Testing: An Essential for 400G Third-Party Optics
There are differences of opinion about when exactly it will happen, but the next big thing in networking will be 400G connectivity. By common consent, millions of 400 Gbps switch ports have been shipped in the past couple of years and they will start to be populated in earnest by 2020. The shift from 100G to 400G represents a welcome boost in bandwidth, in an era of bandwidth-hungry streaming video and hyperscale data centers. Yet 400G also presents challenge, notably in terms of its complex modulation scheme.
Testing is quickly becoming a major requirement for successful implementation of 400G transceivers – an issue that separates the sheep from the goats when customers choose among third-party transceivers. In light of all this, ProLabs has set up a strict testing regime for 400G modules to ensure trouble-free use of its 400G products and maintains a fully functional testing center at its facility.
Modulation Matters
Transceivers below 100G use simple modulation schemes. 100G brought more complex modulation schemes capable of transmitting multiple bits at once. PAM4, the scheme chosen for most 400G networking, is even more complex by combining two bits into a single symbol with four amplitude levels. PAM4 modules can provide a given data rate at half the frequency effectively doubling the network throughout and reducing power consumption.
For all of these advantages, PAM4 creates challenges in terms of the signal-to-noise ratio (SNR), effectively reducing transceiver reach. In 400G, there is a much greater need for forward error correction (FEC) to mitigate the loss of signal integrity. These issues make module testing essential and third-party suppliers that are not capable of providing this testing are at risk in providing low quality products to their customers and losing these customers in the process.
Testing 400G
While 400G testing could be left to third-party testing firms, ProLabs thinks that this area is too important “to leave to the experts.” Not only does maintaining an in-house testing facility demonstrate ProLabs’s commitment to quality, but it ensures that modules can be speedily shipped to customers without some third-party acting as a roadblock. Fast shipping to customers is believed to be one of the factors that pushes customers towards using a third-party transceiver source in the first place.
None the less, testing of 400G modules is complicated. Unlike the testing of earlier generations of transceivers, it focuses less on whether there are errors than on whether the levels of errors are acceptable. Some 400G transceivers operate error-free post-FEC and others will not. A more sophisticated understanding of the error distribution and statistics is required in 400G modules than in earlier transceivers to ensure that error patterns are acceptable and what the root causes are of the errors that do occur. In addition, FEC functionality is itself complex and needs to be tested for both logical errors as well as dynamic power performance.
No measurements possible on the signal after the channel without de-embedding!
• Effect of the channel must be de-embedded. This can either be done using a VNA or a DSO
No longer can testing be confined to just one of the layers, but it must cover the link from the physical layer through to Ethernet. Test results must be able to reconcile where the issues lie and fully validate the margin implications of the FEC channel.
400G Testing at ProLabs
The ProLabs testing lab includes a MultiLane BERT tester, which is essential to much of the testing regime outlined above. There is also a Digital Sampling Oscilloscope (DSO) which can help testing waveforms and whether – in a given module – the electronics is behaving in a way that guarantees efficient information transmission from the module across the network. The testing lab also provides ways of testing how well a 400G transceiver behaves in a network. This includes having Juniper, Cisco and Arista 400G switches in which the 400G transceivers can be tried out. And there is also an EXFO 400G traffic generator, which can be used to test 400G transceivers in simulated traffic conditions, in other words to see how well the 400G transceiver will perform on a network link.
The modulation changes from 100G to 400G are profound and necessitate a significant leap forward in terms of testing at the supplier level – and that includes third-party suppliers. If a customer buys 400G modules from a supplier that does not have its own testing lab, it risks getting poorly performing modules and perhaps – to add insult to injury – receiving them with extended lead times.
Contact our global sales team for more information on our testing capabilities.