[802.3_50G] 答复: [802.3_50G] 200GbE will be Different and Lower Cost/bit than 400GbE

["Wangxinyuan (Xinyuan)" <wangxinyuan@xxxxxxxxxx>]

FlexE is a logic layer definition for bonding 100GBASE-R, clause 82 as in latest Draft 1.3, to get "mX25Gbps" Pipe (FlexE Client).  It doesn't define module form factors to improve port density, the most possible 400Gbps FlexE Pipe will bond CFP2 right now and QSFP28/CFP4 in other application.

FlexE is a better solution in some specified application, for example router to transport connection. But it requires parallel fiber to bond bunch of 100GBASE-R PHYs, this will introduce addition fiber cost and logic cost to deskew fiber length difference, where 400GE will benefit industry for duplex fiber application as in 802.3bs objective.

 

---

发件人: Brad Booth [bbooth@xxxxxxxx]
发送时间: 2016年3月16日 17:51
收件人: STDS-802-3-50G@xxxxxxxxxxxxxxxxx
主题: Re: [802.3_50G] 200GbE will be Different and Lower Cost/bit than 400GbE

FlexE across four 100G ports doesn't result in 400G. It does result in a 400G pipe, but what is the MAC behind it. I could also combine two 2x100G modules via FlexE to create 400G across two ports. Or, I could combine 4x100G in a COBO module and only burn one port.

None of these combinations are truly 400G unless the MAC is 400G.

Cheers,

Brad

On Wed, Mar 16, 2016 at 2:37 AM, John D'Ambrosia <jdambrosia@xxxxxxxxx> wrote:

Brad,

I have talked to both Gary and David on this ? and this is what I understand from my conversations with them.

 

Perhaps you would care to share why you think I am wrong.

 

J

 

From: Brad Booth [mailto:bbooth@xxxxxxxx]
Sent: Wednesday, March 16, 2016 4:29 AM
To: STDS-802-3-50G@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_50G] 200GbE will be Different and Lower Cost/bit than 400GbE

 

John,

 

Your understanding is wrong.

 

Cheers,

Brad

 

On Wed, Mar 16, 2016 at 1:17 AM, John D'Ambrosia <jdambrosia@xxxxxxxxx> wrote:

Brad,

And what happens when someone wants a true 400GbE building block?

 

As I understand it if I have Flex-E with 4 100G ports I get 400G ? but I have also now burned 4 ports ? this will do nothing for face plate density ? which is also an issue that we cannot ignore.  We can’t just look at a single instance.

 

John

 

 

 

From: Brad Booth [mailto:bbooth@xxxxxxxx]
Sent: Wednesday, March 16, 2016 3:34 AM
To: STDS-802-3-50G@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_50G] 200GbE will be Different and Lower Cost/bit than 400GbE

 

Scott,

 

It's a shame you couldn't make it to Macau. You highlight exactly the point some of us are making, "I will argue that FlexE and 100GbE and 200GbE will be more cost effective than 400GbE."

 

If you're correct, then any 8 lane 400G will not satisfy the Economic Feasibility requirement. If it cannot satisfy the EF requirement, then there isn't likely to be any Broad Market Potential for those solutions.

 

Using the 4:1 PMD logic, the only 400G PMD capable of satisfying EF and BMP is the DR4.

 

All I and others are asking for is honesty in our CSD.

 

Thanks,

Brad

 

On Wed, Mar 16, 2016 at 12:03 AM, Scott Kipp <skipp@xxxxxxxxxxx> wrote:

Jeff,

 

You do raise a good point about 2 lane interfaces to servers.  When I interviewed Alan Weckel of Dell’Oro for the Analyst Hour on “50GE and the Path to 200GE and 400GE” ( available for free at https://www.brighttalk.com/channel/6205/ethernet-alliance), he pointed out how they expect about a million ports of 25G servers will ship and another million ports of 50G (2X25G) servers to ship this year (slide 24).  By 2019, they expect 50G servers to outship 25G server ports by about a million ports (5M to 4M), but the servers will be a mixture of the 50G serial lanes by then.  So I have considered 2 lanes servers and 2 is a nice multiple of 4 as well. 

 

While some 8 lane modules are being proposed like QSFP-DD, CFP8 and On Board Optics, 400GbE is not defining an copper interface for servers or an 8-lane MMF interface.    We can’t break a DR4, FR8 or LR8 out to 50G servers.  400GbE is not for breakout to servers.

 

As Gary showed in an excellent presentation in the Study Group phase for the 400GbE (http://www.ieee802.org/3/400GSG/public/13_11/nicholl_400_01_1113.pdf), 400GbE is for routers, not servers.  The conclusion of that slide says, “Early market applications for 400 Gb/s Ethernet will be similar to those seen in early market 100 Gb/s Ethernet”.  I agree.  400GbE will not see high volumes until it has 100G lanes and fit in a QSFP.  That will be in the next decade.

 

What I don’t understand is why people think that 8X interfaces are suddenly going to be low cost when nothing has changed and the speeds are getting more challenging.  Even if we defined an 8X50G MMF interface for 400GbE, the 8X module will not see the volumes that drive cost down like the ubiquitous and beloved QSFP.  Volume drives cost down and investment up.

