MPO Based Fiber Architecture And Next-gen Multimode Fiber

I thought I’d share something that is very near and dear to my heart and has tremendous ramifications for data center infrastructure design. It’s the wonderful combination of pre-terminated MPO-based fiber architecture and the next generation of multimode fiber (fibre), called OM4.

Greetings and welcome to my first attempt at blogging—I’m sure my kids are very proud.

My name is Simon Cowley, and I am a vice president for CommScope Enterprise Solutions’ Global Technical Organization (GTO). I love that acronym as it reminds me of the old Pontiac GTO I had when I was 17.
Purble Pontiac GTO

I’m sure mine never looked this nice and it wasn’t the same color (is rust a color?). Ah, but I digress.
To start our conversation, I thought I’d share something that is very near and dear to my heart and has tremendous ramifications for data center infrastructure design. It’s the wonderful combination of pre-terminated MPO-based fiber architecture (or “fibre architecture” for my European friends, and I have several), and the next generation of multimode fiber (fibre), called OM4.

Let’s navigate through the MPO-based fiber architecture first.

For high-density fiber environments, such as data center SAN and backbone implementations, the combination of the 12-fiber MPO connector and factory terminated assembly is a great solution—as long as the cable diameter is small and the architecture doesn’t limit network design flexibility, application distance, reliability, or number of connections.

With 12 fibers in a connector roughly the size of a duplex LC, the MPO enables you to connect many fibers in a snap. However, because it is 12 fibers in one rectangular housing, it is not readily terminated in field; in fact, it is virtually impossible. But that’s OK, because data center managers don’t want that being done in their spaces anyway, with all the associated disruption and potential contamination. Pre-terminated assemblies provide significant time savings, too, by getting those “revenue-generating applications” up and actually generating revenue much faster.

Today, the typical fiber application runs on duplex fiber, one TX and one RX, so the multi-fiber backbones must be broken out into duplex connections on either end. This is easily accomplished with pre-terminated modules and fan-out assemblies. Once installed, the solution appears and behaves just like a well-planned duplex system. Even better though, with the right MPO architecture, you NEVER have to worry about those dreaded polarity issues – TX will ALWAYS go to RX, so the days of trouble-shooting by flipping patch cord polarity may be gone FOREVER.

But the really big benefit is just around the corner. The new 40Gb and 100Gb Ethernet standards are based on parallel optics using the 12-fiber MPO connector as the equipment interface. Brilliant! Wouldn’t it be convenient and cost-effective if your MPO architecture also migrated easily to these new high-speed applications?? Well it can, provided you’ve chosen the right keying approach (TIA Method B). When the equipment starts showing up with MPO interfaces, simply remove the module or fan-out, use those somewhere else at your site, and connect to the equipment with a 12-fiber MPO patch cord. You can even do a full cross-connect if you like—true structured cabling around 12-fiber increments.

Now that’s driving home performance! The next time you’ll see me in this space, we’ll switch gears (again I picture my old GTO) and discuss how OM4 gets you further on a tank of gas.

So there you have it – I’m blogging! I’m a blogger! I blog! I am Simon the Blogger who blogs. A quarter to whoever is first to correctly identify those two obscure movie references. I look forward to hearing your thoughts on this or any other topic.