Wideband Multimode fiber (WBMMF) was introduced as a new fiber medium in ANSI/TIA-492AAAE, in June 2016. The ISO/IEC 11801, 3rd edition standard is now using OM5 as the designation for WBMMF. OM5 fiber specifies a wider range of wavelengths between 850 nm and 953 nm. It was created to support Shortwave Wavelength Division Multiplexing (SWDM), which is one of the many new technologies being developed for transmitting 40 Gb/s, 100 Gb/s, and beyond.
OM5 is being presented as a potential new option for data centers that require greater link distances and higher speeds. However, many enterprise IT and data center managers are increasingly adopting Singlemode fiber systems to solve these challenges.
So, what are the reasons a data center might consider installing OM5?
“OM5 offers a longer cabling reach than OM4.”
The difference is minimal.
For the majority of current and future Multimode IEEE applications including 40GBASE-SR4, 100GBASE-SR4, 200GBASE-SR4, 400GBASE-SR8 and future 400GBASE-SR4, the maximum allowable reach is the same for OM5 as OM4 cabling. There are only 3 current Ethernet applications that state an additional 50 meter reach with OM5. If a data center is using non-IEEE-compliant 100G-SWDM4 or BiDi transceivers, they would see a 150-meter reach with OM5 – only 50 meters more than OM4. For most cloud data centers, if they have cabling runs over 100 meters, they will likely use Singlemode for 100 Gb/s and greater speeds. Additionally, any installed OM5 cabling beyond 100m may be limited in its ability to support of future non-SWDM applications.
“OM5 will reduce costs.”
It won’t.
OM5 cabling costs about 30-40% more than OM4. In addition, if you look at the cost of a full 100 Gb/s channel, including BiDdi transceivers, the amount per channel is still 40% more than a 100GSR4/OM4 channel. The costs of Singlemode transceivers have declined considerably over the past 12-18 months due to silicon photonics technologies and large hyperscale data centers buying in large volumes. When comparing the price of 100 Gb/s transceivers, 100G-PSM4 using Singlemode fiber is the same price as 100GBASE-SR4 using Multimode fiber.
“OM5 is required for higher speeds.”
Not true.
All of the current and future IEEE standards in development for 100/200/400/800 Gb/s will work with either Singlemode (OS2) or Multimode (OM4). The majority of these next-generation speeds will require Singlemode. IEEE always strives to develop future standards that work with the primary installed base of cabling infrastructure so customers can easily migrate to new speeds. The latest draft of IEEE P802.3db standard includes 400GBASE-SR4 (a lower cost, less complex, more attractive alternative to 400GBASE-SR4.2) which will have the same reach for OM4 & OM5.
“OM5 will create higher density from switch ports.”
It won’t.
It has very been common for data centers using 40GBASE-SR4 and 100GBASE-SR4 to increase port density by breaking out 40 or 100 Gb/s ports into 10 or 25 Gb/s channels. If a data center manager decides to use SWDM4 or BiDdi modules with OM5 cabling, they cannot break out into 10 or 25 Gb/s channels. This is a major disadvantage of using this technology.
“Do the leading switch manufacturers recommend using OM5 cabling with their equipment?”
No, they show OM3 & OM4
Example from Cisco: “In 40-Gbps mode, the Cisco QSFP 40/100-Gbps BiDi transceiver supports link lengths of 100 and 150 meters on laser-optimized OM3 and OM4 Multimode fibers, respectively. In 100-Gbps mode, it supports 70 and 100 meters on OM3 and OM4, respectively.” Example from Arista: “100GBASE-SWDM4: Up to 70m over duplex OM3 Multi-mode fiber or 100m over duplex OM4 Multi-mode fiber”
Siemon does not see any good reason to currently recommend OM5 to large data center operators. For enterprise data centers looking at migrating to 40GBASE-SR4 or 100GBASE-SR4, OM5 offers no additional benefit over OM4. And larger cloud data centers are either already using Singlemode or planning to move to Singlemode in the near future for migration to 800 Gb/s and 1.6 Tb/s without changing out their cabling.
Gary Bernstein
Global Data Center Cabling Solutions Specialist, Siemon
Gary Bernstein is Sr. Director of Global Data Center Sales at Siemon with more than 25 years of industry experience and extensive knowledge in data center infrastructure, telecommunications, and copper and fiber structured cabling systems. Gary has held positions in engineering, sales, product management, marketing and corporate management throughout his career. Gary has been a member of TIA TR42.7 and TR42.11 Copper & Fiber Committees and various IEEE802.3 task forces and study groups including 40/100G “ba”, 200/400G “bs” and 400/800G “df” and 800G/1.6T “dj”. Gary has spoken on Data Center Cabling at several industry events in North America, Europe, LATAM & APAC including 7x24, AFCOM, BICSI, Cisco Live, Datacenter Dynamics and has authored several articles in industry trade publications. Gary received a Bachelor of Sciences in Mechanical Engineering from Arizona State University, is an RCDD with BICSI and a Certified Data Center Designer (CDCD) with Datacenter Dynamics.