 

Can anyone explain how a low volume 400GbE product based on 8 lanes of 50G is going to get to low cost? 

 

Another way to look at this is what will see volume?  A low cost QSFP or a CFP8?  This has played out a couple times now in favor of the QSFP.

 

These topics have been on my mind quite a bit because I’ll be speaking about this next week at OFC on the Thursday panel called “A Rational Assessment of 400GbE”.  I will argue that FlexE and 100GbE and 200GbE will be more cost effective than 400GbE.  I hope you can make it.

http://www.ofcconference.org/en-us/home/exhibit-hall/market-watch/panel-v-a-rational-assessment-of-400g-ethernet/

 

Thanks,

Scott

 

From: Jeffery Maki [mailto:jmaki@xxxxxxxxxxx]
Sent: Tuesday, March 15, 2016 10:47 PM
To: Scott Kipp <skipp@xxxxxxxxxxx>; STDS-802-3-50G@xxxxxxxxxxxxxxxxx
Subject: RE: 200GbE will be Different and Lower Cost/bit than 400GbE

 

Scott,

 

For applications, we see this 4 to 1 ratio such as 200GE and 4 x 50GE OR 400GE and 4 x 100GE. What you do not consider is that the same base technology can do both, where that base technology is 50G lanes. The base technology factors are 4 & 1 OR 8 & 2. You are not accounting for server connects that use 2-lane interconnects.

 

We might argue that 400GE will not be at the sweet spot until it is made using 4 x 100G-lanes and server connects also use a 100G lane.

 

In this way, you do not have to get into saying 400GE is only needed for routers today. There are data center applications.

 

Jeff

 

 

From: Scott Kipp [mailto:skipp@xxxxxxxxxxx]
Sent: Tuesday, March 15, 2016 3:57 PM
To: STDS-802-3-50G@xxxxxxxxxxxxxxxxx
Subject: [802.3_50G] 200GbE will be Different and Lower Cost/bit than 400GbE

 

All,

 

I wanted to comment on some of the presentations that are coming out against 200GbE in the possibly final days of the NGOATH Study Group. 

 

As I showed in kipp_50GE_NGOATH_01_0116.pdf, 200GbE can be designed for low cost QSFP implementations from the start.  This makes 200GbE applicable to the low cost / high volume switch market while 400GbE will be suited for the high cost/ low volume router market.  200GbE will be a continuation of the successful, high volume progression from 40GbE QSFP+ to 100GbE QSFP28 to 200GbE QSFP56.  Highly parallel Ethernet interfaces more than 4 lanes wide like 400GbE have never reached high volume (more than 1M ports/year) and there is no reason to think this will change. 

 

According to Dell’Oro, more 100GbE switch ports shipped than 100GbE router ports last year for the first time.  This year, over 10X as many 100GbE switch ports are expected to ship than 100GbE router ports.  Routers ship low volumes of high speed ports initially, but switches overwhelm routers in port shipments when the technology is ready and the cost is low enough for switching.  This pattern of switching dominance is set to repeat again at 400GbE.  400GbE will ship for routers while 200GbE can come out of the gates in high volume like 40GbE.

 

Here’s a quick comparison of the differences between 200GbE and 400GbE:

 

Characteristic	200GbE	400GbE
Fits in QSFP Family of Modules	Yes, like 40GbE and 200GbE	No, requires 8 or 16 lanes
Supports Low Cost Copper	Yes, twinax and backplane	No copper here
Uses existing MMF cabling	Yes, 8 or 12 fiber MPO	No, requires new 32-fiber cabling
Supports low cost SMF modules	Yes, CWDM4 and LR4 in QSFP28	No, 8 lanes with tight wavelength spacing in higher power module
Target Market	Low cost switching	Premium routers

 

In booth_50GE_NGOATH_01a_0316.pdf, the presentation claims: 200G MAC to MAC provides no value.  I heard similar arguments against 40GbE in 802.3ba and 40GbE took off in switching in 2011 while 100GbE stayed high cost in routers until a 4-lane solution became available.  200GbE will add value by being lower cost/bit.  No Ethernet interface has proven to be low cost when it is highly parallel (>4 lanes).  Ethernet interfaces can sell millions of ports per year when they reach 4 lanes wide or less.

 

In nicholl_50GE_NGOATH_01_0316.pdf, the presentation claims “both a Standards and product viewpoint, 200GbE and 400GbE are likely to come out at the same time”.  I agree that the standards could come out at roughly the same time, but the products will be mainly different in character than in time.  High cost 400GbE will only be available in routers while low cost 200GbE will be in switches.  Go to any systems vendor’s website and there will be completely separate product lines for switches and routers.  Switches are for high volume and low cost while routing has very different and challenging requirements that are suited for high bandwidth deployments. 

 

I’m glad the objectives for 200GbE are already accepted by the Study Group.  I hope enough of you agree to keep 200GbE objectives in the project and prevent last minute changes with little justification beyond FUD.  End users will flock to 200GbE when they see a lower cost/bit performance - just like they did for 40GbE.

 

Regards,

Scott Kipp