The future of Ethernet
With Ethernet becoming the ubiquitous connectivity standard for service providers, enterprises, and military applications, we are letting go of proprietary networking technologies and heading directly into industry standard networking based on Ethernet.
For those of you new to the Ethernet game, let’s first establish that Ethernet standards are collaboratively developed, challenged, and approved by the Institute of Electrical and Electronics Engineers (http://ieee.org/). To ensure interoperability among all electrical and networking devices, this standards body reviews and ratifies and publishes the standards from anything electrical to anything networking. Within computing and networking, the Ethernet Alliance (EA, ) has been established as a global, nonprofit consortium of vendors, industry experts, universities, and governments that work together to bring Ethernet standards to the marketplace.
The EA is key to making Ethernet technology open and innovative, and tried and tested. For all of us tech geeks, the EA previously published a great poster; the latest one is the Ethernet Roadmap.
So where are we now? Ethernet speeds are running anywhere from 10 megabits per second (Mbps) to 400 gigabits per second (Gbps). For easy reference, let’s start somewhere everybody is familiar with: home Internet connections. Although it’s now rare, some rural home connection speeds can be as low as 10 Mbps. Most high-speed Internet home services are currently maxing out at about 60 Mbps for copper-based services, while fiber services are now offering “near-gigabit Ethernet” speeds.
In the enterprise, connectivity options from service providers mean that workstations are equipped with 1000BASE-T network interface cards (NICS) and enterprise LANs run gigabit Ethernet (1000BASE-T) switches and routers to support connectivity between locally connected devices. In enterprise data centers, most server-to-server connectivity – up until the last two years – were running over fiber cables connecting the devices at speeds of up to 10 Gbps. However, with the enormous amounts of data and the advent of cloud-based services, data centers adopted 40 Gbps technology faster than any other Ethernet technology in the history of networking. Now data center-to-data center connectivity (between two data centers or between a data center and a service provider network) is running at either 40 Gbps or 100 Gbps (a standard ratified by the EA and IEEE back in 2010).
Currently being tested and ratified are technologies that will bring speeds up to 400 Gbps, typically used by service providers and large-scale cloud providers, and 50 Gbps (expecting ratification in 2018-2019) that will supplant 40 Gbps connections in data centers because of its more efficient use of electrical “traffic lanes.”
So where does Ethernet sit for military applications? While central data centers and field offices are running at the same pace of traditional enterprises, field applications are lagging behind a bit on the adoption curve. Examples of less-complex network applications include dismounted soldiers or small unmanned aerial systems, where compute and communications devices just need to be able to communicate with each other. Gigabit Ethernet, carrying traffic at speeds of as fast as 1,000 Mbps/1 Gbps, is a de facto standard in mobile platforms that need to support higher volumes of traffic, delay sensitive data (video, sensors, VoIP), or carry multiple devices on the same platform (computer, IP phone, IP camera, sensors, etc).
As military programs look to the future – where mobile platforms such as the Joint Light Tactical Vehicle, tanks, and mobile missile launchers will require hundreds of sensors for situational awareness and feeding high-resolution, high-frame-rate cameras for intelligence – current programs are looking to implement 10 Gbps links for on-vehicle aggregation of this traffic to computers or video cameras.
Forward-looking programs are also looking at 40 Gbps links and communications hubs, such as Ethernet switches, simply because the efficiency of running data through four lanes of 10 Gbps will make data communication almost immediate and deliver zero lag time for on-vehicle communications. Having said that, it is important to note that while 40 Gbps traffic will be efficient while onboard the vehicle, traffic and communications speeds traveling off the vehicle to central command will still be limited by current wireless connectivity rates running at about 150 Mbps. Yes, 150 Mbps. This disparity illustrates the importance of having all possible intelligence and compute power on the vehicle so that only mission-critical data will have to be sent back to central command.
What does the future hold? As of 2017, the EA has already done formal functional and interoperability testing with vendors offering 400 Gbps routers and switches. Between 2018 and 2020, 50 Gbps and 200 Gbps will be tested and adopted. Thousands of 25 GbE servers – and eventually 50 GbE servers in hyperscale data centers, such as cloud service providers – will drive the need for 400 GbE to the metropolitan area networks (MANs) and wide-area networks (WANs). In the not-too-distant future, testing will begin on 200 Gbps, 8,000 Gbps, and, astonishingly, data rate speeds of 1 Tbps and 1.6 Tbps, with expected testing and ratification of the standards by the year 2020.