Catching the Wave With 40G Technology

Catching the Wave With 40G Technology

Have you ever frustrated by your 10G network which is on the verge of capacity limitation? Or you just can’t help but worrying that your design may not provide enough speed for those bandwidth hunger application. Well, you are not alone. So it’s high time to get rolling and move 40G into your data center.

In 2016, just eight years after the adoption of 10 Gigabit Ethernet, IEEE published 802.3ba to pave the way for 40 Gigabit Ethernet. The original driver behind this is the massive application of high-speed, high-demand computing applications and technologies. Let’s see in which way 40G technology transform the data center infrastructure.

40G: Where Budget, Speed and Scalability Meet

The growth in 10GbE connectivity at the server edge will ultimately drive 40 GbE uptake. In order to reduce the increasing complexity associated with managing multiple 10 GbE uplinks form every server rack, enterprises will move to 40 GbE technology. Which pushes the bounds of higher speeds, massive scalability and greater flexibility. Consequently, 40G market is fast maturing - 10G and 40G adoption has reached a break-even point before 2008, and then the latter becomes more economical, with prices likely to drop below the 10G cost.

40G changes the way signals are transmitted, adopting parallel transmission instead of duplex transmission that used in 10G cabling. Aside from the data-rate increase, it provides higher density and lower cost alternative: with 4 times the capacity and the ability to cost effectively migrate to 100G, 40G Ethernet on multimode fiber is proven to be a logical step in the evolution of the network.

Migrating to 40G Ethernet: A Build-to-Suit Trend

When the 40G market started, it was primarily for connecting data centers together and for router links, connecting cities together. A variety of 40G optics are introduced to meet the requirement of  both short and long-haul data transmission.

An Overview of 40G QSFP Transceivers

40G QSFP (Quad Small Form-Factor Pluggable) transceivers offer service providers high density 40 Gigabit Ethernet connections in enterprise data centers. There exist a dazzling array of 40G QSFP transceivers, which is listed in the following chart with details including their interfaces, cabling media and maximum reach.

Mixing 10G and 40G Cabling? QSFP+ Universal Transceiver Came to Secure

Have you ever encountered the situation that, when integrating 10G infrastructure to 40G, the cost and deployment time tends to increase because of the different cabling structure. Then how to streamline the installation and decrease the cost while ensure the smooth transmission? A universal transceiver was introduced by Arista to rescue – QSFP-40G-UNIV or QSFP+ 40G universal.

With a duplex LC port for dual-way transmission, this QSFP+ universal transceiver can be used in a variety of applications. It uses four different wavelengths and can support optical transmission over both single-mode and multimode fiber. It is known that 10G network utilizes two fibers for dual transmission, while 40G applications uses 12-fiber MTP based fiber. So you have to change the cabling structure by laying more fibers for 10G to 40G upgrade. The 40G QSFP+ UNIV module offers two strands of fiber for dual-way transmission that resembles most 10G network, so the cabling structure of the legacy 10G can be saved when migrating to 40G network, the cabling flexibility is further improved.

40G Breakout Cabling for Smooth Integration

As the 40G technology continues to advance, it is also used to integrate with 10G applications by muxing together four 10G connections. Which contributes drastically to cut costs and make more efficient use of available spectrum. Data centers have developed 4×10G Ethernet breakout cables that terminated with one 12-fiber MTP/MPO connector at one end and four duplex LC connectors at the other end. The four LC cables each delivering 10G, for a total of 40G. The same breakout strategy is likely to be used to eventually support 100GbE and 400GbE server interconnects, offering data center an appealing and cost-effective migration strategy into the higher speeds.

How to Test the Compatibility of Your 40G Transceivers?

Third-party optical transceivers have taken a certain fraction of the current 40G optics market. To ensure the compatibility of the transceiver module and device, testing is necessary before putting into use. Here we take FS.com Cisco QSFP-40G-SR4 compatible 40GBASE-SR4 QSFP+ transceiver for example, to test if it is fully compatible with Cisco Nexus 9396PX switch. Referring to the following procedure to perform transceiver testing.

1. Plug the Cisco QSFP-40G-SR4 compatible QSFP+ into the 40Gbps ports of the Nexus 9396PX.

2. Plug the cleaned MTP patch cord into the optical ports of the transceivers.

3. View the state of LED light and more information in the CLI.

Note:

1. Always clean the optical transceiver interface before connection.

2. Wear an ESD wrist strap or anti-static gloves when installing or removing a transceiver.

Minimize the Pain, and Maximize the Financial Gain

The boosting market of 40G has triggered some debates among IT managers. They hesitate whether to deploy 40G technology or wait for the pending 100G. This confusion, however, is fast becoming void since 40G provides design flexibility and cost advantage over 100 Gigabit Ethernet. 40G can be effectively deployed in aggregation links, and it can be commonly applied to access links to connect servers. At least for now, they can co-exist in enterprise data centers, and 100G cannot completely replace 40G with the years to come. Complimenting 40G, 100G is a perfect choice for carrier service providers and core links in data centers. Analysts expect 40G deployments to start ramping up with broader adoption in 2018.