Pushing the Edge of 400GE
Everyone wants to go faster – but raw speed on its own isn’t enough. That extra performance has to be delivered reliably and consistently whenever it’s needed. This is the key challenge facing cloud service providers and telcos, as they upgrade to keep up with demand for bandwidth from enterprises and consumers.
It’s estimated that the bandwidth requirements from Global IP networks—whether local area, data center, access, or core – will triple over the next five years, and so networks and network devices must evolve beyond the newest 100GE speeds and capabilities to keep up. To meet these growing bandwidth requirements, network equipment manufacturers (NEMs) and their customers are racing to bring 400GE products to market.
But how do you go about testing and validating the performance of these ultra-high speed networking solutions, to ensure they can deliver on their promises? You need test solutions that can push these new solutions to their limits, and beyond.
That’s why we launched the world’s first 400GE testing solutions earlier in 2017, to help NEMs ensure their products meet the performance and quality standards expected of them, and to shorten the development and test times, so that innovative networking technologies could be brought to market faster.
To show what goes into developing advanced test solutions for the next generation of networking technologies, Ixia’s senior product manager, Charles Seifert, recently met with editor-in-chief of Light Reading, Craig Matsumoto, for a chat about what’s under the hood of our testing modules.
The challenges of developing 400GE test equipment
The first key point to note is that 400GE technology is not simply a case of running 400 Gigabits per second on a single copper wire: rather, it’s an advanced approach which combines eight serial lines of 56 Gbps. Charles stated that developing this innovation really pushed the boundaries of developing componentry and circuit boards, in order to carry what’s essentially microwave traffic.
He also pointed out that forward error correction needed to be adopted, in order to control errors in data transmission. The higher the speed, the greater the potential for data errors being introduced. Forward error correction was very much a luxury extra with 100GE - but with 400GE, it's considered mandatory. This means extra processing work during the data packet forming process and in the encoding and decoding of the blocks used to transmit data. The FEC adds a lot of design complexity, but it dramatically increases link quality at high speeds.
Performance AND reliability
Speed is no good unless it can be attained reliably, and one of the main benefits of 400GE is the fact that it will reduce the numbers of links, ports and interfaces that IT teams have to work with on their networks. That's a huge benefit.
As Seifert put it, in the past, the ability to replace 10 separate 10GB links with a single 100GB link meant getting rid of nine links that don’t need servicing any more – in other words, nine points of potential failure have been eliminated from the network. 400GE will take this to the next stage, further reducing points of failure and boosting reliability. This in turn confers additional benefits for users, including lower networking costs per gigabit, and less power consumption.
Summing it up, 400GE will dramatically increase both speed, efficiency, and reliability for the next generation of communications. It will provide the bandwidth required by cloud-scale data centers, Internet exchanges, co-location services, wireless infrastructure, service provider networks, and video distribution infrastructure. Getting out in front of 400GE from a test perspective required a lot of innovation from Ixia's engineering team, but we got there first and now our NEMS customers can push the next generation.
Watch the full video of Charles and Craig’s conversation on Light Reading here.