x
Optical/IP

10-GigE Vendors Get Cold Feet

In the last couple of weeks, two performance tests for 10-Gigabit Ethernet have been canceled due to a lack of vendor interest.

The Institute of Electrical and Electronics Engineers Inc. (IEEE) finally ratified the 802.3 specification for the technology on June 13, 2002. And even though several companies claim to have 10-Gbit/s products ready, many are balking at the chance to test their products’ performance. In the past few weeks, a live demonstration at the 10 Gig Inaugural Expo has been canceled, and Meir Communications has called off its 10-Gig Ethernet test. Why aren’t vendors interested?

“No one came right and out and said that they didn’t want to participate because their box couldn’t perform at line rate,” says Mike Bennett, senior network engineer at Lawrence Berkeley National Laboratories, which was putting together the 10-Gbit/s demonstration for the Inaugural Expo. “But let’s just say they haven’t been beating down our doors to be included. You can draw your own conclusions.”

OK. We will.

The Inaugural 10 Gig Expo was supposed to get underway next week in Phoenix (see Show to Demo 10-Gig Throughput). Bennett and his team from Berkeley Labs had designed a test bed using two sets of 12 high-end computers to pump traffic through the network at 10 Gbit/s. Some vendors had already signed up to participate, but many at the last minute declined or pulled out. Finally, two weeks ago, the show’s sponsor, Pinnacle Conference Network, was forced to call it off.

Last week, Meir Communications, an independent Gigabit Ethernet testing service, was supposed to begin testing gear for an article that was scheduled to appear in the August edition of Business Communications Review. Once again, a lack of vendor interest killed plans for that test. Sources say that BCR is still expected to run a story, but instead of testing all the 10-Gbit/s players, it is simply asking vendors to fill out a survey of their capabilities.

There are several switch companies already shipping, or at least beta testing, 10-Gbit/s Ethernet products. Many of them participated in the 10 Gigabit Ethernet Alliance (10GEA) interoperability demonstration last month at Network + Interop in Las Vegas (see Vendors Show Off 10-GigE at N+I). These included switch vendors like Cisco Systems Inc. (Nasdaq: CSCO), Enterasys Networks Inc. (NYSE: ETS), Extreme Networks Inc. (Nasdaq: EXTR), Force10 Networks Inc., Foundry Networks Inc. (Nasdaq: FDRY), and Nortel Networks Corp. (NYSE/Toronto: NT). Riverstone Networks Inc. (Nasdaq: RSTN), which has not yet shipped its 10-Gbit/s product, did not participate in the N+I demo.

Some of these companies, like Foundry and Force10, agreed to participate in the 10 Gig Inaugural Expo demonstration. But many, like Cisco and Riverstone, declined. As for the BCR test, Meir Communications would not reveal which vendors declined to be tested. Force10 says it is still willing to have its gear tested by both testing houses; it will be used in a demonstration staged at the Lawrence Berkeley Lab facility later this month. Cisco says it plans to provide BCR with the necessary specification information for the August story.

“The reason we decided not to participate is basically a resource issue,” wrote Larry Yu, Cisco spokesperson, in an email to Light Reading. “We had to choose between an upcoming third-party 10GigE interoperability test at the end of the month and this. We chose to participate in the interoperability test.”

Cisco is planning to participate in an interoperability test run by the Tolly Group.

It seems clear from the choice of tests that vendors are more comfortable in an interoperability demonstration than in a performance test. And who can blame them? According to Bennett, who is familiar with most of the products available today, Force10 Networks is the only company that even comes close to achieving 10-Gbit/s line-rate throughput (see Force10 Shows Off 10-GigE Switch). He says the startup’s switch typically achieves forwarding throughput between 9.0 Gbit/s and 9.5 Gbit/s, depending on packet size.

“It is somewhat misleading to customers, because if you say it is a 10-Gbit/s product then I expect it to forward at line rate,” says Bennett.

Then, of course, there's the issue of customers. Despite the hype surrounding 10-Gbit/s Ethernet, the fact is that most service providers aren’t ready to offer it. Tier 1 service providers in the U.S. aren’t expected to deploy 10 Gbit/s for at least 12 to 18 months, according to Mark Sue, an analyst with Frost Securities Inc. Large interexchange carriers like WorldCom Inc. (Nasdaq: WCOM) and RBOCs like SBC Communications Inc. are just now rolling out Ethernet services (see WorldCom Unveils Metro Ethernet). Meanwhile, Ethernet service providers like Yipes Communications Inc. and Sigma Networks are filing for bankruptcy protection and selling off assets (see Another Metro Provider Fails: Was Vendor Financing the Difference?) and Yipes Joins Chapter 11 Club).

While large carriers typically adopt new technologies at a slower rate than new players, 10 Gbit/s may also be cost prohibitive right now. Initially, 10-Gbit/s Ethernet will be expensive, with per-port costs as high as $80,000. And as T1 costs drop from $1,000 per month to $750 per month, there is little incentive for customers to switch their service to Ethernet.

But analysts expect prices to drop drastically over the next few years, just as 100-Mbit/s Ethernet and 1-Gbit/s Ethernet did. According to IDC, 10-Gig's prices will decline to $7,800 per port by 2005.

After the pain of the initial investment wears off, the technology will likely take off, says Mark Sue of Frost Securities in a research note he published yesterday. He says that with Sonet OC192c interfaces priced at around $300,000, 10 GigE is already a bargain. But many people disagree. They say that Sonet OC192c port prices have also dropped and are now under $100,000.

The true sweet spot for 10-Gbit/s Ethernet, at least in the near term, seems to be in the enterprise. One of the biggest drivers in this market is the sharp decline in pricing of 1000base-T Ethernet interfaces. These network cards are selling for as low as $50 a piece. And with 1000base-T Ethernet running within enterprise networks, 10-Gbit/s Ethernet switches will become necessary for traffic aggregation. Bennett says he also sees supercomputing as another potential market segment.

— Marguerite Reardon, Senior Editor, Light Reading
http://www.lightreading.com
Scott Raynovich 12/4/2012 | 10:13:09 PM
re: 10-GigE Vendors Get Cold Feet Thanks Michael, greet feedback!
Mike_Bennett 12/4/2012 | 10:13:21 PM
re: 10-GigE Vendors Get Cold Feet Folks,

I would also like to add a few points for the sake of clarification.

