Marcin Bała, CEO at Salumanus, explains the requirements for 800G connectivity and how it will impact the telecommunications industry.
The need for higher bandwidth
According to TeleGeography’s 2023 State of the Network report, global internet bandwidth rose by 28% in 2022 and total international bandwidth stands at 997 Tbps in 2023, representing a four-year CAGR of 29&%. The surge in data consumption, driven by content providers, users and telecom operators, necessitates higher bandwidth. Content providers compete for more users, offering superior quality content, and users demand access to this content with minimal delays.
A recent consultation launched by telco operators at the European Commission on the future of the connectivity sector introduced the possibility of a ‘fair share mechanism’ or ‘telco tax’ in which content providers would contribute to the cost of running an upgrading networks,
According to the same TeleGeography report, content and cloud service providers like Google, Meta, Amazon and Microsoft have become the primary sources of demand. These companies are the dominant users of international bandwidth, accounting for 69% of all used international capacity in 2021.
As a result, telecom operators must continuously upgrade their infrastructure to deliver high-quality content efficiently. Thanks to Industry 4.0 technologies like IoT, VR and AR, the demand for seamless, low-latency connectivity is even greater, putting pressure on operators to reduce the number of active devices and push data centres closer to users.
To achieve the bandwidths necessary to accommodate the surge in demand, optical networks require wider channel widths, measured in Gigahertz (GHz), within the optical spectrum. Channel width determines how much data can be transmitted over a particular wavelength.
In traditional fixed grid networks, channel widths are pre-defined and fixed, typically at specific values like 50 GHz or 100 GHz. This fixed allocation may lead to inefficient use of the available optical spectrum, especially when the required channel width does not match the fixed grid spacing. As data demands increase, fixed grid networks may face limitations in spectral efficiency and data-carrying capacity.
For example, operators using a 50 GHz fixed grid wouldn’t be able to take advantage of high-bandwidth transmissions like 400G or 800G because the grid size would not allow for it.
On the other hand, a flexible grid network is a solution in optical networking that allows for dynamic allocation of channel widths. It enables network operators to adjust channel widths based on specific data transmission requirements, optimising the use of the available optical spectrum and increasing data transmission capacity.
The benefits of 800G
The rollout of 800G technology is already underway, with hyperscale companies like Google, Facebook, Microsoft and Amazon at the forefront of its deployment. This early-stage deployment involves establishing short connections within data centres, then gradually extending to interconnect other data centres.
With this and further testing, it is expected that telecom operators will begin adopting 800G for long-distance transmission. Virgin Media O2, for example, has migrated all core traffic in its backbone locations across the UK to Juniper Network’s transport routers capable of supporting 800G capability.
As the telecommunications industry adopts 800G, users can expect improved streaming quality, faster access to data, and the capability to handle large amounts of data efficiently. Additionally, 800G will be a vital enabler for the adoption of future technologies such as generative AI and quantum computing.
Network requirements
Upgrading bandwidth from 100G to 400G and eventually to 800G brings several challenges. While the transition promises better services, increased energy efficiency and reduced costs, it also demands a shift in network architecture.
Previous generations of coherent technology relied on CFP and CFP2 interfaces, which were unable to work directly with network equipment like switches or routers that had QSFP28 interfaces for 100G ports and QSFP-DD for 400G ports.
For example, Salumanus’ QSFP-DD 400G ZR and 400G ZR+ coherent transceivers can be installed directly in the equipment, avoiding the need to invest in additional muxponders and transponders and cutting the initial installation price by more than 50 per cent. On top of this, both modules are CMIS compliant, which means that parameters like modulation, channel number and FEC type could all be set directly with a switch or a router command line.
