The Accessibility of Data
DATA NEEDS TO BE ACCESSIBLE
For data to be valuable it must be accessible. By 2030, it is estimated there will be 25.4 billion smart devices in service with a projected population of 8.5 billion so there will be approximately three smart devices for every person on the planet. These devices will range from smartphones to smart cars. While all of these devices are wireless, the data they will be accessing will be delivered from a data center.
One of the biggest issues to overcome is latency (data delivery time) from the data source to the device. Anyone that has a smartphone today generally sees almost instant data transfer most of the time, but everyone has also experienced frustration when they have delays in that data transfer. Imagine a smart car that needs data to navigate and to communicate. Latency to a smart car is more than a convenience, it is necessary for life safety. Solving the latency issue is one of the immediate challenges we will be facing as data center designers.
One of the ways this is being addressed is by placing small remote data centers closer to the user. While this will improve latency it becomes problematic. Most of these remote data centers will be small, unoccupied facilities that will present security concerns. These facilities will be carrying data that must be secured and we do not want to improve latency at the cost of the security of the data. Maintenance is also a concern given the facilities will not have an onsite facilities team like the larger data centers.
One of the cities currently leading the way to help solve latency is Austin, TX. Autonomy Institute, a cooperative research consortium, is working with the city of Austin on their Public Infrastructure Network Node (PINN) Pilot at the Texas Military Department. The Autonomy Institute is a corporate research and operational consortium focused on advancing intelligent infrastructure, and AI at the edge. The hope is the program will be expanded in the future to other major US cities and eventually around the globe.
Another challenge for data centers is power. Currently, data centers around the world consume approximately 1% of all the power generated in the world. By 2030, it is anticipated that this will increase to 8%. With the continued push for electric automobiles and the ongoing demand to reduce fossil fuel use, as well as the stresses on our already aging electrical infrastructure, this may not even be possible. To achieve this, it will require us to consider more onsite energy production in our future designs. This includes solar farms, wind farms, and potentially even small nuclear reactors.
I am truly excited about the upcoming challenges of designing data centers in order to address the latency and power needs – while at the same time making these facilities not only net energy neutral, but net positive. This will be crucial as the demands for power increase for the potential of data centers while we also search for ways to reduce the impact on the environment. The next 20 years will be very exciting, and I look forward to leading the way in the design of more sustainable critical facilities.