Ethernet Alliance

Ethernet Alliance Blog

Will 100GBASE-LR4 be the Only IEEE 100GbE Single-mode Link Type?

By Scott Kipp

At the November 2012 IEEE 802.3 Plenary meeting in San Antonio, TX, several straw polls were taken to see if the 40 Gb/s and 100Gb/s Fiber Optic (802.3bm) Task Force was reaching consensus on a new 100G single-mode objective for at least 500 meters.  The straw polls showed that no solution was even close to reaching the 75% majority needed to become a standard, so it looks like 100GBASE-LR4 may be the only 100GbE solution for link distances between 150 meters and 10km.

The most telling straw poll and result was Straw Poll #5 that asked:

I would only support a baseline proposal for a single-mode fiber PMD (Physical Medium Dependent sublayer or link) based on:

a) CWDM – 0%

b) C-BAND – 2%

c) DMT – 10%

d) PSM4 – 24%

e) PAMn – 24%

f) none of the above – rely on LR4 with CAUI-4 – 40%

a: 0    b: 1    c: 6     d: 14     e: 14     f: 23 = Total of 58 voters

This had the surprising result that 40% of the people voted to not define a new single-mode link type and rely on 100GBASE-LR4 for all 100GbE links beyond 150 meters that are supported by 100GBASE-SR10 over multimode fiber.  Without a large change of the votes in the task force, the 100GBASE-LR4 technology will be the only solution and become lower cost as volume production rises.

The 802.3bm project will enable lower cost 100GBASE-LR4 modules by defining the CAUI-4 electrical interface at 4X25Gbps to eliminate the 10:4 Gearbox from inside the optical module.  With the removal of the gearbox from the module, the 100GBASE-LR4 solution can be deployed in the more compact CFP2 form factor that supports 4 lanes of 25Gbps instead of the CFP form factor that can only support CAUI with 10 lanes of 10Gbps.  The 4X25G CAUI-4 interface will do a lot to improve the cost structure of 100GBASE-LR4 and integration of optical components, and volume will drive down prices further.

Many proposals to 802.3bm predict lower cost 500m modules than 10km 100GBASE-LR4 modules.  In addition to going a shorter distance that lessens the requirements on the lasers, Forward Error Correction (FEC) could be used to soften the requirements on the link.  FEC is being used in 802.3bj and may be applied to 100GBASE-SR4 to lighten the requirements and lower the cost and power of the module.    

The straw poll shows that the task force is not yet convinced that they want to move forward with standardizing a new single-mode PMD.  No proposal has yet achieved 25% of the required 75%.  If no proposal is able to achieve the required 75% by May of 2013, then the industry will be left with 100GBASE-LR4 for reaches greater than 150 meters.  

One goal of the task force is to create lower cost and lower power modules than 100GBASE-LR4 for reaches of at least 500 meters.  The lower optical power budgets for 500 meter links enable lower power modules and different technologies than were available when 100GBASE-LR4 was defined.  Given the cost sensitivity of various applications, the reduction in cost and power of 100GbE modules that enable higher switch port densities will be key to the wide scale deployments of 100GbE, especially for data centers.

By Scott Kipp

President of the Ethernet Alliance

Share

The views and opinions expressed in this blog are solely that of the individual(s) and should not be considered the views or positions of the Ethernet Alliance.

Share

The Power and Promise of Energy Efficient Ethernet (EEE): A State of the Union Address By Wael William Diab

By Wael William Diab

Power-hungry IT equipment generates some 150 million metric tons of CO2 annually at a global cost of $16 billion[i]. With an eye toward reducing both energy costs and our carbon footprint, it has become increasingly imperative that technology developers and end users alike seek out and embrace environmentally sensitive solutions.

One innovative technology with the potential to mitigate IT’s impact on the bottom line as well as Mother Earth is Energy Efficient Ethernet, better known as EEE. By reducing power consumption during low data activity cycles, EEE can help lower the amount of energy required by more than half, without interfering with or disrupting equipment operations. It is a promising advancement enabling users to leverage Ethernet’s superior flexibility and reliability in a more cost-effective, environmentally conscious way.

