Berk-Tek, a Nexans company, recently announced that its OM3 and OM4 preterminated fiber-optic assemblies successfully underwent rigorous testing at the Nexans Data Communications Competence Center (DCCC), proving that these cable assemblies far exceed the performance required by the IEEE's 40- and 100-Gbit Ethernet specifications. Specifically, the GIGAlite-10 OM3, GIGAlite10-FB OM4 and GIGAlite10-XB "OM4+" cables were scrutinized at the DCCC under a number of different test configurations and platforms.
During 40-Gbit/sec testing using Avago 40G QSFP+ transceivers, the OM3 fiber assembly successfully supported 202-meter transmission, which Berk-Tek points out is double the reach specified in the 40GBase-SR4 standard. In a 10-connector channel, Berk-Tek's OM4 assemblies successfully transmitted longer than 800 meters, which is more than five times the standard-specified distance. During separate 100-Gbit/sec testing, the OM3 fiber met a 10-12 frame error rate over 252 meters with three connections; the OM4 fiber reached 352 meters with four connections.
Eric Lawrence is vice president of emerging technologies and technical director for Nexans; he directs the operations of the DCCC. Lawrence explains that the testing conducted in the DCCC is far different from the testing sometimes undertaken in an attempt to document a system's capabilities. Tests at the DCCC replicate real-life use conditions as opposed to the near-perfect-condition setups of some test environments. One example of this is the fact that the DCCC conducts frame-error-rate testing, rather than bit-error-rate testing, which is most often cited when test results are announced.
Lawrence elaborates, "We continually push performance boundaries. The fact that we perform frame error rate testing, which is far more stringent than bit error rate testing, means that our tests run hundreds of times longer than the industry requires ... Using multiple connection points is also important because in real-world applications, our customers need to know that our cable can perform reliably and dependably in a variety of configurations." During frame error rate testing a single bit error will cause the entire frame to be discarded, Nexans explains. This approach more accurately reveals the impact of errors on the applications than bit error rate testing does, and is therefore a more meaningful metric, Nexans asserts. This testing arrangement means that systems successfully exceeded the desired 10-12 frame error rate for 40G over five hours rather than the 32-second minimum duration of bit error rate testing. Likewise, the 100G channels were tested successfully for more than two hours instead of the 13-second minimum bit error rate testing duration.
Nexans says other DCCC test characteristics that make the environment more like real-world use sites include the use of multiple test platforms and transceivers. Nexans notes its strong relationship with many component manufacturers has solidified the center's position in the field.
Of the tests on Berk-Tek's OM3 and OM4 products, Berk-Tek's preterminated optical fiber product manager Alfred Flores said, "Obviously we are very pleased with the results. Successfully going 800 meters across 10 connection pairs using OM4 fiber clearly demonstrates what can be achieved when you combine our outstanding-quality optical cable and connectors in support of next-generation communications technologies."
Berk-Tek fiber-optic products business manager Mike Connaughton said, "These types of studies are part of the value that Berk-Tek brings to the customer beyond simply the components themselves. The fact that the Competence Center successfully transmitted both 40G and 100G Ethernet to distances beyond those defined by the standard with our OM3 preterm products was very positive. But more important than that, the stellar results of our OM4 grade fiber, when combined with Avago transceivers, create a new benchmark in the industry."