Some Basics About Migration to 40G Ethernet

Fiber optic links are vital for providing the bandwidth and speed needed to transmit huge amounts of data to and from a large number of sources. Recently more bandwidth is greatly needed to support the use of virtualization and improved space utilization in the data center. 2010 witnessed the ratification of 40 and 100 gigabit Ethernet (GbE) standard, since then leading switch manufacturers offered 40 GbE blades and more than 25% of data centers have implemented these next generation speeds. It is anticipated that by the end of this year, more and more data centers will follow suit. Therefore today’s article will provide an effective solution to help you migrate from current data center applications to 40/100 GbE.

10G OM3/OM4 Connectivity in the Data Center

The IEEE 802.3ae 10G standard released in 2002 included 10GBASE-SR OM3 guidance that can operate at 850nm with duplex fiber serial transmission. Even though duplex fiber implied duplex LC connectivity throughout the channel, 12F MPO terminated cables emerged as a primary choice for deployment in data-center backbone applications. The 12F MPO terminated trunk cable provided the highest fiber packing density to maximize pathway and space utilization in ducts, raceways, and patch panels. For example, 46C3447 is IBM BNT 10GBASE-SR SFP+ that operates over OM3 cable for a distance of 300m with LC duplex connector. An MPO connectorized backbone cable typically is terminated in patch panels using one of two methods that break-out the 12F MPO to six 2F duplex LC (Figure 1).

Standards From 10GbE to 40/100GbE

Similar to how transportation highways are scaled to support increased traffic with multiple lanes at the same speed, the 40 and 100 GbE standards use parallel optics, or multiple lanes of fiber transmitting at the same speed. Running 40 GbE requires 8 fibers, with 4 fibers each transmitting at 10 Gbps and 4 fibers each receiving at 10 Gbps. Running 100 GbE requires a total of 20 fibers, with 10 transmitting at 10 Gbps and 10 receiving at 10 Gbps. Both scenarios call for using high-density multi-fiber MPO style connectors.

According to IEEE 802.3ba standard, multimode optical fiber supports both 40 and 100 GbE over link lengths up to 150 meters when using OM4 optical fiber and up to 100 meters when using OM3 optical fiber. It is important to note that single-mode fiber can also be used for running 40 and 100 GbE to much greater distances using wavelength division multiplexing (WDM) for most data center applications of less than 150 meters. Copper twinax cable is also capable of supporting 40 and 100 GbE but only to distances of 7 meters. Take EX-SFP-10GE-DAC-1M as an example, it is Juniper SFP+ passive copper cable, which is ideal for 10G interconnect application for a link length of 1m.

Migration From Duplex Fiber Transmission to 40/100G Parallel Optics

Based on a MPO system, migration from 10G to 40G to 100G seems to be a simple and easy deployment. Starting with a 10G configuration, a base 12F MPO backbone cable is deployed between the 10G switches. As discussed earlier, modules or harnesses are used at the end to transition from the 12F MPO to LC duplex. These breakout configurations enable connectivity into the switch.

And when the switches migrate to 40G, the 10G module or harness is removed and should be replaced by a conversion module or conversion harness as shown in Figure 3 and Figure 4. Alternatively, an MPO adapter panel can be used. In any of these deployment options, the use of an MPO terminated jumper is needed to establish connectivity between the switches. For 100GBASE-SR4 networks all Figures 4-6 cabling is applicable.

Conclusion

To keep up with the path of future, MPO-based connectivity using OM3 and OM4 is the ideal solution for the data center, which makes the transition to 40/100G more easier and efficient. Fiberstore offers various types of 40GbE transceivers, MPO/MTP trunk cable, MPO/MTP harness cable, MPO/MTP cassette and other assemblies for your 40G network connectivity. For more information about our products, please contact us directly.

The Evolution Path of BASE-T

With the requirements laid on data center increasing rapidly, the ability to flexibly adapt to future demands is tremendously crucial for data center managers. Often this can be achieved by deploying higher bandwidth solutions in a part of the data center, provided that these systems are backwards compatible with existing infrastructure or it may be a cost-consuming method. BASE-T technology featured by its low cost, availability and flexibility is largely favored by data center designers. This article illustrates the migration of BASE-T technology so that people can future proof their data center tomorrow.