One thing, perhaps I am reading into the responses, is that I am GǣdissingGǥ the efforts and accomplishments of the IEEE and 10GEA. Nothing could be farther from the truth. I have the utmost respect for those who have worked so hard to create a high quality standard as scheduled. The folks who solved the extremely difficult problems under difficult circumstances in a downturn economy where human resources were dwindling as the months went by deserve enormous respect and accolades.

Regarding the Pinnacle event, the context of our demonstration is that the standard is done Gǣjust in timeGǥ to support research efforts that take place in institutions such as the national labs, University of New Hampshire, San Diego Supercomputing Labs, and many others since that is where the Gǣearly adoptersGǥ of the technology are. As an end user of the technology, like Roy, I have also used SpirentGs equipment for both demonstration and testing. Spirent has been at the forefront of the technology in my experience for 1 Gbe as well as 10 Gbe, in many cases enabling the work necessary to complete the standard to be done.. I also agree that objectives are quite different between testing and demonstrating a given technology. There are also different objectives in different technology demonstrations. As Mark Fishburn so clearly stated in his response, GǣGǪputting together a 20 hop demo with data passed successfully from a generating system through every system vendorGs equipment was just an outstanding achievement. It featured full reach of the optics over 200km. at a good rate if not full rate and was a tremendous testimonial for the plug and play nature of Ethernet in a network containing a larger group of vendors than may ever be seen in a real deployed network.Gǥ I agree, this is quite an accomplishment in anyoneGs book.

With that said, the purpose of LBNLGs demo is to show that we do have real applications that can use the BW right now, not 12-18 months from now as some folks would have you believe. Will sales folks make their numbers based on the volume of these applications? No. Do my clients care about that? No. The point is that I have to deliver networks with the capacity to handle their application(s) in a cost effective manner. Part of providing a cost effective network is building the simplest network possible to meet the end userGs requirements. Why? Because simple is easier to deploy and manage and that affects the bottom line. So, I could deliver 10 gig with any vendorGs equipment. I could use link aggregation as someone suggested in one of these threads to aggregate 10 individual 1 gig links. Link aggregation is not the simplest solution. Admittedly, 10 Gigabit Ethernet is not the lowest cost solution right now, thatGs the price you pay to be on the Gǣbleeding edgeGǥ of the technology. As the number of ports shipped goes up, the prices will come down, Ethernet will become more pervasive outside the traditional LAN and when Ethernet in the First Mile is done, we network engineers will not have to deal with as many protocol translations, perhaps none at all. In the long run, that makes things simpler, cheaper, and everyone wins.

Obviously, I am an Ethernet advocate. While the squabbling here and elsewhere will continue regarding the individual vendorGs products being able to deliver line rate, the choice ultimately lies in the hands of the end users. The accomplishment of the IEEE to produce the high quality standard in time to meet our needs transcends all of that. Someone once said, GǣIf you build it, they will come.Gǥ Such is the case for 10 gig and will be the case for the technologies that follow. I, for one, am quite thankful that there are so many brilliant and capable engineers out there willing to take on the task and deliver as promised.

Respectfully,

Michael J. Bennett
Sr. Network Engineer
Lawrence Berkeley National Lab
hoffmane 12/4/2012 | 10:13:32 PM
re: 10-GigE Vendors Get Cold Feet Based on what I have seen of the major vendor 1 Gbe switches, their 10 Gbe will be jitter (latency variation)inducers.

It does not matter if the transmission line is jitterless, going through a layer 2/3/4 switch may produce jitter effects all of its own.

Perhaps Cisco, Foundry, Extreme et.al. should publish figures on multihop latency build up and also transportation of jitter sensitive traffic like video and voice (and not just 1-5 streams, try 80-100+ streams).
hoffmane 12/4/2012 | 10:13:33 PM
re: 10-GigE Vendors Get Cold Feet Probably including the cost of a major OC192 ADM or DCC in this price (expensive, large ADM/DCC).
Roy_Bynum 12/4/2012 | 10:13:33 PM
re: 10-GigE Vendors Get Cold Feet sigint: "Surely, the equipment could derive its timing from the upstream SONET equipment. This could be provided as an option where required."
_________________________________________________

This functionality was specifically excluded from the 10GbE WAN PHY functionality. Ethernet interfaces are peer level respecting transmission clock timing, each transmitter clocked independently, with the far end receiver deriving bit clocking from the embedded data signal. 10GbE was not intended to be a client on a data transmission network, so network source timing was considered to be "out of scope" of this Ethernet development project. If some vendors implement this functionality, it is not part of the 10GbE standard. In the future, there may be a need for 10GbE to be a client to a 40Gb SONET/SDH system. At that time, Ethernet switch vendors may make it a practice to include full support for the ANSI standard for SONET client interfaces.
Roy_Bynum 12/4/2012 | 10:13:33 PM
re: 10-GigE Vendors Get Cold Feet MarkFishburn: "it should be pointed out that putting together a 20 hop demo with data passed successfully from a generating system through every system vendorGs equipment was just an outstanding achievement. It featured full reach of the optics over 200km. at a good rate if not full rate and was a tremendous testimonial for the plug and play nature of Ethernet in a network containing a larger group of vendors than may ever be seen in a real deployed network."
______________________________________________

To those reading this board,

Having put together multi-vendor test beds, I know that it is often more difficult to generate "full rate" data aggregation through multiple vendor and multiple hop/source/sink data links and sources than one might imagine. While it is often easy to "daisy chain" Ethernet switches, making sure that the configurations from the different vendors systems (each with its own set of defaults) do not interfere with other, takes more patience than most realize. Given the nature of many demos, I am surprised that you were able to demonstrate even a Gǣgood rateGǥ.

There is a big difference between demonstrating a technology and stress testing the technology. I am familiar with Spirent test gear as well as other vendors data generation and communications test systems. I have used it to not only demonstrate technology architectures, but also to do stress testing of deployment configurations and new technologies. There are also test system configuration differences between demos, that run for long times, and GǣtestsGǥ that often run for short periods of time.

Running a demo at Gǣa good rateGǥ over a period of time does not mean that either the testing systems or any vendorGs systems that might be put under a bandwidth performance test would not pass a full bandwidth performance test. It means that running the demo over an extended period of time was done at an optimal data transfer rate for the systems and length of time that the demo was running. Actual performance information will have to come from the individual vendors. Even the third party testing labs often do not publish that information because of current NDA practices.

Roy Bynum

sigint 12/4/2012 | 10:13:34 PM
re: 10-GigE Vendors Get Cold Feet Huub_van_Helvoort wrote:

> The whole point of making the 10GigE PHY
> standard was so that> SONET/SDH and 10GigE > would be compatible.

