400G标准启程 以太网创新再续传奇

400G标准启程 以太网创新再续传奇

诞生40年来，以太网创造了一个崭新的连接世界的奇迹。而今，迈入不惑之年的以太网仍然生机勃勃，继续书写着创新的传奇篇章。

前不久，国际标准组织IEEE802在北京召开全体会议期间，《网络世界》出版人兼总编辑高辉采访了作为IEEE802执行委员会记录秘书的以太网联盟主席、戴尔网络首席技术官办公室以太网传播总负责人John D'Ambrosia，请他就以太网最新技术进展与未来趋势和本报读者分享看法。

400G标准制定进程启动

在以太网历经1G、10G、40G、100G的速度提升之后，如今正在迈向令人激动的400G。本次IEEE802全体会议期间，802执行委员会首先批准了400G以太网标准工作组成立(其后的3月27日，在获得802标准委员会通过之后，名为P802.3bs 的400G标准工作组正式成立，由John D'Ambrosia担任工作组主席，第一届工作组会议将于今年5月在美国弗吉尼亚州北部举行)。

在之前John D'Ambrosia负责的400G以太网研究组工作中发现，在计算领域，带宽能力每24个月翻一倍，而网络带宽则每18个月翻一倍。据其预测，平均下来，到2015年网络必须支持TB每秒的能力。到2020年必须支持10TB每秒的能力，这一数字是2010年的 100倍。

“业界一直称我为致力于让以太网向更高速发展的推动者。然而，以太网的快速发展，让我这个一直强调高速率的都感到吃惊!” John D'Ambrosia这样表示。他说：“自从40Gb/s 和 100Gb/s以太网标准制定之后，以太网联盟看到了业界对于更高速度的不断需求。例如，车载以太网应用就将带来海量用户和应用，因此导致巨大的带宽需求。为了满足这样的带宽需求，以太网必须向更高速度进发。”

John D'Ambrosia对以太网创新带来的巨大可能非常兴奋，他告诉《网络世界》记者，他领导的以太网联盟非常欢迎IEEE P802.3bs的成立，并已成立了400G分委员会，从而帮助400GbE标准的制定，为促进以太网向下一代更高带宽演进提供支持。

以太网创新并非轻而易举

由于应用层创新异常丰富并能够让最终用户感同身受，相比而言，作为基础联网技术的以太网就与人离的较远，因此许多人就认为以太网已经非常成熟，而近年来创新的东西似乎并不太多。对此，John D'Ambrosia表示反对。

“以太网领域有很多层出不穷的创新。在过去短短两年当中，我们就有12个项目在进行。比如我们做了很多工作来研究如何改善以太网供电、支持车载以太网、不同的铜介质和光学介质的以太网传输方式，以及提升以太网速度等等。所以在我看来，市场上关于以太网没有太多创新这样的说法是因为被宠坏了——好像以太网实现的一切都是理所当然，直接接入就可以了。事实上绝没那么简单。” John D'Ambrosia还认为，现在的问题并不在于创新不够，而是在于创新产生的新技术产品的成本是目前市场暂时不能承受的。

这几年网络业的热点变化非常快，从数据中心网络[注]的TRILL和SPB标准之争，到目前的SDN[注]、NFV[注]、OpenFlow大热，短短几年时间里，热点层出不穷且迅速转移。对此，John D'Ambrosia说：“把自己陷入一些混乱的技术术语当中是件非常容易的事情，一些人可能希望用这样一种方式使自己显得与众不同或显得更有知识和能力。如果我们忙于迎合所有的新概念，迎合所有人的需求，那就会失去自己的方向。所以在过去几年，我一直提醒自己要专注在重要的关键领域当中。”

的确，创新无诀窍，只有依靠一步一个脚印的艰苦工作才会有收获。让我们一起加入以太网创新洪流，为之推波助澜。

英文译文

400G Standard Sets Sail, Ethernet Innovation Builds Upon Its Legend

BY Meng Ke,China Network World

40 years ago, Ethernet was born and created a technical breakthrough by connecting the world. Today, Ethernet continues to build upon its legacy in innovation.

During the IEEE 802 Plenary in Beijing, Network World publisher and chief editor Gao Hui interviewed John D'Ambrosia, Chief Ethernet Evangelist, CTO Office, Dell and Chairman, Board of Directors, Ethernet Alliance where he discussed Ethernet’s latest advancements and future trends.

Kicking off 400G standards

After 1G, 10G, 40G and 100G, Ethernet is breaking through yet again with exciting 400G. At the IEEE 802 Plenary, the IEEE802 executive committee first approved the foundation of the 400G Task Force. (On March 27, after being approved by the IEEE802 standard committee, the 400G Task Force called P802.3bs was officially founded with John D'Ambrosia as the chairman. The first meeting of the Task Force will be held at Virginia in May.)

Research from the 400G Ethernet study group led by John D'Ambrosia had several findings: for computing space, bandwidth capabilities were doubling every 24 months, while network applications were doubling every 18 months. It is forecasted that on average, networks would need to support terabit per second capacities by 2015 and 10 terabit per second capacities by 2020. This is 100 times more than what was required in 2010.

“The industry has always viewed me as the driver of higher speed Ethernet. However, its rapid development surprises even me, a person who has always been pushing for higher speeds,” said John D'Ambrosi. “Since 40Gb/s and 100Gb/s was defined, the Ethernet Alliance saw the industry’s ongoing demand for higher speed. For example, in-vehicle Ethernet applications will bring more users and create more applications, which will create enormous bandwidth demand. To meet these demands, Ethernet has to increase speeds.”

John D'Ambrosia is very excited about the new possibilities created by Ethernet innovations. He said the Ethernet Alliance under his leadership welcomes the establishment of IEEE P802.3bs and has established a 400G committee to help with 400GbE standards and support Ethernet’s evolution of higher speeds.

Ethernet Innovation is Not Easy

While application innovations are abundant and clearly visible to end users, Ethernet, a basic connectivity technology, is more hidden to end users. Therefore, many believe that Ethernet has already matured and innovations in recent years are few in numbers. John D'Ambrosia disagrees with this.

“We are working on a number of innovations in Ethernet. Within the past two years alone, we’ve taken on 12 different projects. For example, we’ve conducted in-depth research on how to improve power over Ethernet, in-vehicle Ethernet, Ethernet transmission over different copper connections and optical connections, as well as improving overall Ethernet speed. I believe that the notion of a lack of Ethernet innovation is unfounded. It seems like everything about Ethernet has been taken for granted and people believe that you can just connect. It’s not that simple,” said John D'Ambrosia. The problem is not that there is a lack of innovation, but that the new technologies and products available today are not affordable.

The networking industry’s hot spots have changed rapidly, from TRILL and SPB of datacenter networking to SDN, NFV and OpenFlow in recent years. John D'Ambrosia said: “It’s very easy to get caught up in the technical jargon and some even use this to show off. If we just embraced any new concept and demand, we would lose our direction. So I’ve been reminding myself to focus solely on the important aspects over the past years.”

Indeed, there is no shortcut to innovation. You can only reap when you have spent the time and effort to sow. Let’s continue to ride the wave of Ethernet innovations.