In the aftermath of COVID-19, hyperscalers have become mainstream. Unlike the past, when data centers were largely situated in locations such as Sweden, Iceland, in the post-COVID era that logic has almost turned on its head!
Take a look at the numbers. The acceleration of the digital economy is reflected on the graphs of Internet Exchange Points (IXPs) worldwide. Exchange Points, such as AMS-IX in the Netherlands, have significantly increased their Internet traffic between March 2020 and March 2021. AMS-IX’s traffic, for instance, has increased from about 5 Tbps to about 7 Tbps, according to research by Eurofound.
Naturally, demands of a data center has exponentially increased. In this scenario, Hyperscale and global-scale clients looking for the right building blocks to enable building of a flatter network—one that delivers much better performance and redundancy.
Unlike the past, increase in data traffic puts a lot of pressure on the network and this warrants higher capacity as well asn energy efficient switches. Back in 2014, when the 25G Ethernet Consortium proposed single-lane 25 Gbps Ethernet and dual-lane 50 Gbps Ethernet, it created a big fork in the industry’s roadmap, offering a lower cost per bit and an easy transition to 50G, 100G and beyond.
In 2020, 100G hit the market en masse, driving higher and higher fiber counts—and larger hyperscale and cloud-based data centers confronted their inevitable leap to 400G. With switches and servers on schedule to require 400G and 800G connections, the physical layer must also contribute higher performance to continuously optimize network capacity. In a changed network topology, how can hyperscalers ensure that is the question.
Aligning the Network
Data center managers are looking to take advantage of faster, higher capacity switches.
To do this, data centers need to provision more ports at higher data rates and higher optical lane counts per port. Among other things, this requires thoughtful scaling with more flexible deployment options, points out Sivashish Jena, APAC Data Center Business Lead CommScope.
Simplifying and supporting the required design, installation, operations and migration paths means infrastructure and network teams must collaborate to ensure the cabling architectures align with the network configurations. What we’ve found with our hyperscale and global-scale clients is that they need the right building blocks to enable a flatter network—one that delivers much better performance and redundancy. The evolution from four-lane quad designs to eight-lane octal has enabled the migration to 400G, 800G and eventually 1.6T and beyond. The 16-fiber configuration that supports octal technology is the primary building block Sivashish Jena, APAC Data Center Business Lead CommScope.
This, in turn, is driving changes in network topologies and raises the question – how can all that capacity be distributed most efficiently, switch to switch and switch to server?
We know that the trend is to flatten the network by reducing switch layers where possible. But each use case is different; therefore, a high degree of flexibility is necessary Sivashish Jena, APAC Data Center Business Lead CommScope.
That means providing the widest range of options and configurations that support legacy applications. These are just some of the changes that are helping to reshape the requirements and designs of today’s hyperscale and cloud-scale data center networks.
Propel Forward
It is precisely here that a solution like Propel fits in. Propel is an end-to-end, high-speed, modular fiber platform that addresses the current and future needs of today’s larger, high-fiber count data center networks. “Propel actually grew out of our High Speed Migration platform. Shortly after launching that a few years ago, we began working on what would become Propel. We went to our customers, sales team, and integrators and installer partners and asked them: If you could have anything in a fiber platform, what would it be? If you’re looking at the applications in the market, how would you want to support them?” Their answers echoed the same themes repeatedly—16-fiber migration, ultra-low loss optical performance, front and rear panel access, interchangeable components,” explains Sivashish Jena, APAC Data Center Business Lead CommScope.
The ability to evolve the physical layer infrastructure in the data center is ultimately key to keeping pace with demand for the low-latency, high-bandwidth, and reliable connectivity that subscribers demand.
Needless to say, the industry continues to grow. At CommScope, we’re always looking at what’s next and what’s at the forefront of the ever-evolving data center landscape.
Contact us if you’d like to discuss your options when preparing for migration to higher speeds.