So why did the IEEE make it not fully compatible?
There are enough differences to make it even
incompatible.

At the last ITU-T SG15 meeting operators claimed
that they would NOT transport this signal because
it will not meet the +/- 4.6 ppm clock accuracy.
_________________________________________________

Surely, the equipment could derive its timing from the upstream SONET equipment. This could be provided as an option where required.
Roy_Bynum 12/4/2012 | 10:13:34 PM
re: 10-GigE Vendors Get Cold Feet Hub_van_Helvoort: "So why did the IEEE make it not fully compatible?
There are enough differences to make it even
incompatible.

At the last ITU-T SG15 meeting operators claimed
that they would NOT transport this signal because
it will not meet the +/- 4.6 ppm clock accuracy."
_________________________________________________

The WAN PHY was NOT intended to be compatible with SONET or SDH. It borrowed the overhead of SONET/SDH to provide consistant OAM capability over optical facilities. There is a deliberate need for SONET/SDH translation at the section and line level of the overhead. Only the path overhead and payload was intended to be carried unmodified over service provider optical systems. The translation system is referred to as the "ELTE" by P802.3ae and is referred to as a "Path Relay" by ANSI. The transmission clock tolerance was changed from +/-100PPM to +/-20PPM at the instance of ITU SG15 to support Class B regenerators. (+/-4.6 PPM clock accuracy is for Class A regenerators and Stratum 1 sources.)
st0 12/4/2012 | 10:13:35 PM
re: 10-GigE Vendors Get Cold Feet Markfishburn said:
"It featured full reach of the optics over 200km. at a good
rate if not full rate and was a tremendous testimonial for the plug and
play nature of Ethernet in a network containing a larger group of
vendors than may ever be seen in a real deployed network".
-----------------
Agree about good progress in technology.

I know it is Ethernet. Net has this wonderful rate depression: 14.4 modem usually gives you 2.5, 56 gives a bit of high than 32 and so on. As we talk about killer application, we explore the full rate capability, voice the feature and benefit of the high rate. When it comes to deliveble, we cheer the "near full rate" as "good rate", without even shame about "good rate" is hiding under the name of the "full rate".

Let's just call it as is: 8GbE or 10GbE-(minus). If we can't be honest on the tech issue, may be we should just go to Wall Street join the analysts.

I would prefer someone to state the cause of short coming and the "fixes" implementation timeline in order to achieve the full rate, as we set up to deliver at 1st place. At the moment, just simplely state: we are not there yet (over promise and under deliver).

st
hoffmane 12/4/2012 | 10:13:37 PM
re: 10-GigE Vendors Get Cold Feet 15 switch routers in the same box, connected to a backplane using 8 Gb/s as a link.

So really you have the equiv. of 15 small switches using 802.3AD to link to a backbone switch? Yes?

Must get complicated to setup for a normal routing environment?, Tell the truth is sounds similar to Fore's ASX 1100, smaller switches meshed on the backplane, and and was a nightmare to work with.
hoffmane 12/4/2012 | 10:13:37 PM
re: 10-GigE Vendors Get Cold Feet Cisco was not honest on the 5K, it could never reach 1.2Gb/s. It was lucky to reach 600Mb/s.

The 5.5K was just 3 5Ks welded together, never was there a chance that you could GET 3.6, just around 400-600Mb/s x 3

They also lied when the data came out on their Illegal interfame gap (10Mb/s boards), they told customers it would be fixed in the next release, they are still waiting.

Cisco is the most dishonest one of the vendors out there. I hope the SEC is investigating their use of "Vendor Financed Deals" as normal revenue practices.
hoffmane 12/4/2012 | 10:13:37 PM
re: 10-GigE Vendors Get Cold Feet Check Out Alcatel's new Falcon (8800) specs, same shit, claims big backplane, with 384 Gigabit Ports, only claims 128 ports as wirespeed.

Cisco started this and everybody does it now.
Iipoed 12/4/2012 | 10:13:41 PM
re: 10-GigE Vendors Get Cold Feet 2fast- agree 100%. unfortunately this thread kept going on and on. I even suggested it be put to bed a few days ago. Anyway i can assure you I am done on this thread as well as most others due to the fact I am mostly responding to very technical types with a sales mentality.

If I am positive I can be of value or a topic I am very clear on this board may see a post from me in the future.

I have learned my leason.
2fast 12/4/2012 | 10:13:41 PM
re: 10-GigE Vendors Get Cold Feet wiThDraWing lasT sTaTeMent Sorry lipoed GEttiNg Frustrated with the White BoarD BickeRing. These Message boaRds SHould be MoRE used for ShaRiNG info
2fast 12/4/2012 | 10:13:42 PM
re: 10-GigE Vendors Get Cold Feet idiotic person to be driving customer statistics on white boards! Foundry has Nothing to sell
Iipoed 12/4/2012 | 10:13:42 PM
re: 10-GigE Vendors Get Cold Feet Light-headed you are 100% correct. Next time you are in So. Cal let me know and I will buy lunch. Salesguy's technical speciality. My old SE would not even let me use the white board.
Iipoed 12/4/2012 | 10:13:44 PM
re: 10-GigE Vendors Get Cold Feet geez, never said it was a 480 gbs backplane did I.
There is 480 gbs of total switching capacity in the 15 slot chassis when fully populated.

dude, do yourself a favor since this so important to you and I cannot give you what you are looking for. got to foundrynetworks.com find the nearest sales location contact the SE. Or go directly to Foundry corporate ask for BigIron or NetIron product management.

I am just trying to make sure you get the correct answer in as technical format as you seem to require
MarkFishburn 12/4/2012 | 10:13:47 PM
re: 10-GigE Vendors Get Cold Feet Some of the comments in Maggie's article need clarification.

The comment on the standard Gǣfinally being ratifiedGǥ implies it's late. For the last two years, the 10GbE roadmap published by the IEEE and 10GEA has always showed ratification in Q2 2002. This was achieved exactly as scheduled.

Next. Lets separate Test and Demo.
As Maggie does mention at one point, but not in the headline, the 10GbE Inaugural Expo did not contain a performance test. It was a demonstration.

