As cloud computing, big data and the deployment of mega-scale data centers accelerates, organizations need to continually recalibrate and evolve the network. This challenge has led to the development of new technologies and standards designed to increase and optimize network capacity, security and flexibility, all while keeping a lid on cost. Here are the top five trends as we see them:
* Rapid Adoption of 802.11ac. Tablets and smartphones are becoming ubiquitous in the workplace. As the number of mobile devices and the deployment of cloud-based enterprise services continues to scale at a dramatic rate, the keepers of the network must reconsider how they provision, secure and control enterprise computing resources and information access.
IEEE 802.11ac is 3x faster and 6x more power efficient than its predecessor, 802.11n, while remaining interoperable with 802.11n as well as legacy 802.11b/g/a air interfaces. According to recent data from Infonetics, sales of 802.11ac access points are up almost 10-fold over the past year. This has only increased the pressure on enterprise technical officers to better manage and secure their network connection points.
* 2.5G and 5Gbps in the Wiring Closet. As Wave 2 of 802.11ac Wi-Fi rolls out, traffic rates through access points will surpass multiple gigabits per second, requiring both the access point and the Ethernet switch ports to scale beyond 1000 Base-T (1G). However, with literally millions of feet of installed Category 5e/Category 6 cabling, enterprise operators are looking to fill the gap between 1G and 10G over this legacy unshielded twisted-pair copper cabling. (Also read: Shopping for 802.11ac Wave 2 Wi-Fi? Don't rush into it.)
The Wave 2 addendum to the original 802.11ac wireless specification will usher in multi-user MIMO antenna technology and other advancements to help increase theoretically maximum wireless speeds to 6.93Gbps. This has created opportunity for faster Ethernet speeds and a fast upgrade cycle to 2.5G and 5G, versus the 10x speed upgrade path typical of Ethernet. While new installs will be wisely using Cat 6A cable to enable 10G, most of the legacy installed base will stay with 1Gbps or adopt 2.5G/5Gbps when they need an uplink rate of greater than 1Gbps.
While multiple industry proposals exist to enable 2.5G/5Gbps over existing Cat56/Cat6 cables, the industry needs to get behind a single standard to avoid confusion and non-interoperability among different solutions. This could potentially adversely affect the potential success of this emerging market. The good news is the IEEE 802.3 standards body is focused on bringing multi-rate Gigabit Ethernet BASE-T to enterprise access points via a 2.5G and 5G proposal.
The emphasis is on achieving this near term and over current Cat 5e (2.5G over 100m) and Cat 6 (5G over 100m) installed cable. Given the limited bandwidth, the obvious route to achieving this is to use higher modulation schemes.
The addition of 2.5G and 5G Ethernet link protocol speeds in the wiring closet will enable the cost-effective scaling of network bandwidth to enterprise access points and provide IT professionals with more data-rate options.
* 25G, 50G and 100G in the Data Center. Enterprise and cloud network operators must build out their networks following carefully planned technology roadmaps that scale according to changing needs. As always, these operators will have to balance performance optimization while maintaining the lowest possible capital and operating expenditures (CAPEX/OPEX). One way to accomplish these often-conflicting goals is by transitioning to higher-speed Ethernet technologies.
Experts predict that the largest cloud operators will shift to 100G Ethernet fabrics while cost-efficient 25G and 50G will remain the workhorses for most of the other well-known data-center companies. The 25G/50G Ethernet Consortium – a relatively new body – has made its 25G and 50G Ethernet specification open to all data center ecosystem vendors - royalty free - to create and deploy compliant, interoperable product implementations. The standard should hasten the widespread deployment of 25G and 50G Ethernet ports that will coincide with the rollout of 100G Ethernet for cloud fabrics.
A 25G/50G standard may seem like a step backward, because 40G and 100Gbps Ethernet already exist, but it’s all about the need for more cost-effective speed, specifically from servers in cloud data centers. For example, 25Gbps cabling is about the same cost structure as 10G at 2.5X the performance. Similarly, 50G is half of the cost of 40G with a 25-percent increase in performance.
* Cloud Scale Networking. How can companies drive down the cost of mobile networking while driving up efficiency and flexibility of their IT infrastructure? The virtualization of networks, storage, and servers is reshaping the way organizations use IT. Cloud computing plays an essential role in this process, providing Internet access to complex applications and massive computing resources.
The cloud delivers the additional capacity required to satisfy growing demand to an enterprise or small business from a third party, giving organizations a way to crank up capacity without investing in a new IT structure. Being able to host processing and data in the cloud frees network administrators to relocate that capacity to geographic locations where the data is being created, as well as control the devices generating the data. A well-designed Ethernet network delivers the low latency and high bandwidth required by cloud computing to deliver the full value of its efficiencies.
* Open-Source Networking. Cloud scale networking is useless without software — every switch needs an operating system. Software must deliver a hardware-agnostic network operating systems that enable an abstraction layer on the merchant-based switch silicon, just like Linux and Windows provide on the server side.
Removing proprietary boundaries and allowing developers to collaborate, debate, compromise and inspire each other is a key benefit of open-source networking. Gartner forecasts that open-source technology will be included in 85% of all commercial software packages by 2015, and 95% of mainstream IT organizations will leverage some element of OSS.
For decades, customers have been sold proprietary combinations of servers, data storage, networking and software. Now, these same customers want systems that can be arranged in different ways, combining open-source software with off-the-shelf hardware and their own in-house legacy systems. This mix-and-match approach promises advantages in speed, cost and innovation.
The traditional, one-vendor, proprietary solution is transitioning to solutions involving many suppliers. As with any market, this offers customers with significant cost savings and performance optimization. The challenge now shifts to the successful integration of systems from those many suppliers. Fortunately, the server and PC market have provided a template to handle such integration with new open sourced hardware, open source software and relevant standards to tie it all together.
Ilyadis is Vice President and Chief Technical Officer of Broadcom’s Infrastructure and Networking Group (ING), where he is responsible for product strategy and cross portfolio initiatives for a broad portfolio of Ethernet networking products. Reach the author @broadcom.