So as the year 2012 winds toward its inevitable close, the time is right for a frank assessment of the ever-changing EEE landscape – how far we’ve come, and how much further we need to go before this powerful concept becomes a true reality. The good news is that this year has been fruitful for the forward-looking individuals and enterprises that are the driving force behind EEE’s continued growth and evolution. This dedicated group of key stakeholders has accomplished much throughout 2012, including the adoption of new EEE-focused objectives across multiple new IEEE 802.3TM standards projects.

For example, the IEEE P802.3bjTM 100 Gb/s Backplane and Copper Cable Task Force recently elected to modify its scope of work to include specification of optional EEE for 40 Gb/s and 100 Gb/s operation over backplanes and copper cables for both the newly specified interfaces by the P802.3bj, as well as the relevant interfaces specified in IEEE Std 802.3baTM-2010. However, the IEEE P802.3bj Task Force isn’t alone in its efforts to advance EEE across the IEEE 802.3 ecosystem; momentum is also growing within other IEEE 802.3 groups, generating a new round of interest and activity from group members.

Other IEEE groups following in the footsteps of IEEE P802.3bj are the IEEE P802.3bmTM 40 Gb/s and 100 Gb/s Operation Over Fiber Optic Cables, the IEEE P802.3bnTM EPON (Ethernet Passive Optical Networks) Protocol over Coax Networks (EPoC) and IEEE P802.3bpTM Reduced Twisted Pair Gigabit Ethernet (RTPGE) PHY Task Forces. All three groups have adopted new objectives covering the specification of EEE for their respective technologies:

  • IEEE P802.3bm: Similar to the IEEE P802.3bj project, the IEEE P802.3bm project is adding EEE capability to both the newly specified 40G and 100G optical interfaces by P802.3bm, as well as the relevant interfaces specified in IEEE Std 802.3ba-2010.
  • IEEE P802.3bn: This project seeks to expand the popular EPON access technology into new and existing coax networks so that the EPON protocol can transparently run over either fiber or coax. At the November plenary meeting, the group added an objective for EEE operation, reflecting wide interest from key stakeholders and industry in power savings.
  • IEEE P802.3bp: This project looks to expand Ethernet into automotive and other new markets by reducing the number of pairs for a gigabit link over twisted pair copper. With encouragement from participants affiliated with the automotive industry, the group is pursuing the development of in-vehicle Ethernet connectivity that incorporates support for EEE, which has been reflected in an EEE objective.

Additionally, we’ve seen new activity in the rest of IEEE 802.3, such as the Next Generation BASE-T study group, which is studying the next generation of the popular BASE-T family. As of the writing of this post, the group has adopted a 40G rate along with an EEE objective.

Ethernet has become the ubiquitous wired layer-2 connectivity. In an environment where energy consumption is at the forefront of users, manufacturers, and deployers minds, it stands to reason that there we will continue to see standards innovate to accommodate this concern. With the New Year underway, we’re ready to move beyond the question of can we make EEE a reality within these projects and instead focus on the “how”. As these projects progress through the standards development cycle, we’re ready to select needed technical proposals and produce the resulting standards. These efforts will help complement the EEE deployments associated with the widely adopted IEEE Std 802.3azTM-2010, making EEE a ubiquitous reality in all the major markets that Ethernet services.

About The Author

Wael William Diab, is a Senior Technical Director in the Office of the CTO, Broadcom Corporation, Vice-Chair of the IEEE 802.3 Ethernet Working Group, serving in that position since 2008, Vice-Chair of the IEEE-SA Corporate Advisory Group (CAG)



[i] Source: ITU World Summit on the Information Society

 

Share

The views and opinions expressed in this blog are solely that of the individual(s) and should not be considered the views or positions of the Ethernet Alliance.

Share