This conference was scheduled to run a few weeks after Supercomm in Pheonix, Az., in a difficult market for travel expenditure. These events need paying attendees and support from vendors to exhibit and sponsor. The reasons for its postponement or cancellation of the entire conference may have included lack of interest or available equipment for participation in the demo, but this may not have been the only or overriding reason. As a participant in this demo and all the others discussed, and to give comfort to those concerned with the technology, Spirent (my company) has been able to generate and analyze data at full rate with minimum packet sizes since April 2001, so when the vendors are ready to go full rate we will be ready with the test equipment. As with the well-known LihgtReading test on Core routers last year, that Spirent was involved with, we are also ready to explore how the control plane of these new devices (routing, MPLS, etc), has managed the jump to 10GbE.

On the issue of operating at wire rate: while products in this market mature, already currently shipping products that can already demonstrate more than (say) 4-5 Gbps can show the benefits of 10GbE even below full line rate, in, for instance, link aggregation applications. Add to that the anticipated rapid decrease in port costs for 10GbE anticipated over the next 12 months.

Finally I think that the comment Gǣmore comfortable with interoperability demosGǥ does not do justice to achievements in this area. As someone intimately involved with the 10GEA interoperability demos at N+I and Supercomm, it should be pointed out that putting together a 20 hop demo with data passed successfully from a generating system through every system vendorGs equipment was just an outstanding achievement. It featured full reach of the optics over 200km. at a good rate if not full rate and was a tremendous testimonial for the plug and play nature of Ethernet in a network containing a larger group of vendors than may ever be seen in a real deployed network. Even the 3 day hot-stage sponsored by Key3media completed the raw configuration in the first morning. Congratulations again to all the companies involved and the work of University of New Hampshire and San Diego Supercomputing Lab in this effort.
light-headed 12/4/2012 | 10:13:49 PM
re: 10-GigE Vendors Get Cold Feet you are trying to have a "technical" discussion with a sales guy about marketing numbers... i hope that sounds as funny to you as it does to me!

the architecture will only support 8 gig per slot at close to wirespeed... end of story. the rest is just cisco/3com/extreme/cabletron marketing spin as done by fdry. they cannot even put in a 10 gig card without it being somewhat blocking by their own admission:

GǣWe designed the BigIron back in 1998 before 10-Gbit/s Ethernet was even on the radar,Gǥ says Marshall Eisenberg, director of product marketing for Foundry. GǣWhen itGs fully populated, the switch will be 20 percent blocked. ThatGs just the limitation of the hardware.Gǥ

http://www.lightreading.com/do...
ntwkeng 12/4/2012 | 10:13:50 PM
re: 10-GigE Vendors Get Cold Feet It sounds like disingenuous marketing and byte doubling to me.

Please explain how a 480Gb/s Backplane is arrived at with 8Gb/s backplane interconnect.

This would be identical to Cisco's 6500 with 32Gb/s local switching (checked the web site) and 13 slots. But I do not see them billing it as a 416Gb/s switch.

Also- how do you get 32Gb/s with 16GbE interfaces?
Iipoed 12/4/2012 | 10:13:52 PM
re: 10-GigE Vendors Get Cold Feet Each blade has 32 gbs of local switching capacity.
The backplane has no actual capacity. A fully loaded 15 slot chassis equate 15 x 32 = 480 of total switching capapcity.

Hopefully you understand Foundry's architecture. Each blade is in fact a switch/router. You can switch or route between ports on the same blade without hitting the backplane. Each blade does connect to the switch fabric via a full duplex 8 gig connection.

I have asked one of Foundry's SEs to further respond to help clear up this issue. I will forward
their response when I receive it.
ntwkeng 12/4/2012 | 10:13:53 PM
re: 10-GigE Vendors Get Cold Feet So you claim that Foundry does not double count their backplane right? I believe that was your previous statement.

So you have 32Gb/s of switching per slot. Now each slot supports 8GbE interfaces. Each slot has 8Gb/s of bandwidth into the crossbar backplane.

How do you get 32 without doubling?

But that is forwarding capacity, and not backplane. The backplane is 8gb/s per slot. At 15 Slots cpaacity your maximum backplane is 240Gb/s with doubling.

Huub_van_Helvoort 12/4/2012 | 10:14:07 PM
re: 10-GigE Vendors Get Cold Feet > The whole point of making the 10GigE PHY
> standard was so that> SONET/SDH and 10GigE > would be compatible.

So why did the IEEE make it not fully compatible?
There are enough differences to make it even
incompatible.

At the last ITU-T SG15 meeting operators claimed
that they would NOT transport this signal because
it will not meet the +/- 4.6 ppm clock accuracy.

.....
Iipoed 12/4/2012 | 10:14:13 PM
re: 10-GigE Vendors Get Cold Feet each connection is full duplex.
The Foundry blades are in reality a standalone switch. If you routing or switching between ports on the same blade you do not touch the backplan. Each blade provides for 32 gigabits of throughput.

All: remember I was a sales guy so if we need to go farther I will get an SE to come into this thread. You can also got the Foundry web site.
Neal 12/4/2012 | 10:14:14 PM
re: 10-GigE Vendors Get Cold Feet
Message 18 mentioned Foundry's demo at N+I Las Vegas of 14 10 Gig Ethernet ports.
The 10 Gig Performance Test equipment was supplied by Spirent Communications

Neal
StartupGrunt 12/4/2012 | 10:14:15 PM
re: 10-GigE Vendors Get Cold Feet To support this, note that the ATCA PICMG 3.1 group elected to standardize backplanes that would support 1GE, 2GE, or 4GE connections in addition to 10GigE XAUI on the theory that 10GigE is not going to be cost-effective for a couple of years.

I also suspect that it isn't going to be needed for that time either, given the likely processing power of CPU and DSP board architectures.
light-headed 12/4/2012 | 10:14:15 PM
re: 10-GigE Vendors Get Cold Feet they are counting useless bandwidth... like local switch capacity vs. capacity to the backplane... you could also count 2x switch fabric speedup which is needed to get non-blocking but is useless as well... all vendor BS... they do 120 Gig, 240 if you use cisco math.
ntwkeng 12/4/2012 | 10:14:16 PM
re: 10-GigE Vendors Get Cold Feet This is a fun thread. So lets take a stab at the math for a second.

8Gb/s per slot. 15 Slots. 15*8 = 120

120G/s in a BigIron 15000. How do you get 480? Apologies for the 512, got it confused with an Alcatel... Double the 120 because EVERYONE since Cabletron started it double counts across the backplane and you get 240.

Cisco says 256Gb/s. 16Gb/s per slot. 8 Slots for payload. 8*16=128. 128 * 2 (double count) = 256Gb/s

Can't say I like it, but I get it. They were honest on the 5K's at 1.2Gb/s and subsequently 3.6Gb/s. No doubling there.
light-headed 12/4/2012 | 10:14:16 PM
re: 10-GigE Vendors Get Cold Feet ok scott... apologize to andrew for me.