Why BASE-T Is so Popular?
To be short, three main advantages will be concluded in the following part.
1. Least cost access layer alternative when compared to other interconnect technologies

• Optical (e.g. SR, LR)
• Direct-Attached

2.Structured topology

• Common physical interface (RJ45)
• Flexibility and longevity
• Optimized for small to medium-sized data centers (< 20K square feet)

3. Supports auto-negotiation and Power-Over-Ethernet

• Simple plug and play installation
• Ubiquitous RJ45 interface simplifies 10GBASE-T to 40GBASE-T upgrade path

1000BASE-T—Gigabit Ethernet Over 4-pair Cat 5 Cabling
1000BASE-T (ratified in 1999) is a Gigabit Ethernet standard over copper wiring at the speed of 1000 Mbps. Each 1000BASE-T network segment can support a maximum length of 100 meters, and uses Category 5 cable or better (including Cat 5e and Cat 6). 1000BASE-T also uses a symbol rate of 125 Mbaud and all four pairs for the link and a more sophisticated five-level coding scheme. The 1000BASE-T SFP operates on standard Category 5 unshielded twisted-pair copper cabling of link lengths up to 100 m.

Realizing 10BASE-T
Upgraded from 1000BASE-T, 10GBASE-T (certificated in 2006) offers the most flexibility, the lowest cost media, and is backward-compatible with existing 1 GbE networks. 10GBASE-T connected with Cat 6 and Cat 6A (or above) cabling supports a length up to 100 meters that gives IT managers a far greater level of flexibility in connecting devices in the data center. 10GBASE-T and Category 6A cabling costs less than using either optical fiber or SFP+ direct attach copper (DAC) options that have been widely deployed to date center for 10 Gb/s. For example, EX-SFP-10GE-DAC-1M can only support a link length of 1m that largely limits its application. Figure 1 presents a comparison between 1000GBASE-T and 10GBASE-T.

Road to 40GBASE-T in Data Center Networks
If there is a 10GBASE-T for switch-to-server and switch-to-switch connectivity, there will be a 40GBASE-T over twisted pair cabling for the 40G data center deployment according to the IEEE. Twisted pair cabling with the RJ-45 connector has always been the first choice for IT professionals, based on its low cost and ease of use. Unlike fiber or twinax solutions, twisted pair cabling can automatically switch to different data rates, such as from 100MbE to 10GbE. Therefore migration to 40GBASE-T does not require a upgrade of all the equipment of the data center, which will reduce of the overall expenditure of the data center.

The advantages of 40GBASE-T are clear, but the path from initial ratification to commercial availability is not always smooth. There still a few months off for 40GBASE-T standardization, here comes some good news, as well as some considerations.

40GBASE-T is specified with transmission performance up to 2 GHz (four times the bandwidth of Category 6A) with a lot more stringent alien crosstalk requirements. Since initial 40GBASE-T applications would be limited to data centers, the traditional twisted pair Ethernet 100m link length is not essential. Additionally, Industry players helping in the development of an industry standard for 40GBASE-T have to ensure that it could be supported and rolled out cost-effectively. The new standard will minimize the time it will take to develop new electronics for switches and servers that can support 40GBASE-T connectivity, by building on the work already completed to support 10GbE connections. The standard will also support the ubiquitous RJ-45 connector. Sooner or later, 40GBASE-T will be upon us. Nowadays 40GBASE-LR4, 40GBASE-SR and 40G QSFP+ cables are there to help with the deployment of 40G connectivity. Take JG330A as an example, it is QSFP+ to 4SFP+ Passive Copper Cable available for short reach application. Figure 2 shows a data center twisted-pair migration roadmap.

Summary
BASE-T technology (1000BASE-T, 10GBASE-T or 40GBASE-T) always retains the traditional advantages—low cost, easy to deploy and auto-negotiation for plug and play and backwards compatibility. 1000BASE-T and 10GBASE-T have already brought benefits to people. But no one can foresee that 40GBASE-T will be used in the future but future-proof planning of the cabling is important, given the long life of the cabling systems. Fiberstore provides a full range of BASE-T products including 1000BASE-T SFP, 1000BASE-T media converter, 1000BASE-T GBIC transceiver, etc. And 40GBASE-T devices will be coming soon. If you have any request of our products, please send your inquiry to us.