments with high bandwidth such as cameras and remain competent over their design life spans. As the need increases for subsea information, so will the requirement for data rates. Observatories need to be able to upgrade throughout their lives to ensure they are able to meet the needs of future experiments. Subsea observatories also maintain the capability to be expandable into an entire subsea local area network (LAN) of sensors creating a need to evaluate step-out distances and repeaters. Ethernet-based communications have a maximum length of 70 meters making them ideal for specific regions of interest. As some applications have found, the regions of interest can span many more square kilometers than is logistically viable for Ethernet-based communications architectures to support, leading to the need for an opticallybased communications architecture which is able to support much longer step-outs. Using a standard set of existing equipment that is designed for, and can be configured for, each specific application's environment will minimize the cost and improve the reliability of the system by removing the need for engineering and qualification of new and unproven hardware. This paper will present an overview of the latest generation of subsea observatories; focusing on the system architecture and discussing enabling interconnect hardware currently available, as well as highlighting future possibilities. Current technologies in each of the three main factors will be explored and discussed with special emphasis on how t...