LH
Iipoed 12/4/2012 | 10:14:17 PM
re: 10-GigE Vendors Get Cold Feet I showed my double ccie contact I was referring to the discussion and he pointed out that although he smells cisco kool-aid on your breath, yourpoints have a decent amount of validity. He was also upset that I even mentioned him and completely misquoted him, so I apologize for that.

Your comment: 1. The fact that FDRY has an 8gbpsbackplane means that they can send 8gbps of DATA across the backplane. The L2 header
information does not need to be sent across the backplane, however L2 header information does eat up space on the line, so line rate is completely possible.

His comment: Yes its an 8gbps backplane interconnect, it was designed as a gige switch and supporting 8x1gige was a valid design decision, it
obviously wasnt built to scale to line-rate 10ge. As far as line-rate being possible in certain environments (such as all min-sized packets
due to 28% of the data being stripped at the LC, sure, but are peoplegoing to buy a box where line rate is ONLY achievable with min-sized
packets? And that brings up the question of whether or not the network processor can keep up w/min-sized packet lookups at line-rate, doubtful.



Your comment: 2. The 6509 does not need to be put out to pasture like the Cat 5500. The 5500 had a 3.2Gb/s bus, where the 6509 has a 32Gb/s
bus. To argue that simply because a Bus is used, it is obsolete technology, the entire JNPR M series should be put to pasture.

His comment: The idea of bus architectures not scaling is not limited to a secret society of uber-elite router architects. Cisco realized this a
while back, Juniper realized it recently (see T-series). But depending on the requirements of the box, a bus might do just fine.



Your comment: 3. The 6509 is full extensible with a crossbar switch fabric, while still retaining investment protection for bus based cards.
It satisfies current customer needs, and it scales to where customers will be in the next 3-5 years.

His comment: Point taken, investment protection is a very valuable characteristic. Also, for your own good, consider a marketing position somewhere, you would shine.





Iipoed 12/4/2012 | 10:14:17 PM
re: 10-GigE Vendors Get Cold Feet light-headed, appreciate your response and totally agree. I however do take offence at being called a big mouth. I worked my butt off in the industry for 20 years until my retirement. I never worked for a more focused, dedicated and straight shooting company as Foundry.

They are upfront about their limitations, they have to be as 90% of their business is direct not via distributors that say anything to get the sale.

For you to publically name Andrew was also tasteless, sure he is one of the sharpest minds around, and I admit I also made a mistake in referrencing his background. I did attribute a comment to him that was not quite correct. My problem not his.
Iipoed 12/4/2012 | 10:14:18 PM
re: 10-GigE Vendors Get Cold Feet Each blade provides 32 gbs x 15 slots =480.
not quite sure where the 512 came from.
light-headed 12/4/2012 | 10:14:18 PM
re: 10-GigE Vendors Get Cold Feet Beo:

You shouldn't compare PoS running as a port on a router or switch with real SONET... PoS is usually only p2p and the only advantage is perhaps APS and more packets through one pipe than gig (with oc48 ports)... PoS can easily be replaced by 10-gig or even aggregated gig in most enterprise/fortune 1000 companies.

Big Difference in costs and capabilities here...

What makes you think that enterprises turn over equipment faster??? big fallacy... SPs have to upgrade/change the equipment every 1-2 years (now 3-5 maybe) to keep up with demand. I know enterprises that leave the equipment alone for 8-10 years! Some large well-known enterprises still have FDDI backbones...

I would not say there is a LOT of SONET in big organizations... there is SOME PoS... no SONET muxes, no SONET rings, No DCS or BDCS...

Real SONET and TDM infrastructure will not be replaced soon... PoS can be replaced very, very easily if desired...



light-headed 12/4/2012 | 10:14:18 PM
re: 10-GigE Vendors Get Cold Feet lipoed:

pretty sloppy to mention that your SE was andrew... you might as well post with your real name now. hmmmmmm you successfully IPOed with who... let me guess... FDRY! duh... another big mouth FDRY salesguy.

durtyphiber and the rest:

Crisco does as much as they have to and they always will. they are not about speeds and feeds. The xbar switch fabric will keep their ailing architecture going until they buy F10 or somebody for a box of peanuts...

FDRY is a decent box but they cannot do 10gig... do not give me this L2 header crap. They have to attach their own data to the packet when it goes anywhere. They are not even true wirespeed across the backplane with 8 gig like they claim. They always wrap one port locally when doing that testing. Do a true Full mesh and you will see that they are only wirespeed at around 80 byte packets... go ahead, prove me wrong!

that said, it doesn't matter because no one in the real world is going to see wirespeed 64 byte packets at 10 gig or 8 gig... and no one in the enterprise for sure which is where FDRY and F10 belong.
ntwkeng 12/4/2012 | 10:14:20 PM
re: 10-GigE Vendors Get Cold Feet Cisco does double the Gb/s, you are correct their. I have never seen them double the packet counts.

I would challenge your assertion that Foundry does not double or quadruple the bandwidth counters though.

Please help me understand how with 8Gb/s per slot they claim a 512Gb/s backplane on a 15 slot chassis? I would be very interested to see the math worked out for me.

M_Eisenberg 12/4/2012 | 10:14:20 PM
re: 10-GigE Vendors Get Cold Feet All,
One item that Maggie overlooked in reporting on this article was the fact that the IEEE 802.3ae and 10GEA scheduled 10 Gigabit Ethernet interoperability events prior to the Phoenix Trade Show (and associated performance test) which put very heavy demands on the same set of resources from most (if not all of the vendors). On top of that, a number of trade publications tried scheduling tests within the same time frame.

We vendors are often asked to participate in tests and in doing so, commit many resources (both internally and externally) to support the effort.

This is simply one situation where there were too many test events scheduled for exactly the same set of dates - if readers want to voice their dissatisfaction, those questions should be directed towards the individuals and editors who concurrently schedule similar events at the exact same time.
tasmanian 12/4/2012 | 10:14:21 PM
re: 10-GigE Vendors Get Cold Feet Forget the argument over ethernet vs. SONET for now.

10GBE is not selling yet due to the low cost of 1GBE ports. Customers are channelizing 1GBE and getting a much cheaper 10GBE solution than the current 10GBE port vendors can provide.

