Previously, I wrote about the coming demand for data servers and the electricity required to power them. The picture can look dire with conservative assumptions that may underplay continued progress in processing efficiency. So, we can change the assumptions in my model to show greater new efficiencies in data center processing and operation, significantly easing the stress data centers will put on worldwide grids.

But we can't expect to handle the expected data growth without serious digging into our own, local issues in deploying data centers and the power that will feed them. Projections show all of the world's data reaching 1 yottabyte (1,000 zettabytes) in 2034, with Internet traffic exceeding 100 zettabytes in 2037 and reaching a full yottabyte sometime in the 2040s.

The Zettabytes Are Piling Up
Internet traffic is projected to reach 4.8 zettabytes this year, in 2023 – an average of 152 terabytes per second over the course of the year. Using the EU and North America as benchmarks, we estimate developing Africa to have a share of this traffic close to 50% proportionate to its population, with developing nations in other regions reaching about an 80% proportionate share.

This means Africa, with about 1.2 billion people, or 15% of the world's population, and Southeast Asia, with about 700 million, will each generate about 7% of the traffic. This represents a bandwidth demand of more than 10 terabytes per second for each of the regions.

It also represents an electricity demand of 5 to 6 gigawatts for the data centers alone; add a significant percentage of this to power the devices. Using figures from the US EPA (Environmental Protection Agency), we can calculate this demand will emit around 50 million metric tons of CO2.

Current Conditions, Future Needs
Even though nations in the developing world typically get by with a small fraction (between 1% and 6%) of per-capita electricity consumption when compared to developed nations, our numbers show that overall electrical grids in Southeast Asia and Africa can hypothetically handle this demand.

But the existing digital infrastructure cannot. Mobile traffic is proportionally higher in both regions, particularly throughout Africa, then in the world, which mitigates this situation to some degree.

Developing a nation-by-nation look of expected IP demand and the electrical and digital infrastructure needed to support it presents a series of significant tasks for analysts, government planners, investors, and operators. Our growing research effort at IDCA is equipped to work with interested parties in taking on these tasks.

The metaverse is kicking IP traffic demand into a much higher gear, and all nations in the world need to be ready for this.