Perhaps a middle of the road GBE spec will come out before the price of 10GBE drops -- who knows?
tasmanian 12/4/2012 | 10:14:21 PM
re: 10-GigE Vendors Get Cold Feet Probably on Purnell's staff.
durtyphiber 12/4/2012 | 10:14:22 PM
re: 10-GigE Vendors Get Cold Feet I will grant that I do not know the details of the FDRY architecture. I generalized the backplane requirements. The backplane does not need to be 10gb/s to xfer 10Gb/s in the port.

Frame comes in, lookup done, Data portion (w/o header) switched to egress LC, new L2 header applied, frame switched out port. At this point, the avg frame size will determine how much of the 10Gb incoming stream is actually overhead, the smaller the frames, the more is overhead, reducing fabric bandwidth requirements. (Hence, GSR for instance can do 3 ports of GE at LR on 2.5Gb/s fabric interface)

I would love to hear exactly how my No.3 item is so far from the fact it does not deserve discussion. Your CCIE status does not concern me (got one to), nor is your employment at procket relevant to the discussion. (However, as I'm pretty sure I know who you are now, I retract my statment about ignorance.)

Simple fact is, the 6509 supports a full-xbar architecture via the SFM's and has distributed switching capabilites via the DFC's.

You also state "IT Managers", which is indicative of enterprise environments. How does the 6509 not scale to current needs and 3-5 years out? We are not discussing 40Gb/s interfaces here, we are talking GE aggregation, or high density FE aggregation, with GE or 10GE uplinks.

-dp
Iipoed 12/4/2012 | 10:14:22 PM
re: 10-GigE Vendors Get Cold Feet Sorry for the confusion, I am not a CCIE, my sourse to further validate my posts regarding architecture and my previous Sales engineer is a dual CCIE and is now at Procket.

This thread is becoming somewhat personal and really accomplishes little. There will always be Cisco bigots out there as there will be those looking at the new architectures to increase the performance of their networks.

Remember no one would ever leave AT&T for telephone service or IBM for computers. Cisco is going in the same direction by trying to be everything to everyone. They have long since stopped being an engineering driven company and marketing has taken over.

lets end this as the other posters I'm sure by now are gaining nothing relevant from continuing on.
Iipoed 12/4/2012 | 10:14:24 PM
re: 10-GigE Vendors Get Cold Feet meant to say "now a development engineer at procket",
Iipoed 12/4/2012 | 10:14:24 PM
re: 10-GigE Vendors Get Cold Feet Durtyphiber.

First of all each blade on Foundry's chassis connects to the backplan via 8gig. The backplane itself has, depending on the chassis config, 4 slot 8 slot or 15 slot 64 gig, 128 gig or 256 gig of throughput. This is a true non blocking wire speed architecture. The 6509 will always be a blocking box.
Your no. 3 item is so far from the fact that it does not deserve discussion. This from a dual CCIE that is not a development engineer at Procket.

Know what you are talking about before you make such ridiculous comments.
wilecoyote 12/4/2012 | 10:14:25 PM
re: 10-GigE Vendors Get Cold Feet Iipoe'd: I agree, the Cisco Fear Uncertainty and Doubt sales model is one of the three or four pillars of Cisco's success. Interestingly, many of Cisco's early managment team spent time at IBM which pretty much invented FUD for technology companies, and perfected it. But you know what? I/T people got really tired of IBM's FUD and started buying Amdahl mainframes, as an example, and IBM really stumbled. They re-invented themselves eventually but it was a painful transition. You don't see a tremendous amount of FUD anymore from IBM but Cisco still thinks they can get away with it indefinitely.

People know Cisco's technology sucks. They know there hasn't been anything new out of the place for many years. But it doesn't matter. There's too much critical mass there to fight it. Cisco is the Microsoft of communications.
durtyphiber 12/4/2012 | 10:14:25 PM
re: 10-GigE Vendors Get Cold Feet Your ignorance is amazing, and the fact that you feel the need to display it so flagrantly.

1. The fact that FDRY has an 8gb backplane means that they can send 8gb of DATA across the backplane. The L2 header information does not need to be sent across the backplane, however L2 header information does eat up space on the line, so line rate is completely possible.

2. The 6509 does not need to be put out to pasture like the Cat 5500. The 5500 had a 3.2 Gb/s bus, where the 6509 has a 32Gb/s bus. To argue that simply because a Bus is used, it is obsolete technology, the entire JNPR M series should be put to pasture.

3. The 6509 is full extensible with a crossbar switch fabric, while still retaining investment protection for bus based cards. It satisfies current customer needs, and it scales to where customers will be in the next 3-5 years.

-dp
Iipoed 12/4/2012 | 10:14:26 PM
re: 10-GigE Vendors Get Cold Feet The 6509 needs to be put out to pasture like the Cat 5500. The 6509 is still the same old Bus architecutre it always has been.But Cisco will milk their oldest technology as long as IT managers continue to think if they don't buy Cisco they might lose their jobs. When they should be thinking of what their organizations will be best for and require over the next 3 years.
seeker 12/4/2012 | 10:14:26 PM
re: 10-GigE Vendors Get Cold Feet Looks like Force10 will eat into Cisco's lunch. The 6509 is the money making machine.
Iipoed 12/4/2012 | 10:14:27 PM
re: 10-GigE Vendors Get Cold Feet Foundry's Switch/Router Router? (supports BGP, ISIS, MPLS, OSPF) 10g blade connects to the backplan via
8 gig. So obviously only 8gig throughput. Keep in mind the backplan is 4years old but it will still support the 10g module. Their new chassis, Mucho Grande (code name) will be right up their with Force 10 and most likely be able to support 2 10Gig ports per blade. As with all Foundry's chassis based products all ports are wired speed. Something that Cisco will never be able to accomplish with their current architecture.

As to features, the BigIron and NetIron ranks right up their with the current crop of high end competitors yet with the smallest footprint, highest density and by far the best price performance.
jepovic 12/4/2012 | 10:14:31 PM
re: 10-GigE Vendors Get Cold Feet Just a carrier's 2-cent perspective: It's just technology, so let's not get religious. SONET oc-192 has been here now for a while, it interops (which is pretty cool if you want to connect your router to the world), it delivers line rate and it has some nice features like APS if you like that stuff.

10 GE is still an emerging technology, but in a couple of years it will be mature and the business market will drive the prices down. If it is cheaper, I'll buy it, at least to connect customers, peers and access level equipment (I dunno about WAN links). That's how we use GE today. However, I think it might be that 10GE will drive down the prices of all 10G components, bringing the oc-192 prices with it. But hey, what do I care, I'm selling services not protocols.

But saying that SONET is dying - get outta here! We've been trying for a couple of years to get rid of ATM and SRP, and we're still not done. We're building PoS for WAN, mixed PoS and GE for MAN&LAN.

Why we won't buy any switches - they don't offer the same range of interfaces. The local loops are TDM! What am I suppose to tell the customer, "Oh, you have to wait until the ILEC delivers ethernet loops"?

Also, I'll tell you why Force10 or any startup will have a really hard time - we don't want more vendors! We worked our asses off to convince the engineers to learn JunOS, and we don't want to go through that again.

Price/Performance is just one parameter - services are more important. Can Force10 deliver MPLS-VPNs, multicast and DiffServ, together with the basic BGP and IS-IS routing? Do you have IPSec encryption modules? How about your SNMP? Can you keep up with the standardisation and development in all of these areas? If so, impressive!!
switch_this 12/4/2012 | 10:14:32 PM
re: 10-GigE Vendors Get Cold Feet
I can tell you the exact reason why every vendor except Force10 and Foundry did not bow out of this performance test. None of them can hold a candle to Force10's switch/router when it comes to performance, port density, and price. For a few examples, Cisco's CAT6500 drops packets in the 5-6Gig range, Foundry's NetIron switch (by their own admission) can only switch 8Gig per port. Cisco's gear is pathetic.

Read this, no one comes close to the performance the engineering team at Force10 has put together in their switch/router - yes, you read that correctly - I said "ROUTER" too! I'm predicting right here that Force10 will take business away from Cisco, Foundry, Extreme, Juniper, and the others if market conditions begin to improve. Force10 is the only company that can run 10GE at line rate, that also has the port density and features to go with it.

Yes, Force10 is a startup. Yes, there is a huge cloud of uncertainty in the telco/ISP/enterprise markets right now, but this startup has a legitimate shot at hitting a home run. There are only two things that could keep Force10 from making it: market conditions and self implosion.
rjmcmahon 12/4/2012 | 10:14:32 PM
re: 10-GigE Vendors Get Cold Feet Worse yet, the ops-cost vs. price ratio is worse when Service Providers need to deploy a NEW Service. Service Providers incure immense operational costs whenever they deploy a new technology, and it's a bigger hump for them to get over initially. Off the top of my head, let me list just a few the operational costs involved in deploying a new technology: training, QA, documentation, order entry system upgrades, provisioning system upgrades, billing system upgrades, purchasing new network management software (and integrating it into their NOC), working out their marketing strategy, and getting their sales force geared up to sell the service. And I've probably overlooked another dozen cost centers. Yuk!
_____________

This problem IBM faced as well. They couldn't retrain the majority of their workers to understand the new technologies.

Ops-costs seem most easily solved by building brand new companies using lower cost technologies. (Or possibly by massive consolidation, selling of divisions, and massive layoffs.)
beamer 12/4/2012 | 10:14:32 PM
re: 10-GigE Vendors Get Cold Feet Another one of those typical HarveyMudd (=bobbyMax) messages, which can be trashed.

BTW, heard that Cisco comes no where near to 4 gig on their 10gig blade, for some packet sizes and foundry for a well known fact cant switch 10G across a 8g plane. So this could be why those supposed to be big players are holding off!!!
Can someone shed any light about the other 10gig players??
rjmcmahon 12/4/2012 | 10:14:32 PM
re: 10-GigE Vendors Get Cold Feet But there's a whole lot of Fortune 500 companies and other organizations that use SONET to provide humungous amounts of bandwidth (such airports, commodity trading organizations, universities, big government agencies, and research institutions).
__________________

You probably know more than I do, so I'll take your word for it.

It seems likely the high bw/low cost interconnects will be the runs from wiring closets to the data centers. What is the expected SONET to 10GE ratio there? And how many such links are there? Enough to make a volume difference that matters?
rjmcmahon 12/4/2012 | 10:14:33 PM
re: 10-GigE Vendors Get Cold Feet So tell me again why SONET will go away? Especially when the big carriers have so much invested in their SONET infrastructure already.
_________________

Disclaimer: I have not enough information nor a working crystal ball and hence must take wild guesses at what will happen to the carrier industry.

With that disclaimer, the carrier industry seems to have enough potential that at least some few will emerge as the new leaders within 5-10 years. It's hard to imagine that these leaders will come from yesterday's laggards.

I think of it like IBM and all those PCs. Who would have predicted some UT undergraduate would emerge and take the world leadership position? He didn't sell mainframes nor even workstations -- rather commodity PCs which solved customer problems.
beowulf888 12/4/2012 | 10:14:33 PM
re: 10-GigE Vendors Get Cold Feet Taz wrote:
"So, feel me now, if everything in the line card except the port phy (not the backplane phy) is ethernet based, then adding that pesky SONET module drives the cost up and more importantly, is a pain in everyoneGs side -- SONET is close to death, sorry."

Well, I suspect that SONET as we know it will be around for another decade at least. Heck, there are still X.25 networks out there! -- so tell me again why SONET will go away? Especially when the big carriers have so much invested in their SONET infrastructure already.

cheers,
--Beo
willywilson 12/4/2012 | 10:14:33 PM
re: 10-GigE Vendors Get Cold Feet If you look at all of the elements that make up the cost of a 10GE or OC-192 system; the optics, chips, PCBs, sheet metal, labor, etc. how much would you say is common between the two? 10%? 50%? 90%? So the difference in cost of the two is determined by difference in cost of the non-common elements in one versus the other, right? Can you do the analysis and the math and show me how a 10GE switch will be SO MUCH cheaper than an OC-192 Sonet switch to manufacture?

-----------------

Here's a better way to win this argument: Portray SONET as "next generation" and maybe hire George Gilder to write some obfuscatory hagiography on the subject, to be presented at an "investment" conference on the new companies who have coincidentally kicked in $100,000 sponsorship fees. Hire some other publicity consultants to write press releases and place articles in the trade journals.

What? You mean that formula is yesterday's news? The horror! The horror!
beowulf888 12/4/2012 | 10:14:33 PM
re: 10-GigE Vendors Get Cold Feet rjmcmahon wrote:
"When running a business, it seems like volume is a common mechanism used to reduce expenses. In that regard will enterprise markets drive 10GE to significant volume and reduce the costs? It's hard to imagine enterprise markets would ever installing SONET, so no need to ask about that ;-)"

Are you kidding? Sure the big-SONET switches get sold to Service Providers. But there's a whole lot of Fortune 500 companies and other organizations that use SONET to provide humungous amounts of bandwidth (such airports, commodity trading organizations, universities, big government agencies, and research institutions). I'd hazard a guess that these days it's the PoS type of SONET predominates at this level -- since organizations such as those listed above turn over their equipment faster than Service Providers. No doubt GigE has made huge inroads into this market, as well. But there's still a lot of SONET in big organizations.

--Beo
strands555 12/4/2012 | 10:14:34 PM
re: 10-GigE Vendors Get Cold Feet re: "look at the
overall vendor cost (not price) to deliver the
port at 10GE from the backplane to the outside
fiber. Thats the number that matters."
=====
There are lots of different perspectives to determine 'the number that matters.' Below link is a paper by Peter Sevcik that addresses SONET, ATM, GbE costs, beyond just equipment and including operating costs (well worth a read):
http://www.netforecast.com/Art...
beowulf888 12/4/2012 | 10:14:34 PM
re: 10-GigE Vendors Get Cold Feet Iipoed, you asked:
"Why pay the premium for large carrier level switches when for 1/10 the cost you can set up a basic, no frills, not totally redundant and most likely not even NEBS compliant switch to collaspe your wiring closets via 1 gig to a 10Gig MAN."

Well, if you building a brand new MAN from scratch, you may be right. But in the real world that seldom happens (at least after the CLEC bubble burst ;-).

Even if 10GigE is 1/10 the cost of OC-192 (and I haven't yet seen any data on this thread that supports that particular price ratio), the actual cost of the hardware is less than third of the total cost of ownership.

Worse yet, the ops-cost vs. price ratio is worse when Service Providers need to deploy a NEW Service. Service Providers incure immense operational costs whenever they deploy a new technology, and it's a bigger hump for them to get over initially. Off the top of my head, let me list just a few the operational costs involved in deploying a new technology: training, QA, documentation, order entry system upgrades, provisioning system upgrades, billing system upgrades, purchasing new network management software (and integrating it into their NOC), working out their marketing strategy, and getting their sales force geared up to sell the service. And I've probably overlooked another dozen cost centers. Yuk!

Buying that same tired SONET gear doesn't look that unattractive when you start getting into the nitty-gritty details.

best regards,
--Beo

Lopez 12/4/2012 | 10:14:34 PM
re: 10-GigE Vendors Get Cold Feet The only Cisco product that comes close to Foundry's BigIron and NetIron products is the GSR12000 which is 3x the size and 3x as expensive and still on delivers48 gig of throughput versus 96 in the Foundry 8000.

A 12406 is approx the same size as a BigIron 8000 (quarter rack) and can hold 5 10GE cards, each processing 25Mpps, for a total of 125Mpps vs 96Mpps for the BigIron. Two 12404's could occupy the same space and provide 6 10GE ports, which is one less than the Foundry box. Density may be slightly lower, but performance is higher.

tasmanian 12/4/2012 | 10:14:35 PM
re: 10-GigE Vendors Get Cold Feet As long as the enterprise dictates the number of ports in market, the price of OC-192 ports will be artificially higher than 10GE.

Point being, the high volume of ethernet in the enterprise will keep the 10GE port COST much lower than OC-192.

Eventually, purchasing SONET modules in small amounts for the most stubborn, legacy system customers, then converting to ethernet packets before they hit the first asic will be the norm -- ala the ONS 15454, there are packets inside that chassis, not frames

So, feel me now, if everything in the line card except the port phy (not the backplane phy) is ethernet based, then adding that pesky SONET module drives the cost up and more importantly, is a pain in everyoneGs side -- SONET is close to death, sorry.

rjmcmahon 12/4/2012 | 10:14:36 PM
re: 10-GigE Vendors Get Cold Feet There will be no substantial cost difference between 10GbE and OC-192 as long as they run the same or similar shipping volume range.
______________

When running a business, it seems like volume is a common mechanism used to reduce expenses. In that regard will enterprise markets drive 10GE to significant volume and reduce the costs? It's hard to imagine enterprise markets would ever installing SONET, so no need to ask about that ;-)
Iipoed 12/4/2012 | 10:14:36 PM
re: 10-GigE Vendors Get Cold Feet 10 Gig is being built for the MAN. The primary need to address is taking multiple gigabit ethernet ports and intelligently using fiber by pushing everything through a 10 gig pipe. Why pay the premium for large carrier level switches when for 1/10 the cost you can set up a basic, no frills, not totally redundant and most likely not even NEBS compliant switch to collaspe your wiring closets via 1 gig to a 10Gig MAN.

Is it just me or do all the engineers out there make things to complicated when all they needed to do was to ask for a salesperson's input.
KayakSport 12/4/2012 | 10:14:36 PM
re: 10-GigE Vendors Get Cold Feet wdog,
you are probably right about this. there will be no substantial cost difference between 10GbE and OC-192 as long as they run the same or similar shipping volume range. 1GbE is a different story, though. Its volume has driven down the entire system cost and some carriers love to use the technology in the Metro access because of this lower per-bit cost.
BobbyMax 12/4/2012 | 10:14:37 PM
re: 10-GigE Vendors Get Cold Feet It has been known to a lot of people in the industry that 10 Gigabit Ethernet Technology is not suitable for the carrier environment. Fortunately it does not take Albert Einstein to figure this out.

When a new technology is announced, the serial enterpreaners from California run to VCs. These together along with their VCs can say anything that comes to their mind.

Cisco, for example, has never participated in any testing. It is always scared of getting exposed as for performance and reliability is concerned.

I must say in all fairness that the lab testing does not provide a realistic setting for the testing of 10 Gigabit Ethernet products. To expose the weaknesses of these products they need to be tested in the real RBOC environment.

Most of the time SONET/SDH equipment has to work in a meshed network environment. It is not very clear that ethernet switches can work in this enviroment.

Also none of these vendors, for example. Force 10 have their products put to OSMINE test. Most of the switch vendors do not good switching fabrics and queueing mechanism, yet they have very loud mouth to speak abour their products in a conference. They also half right and half wrong technology papers. Force 10 has amassed about $160 Million. One wonders what do these start-up tell the VCs.

It is hard to determine why the California VCs and start-up engage in behavior that reflects dishonesty.

These Gigabit Ethernet vendors want to sell their products first before being tested. These vendors also think many important system chracterstics are not important. They play the game of ASICs, a brick level technology. ASICs can be used as a selling point of these features.

wdog 12/4/2012 | 10:14:37 PM
re: 10-GigE Vendors Get Cold Feet "There is no inherent reason that 10G ethernet is/or will be cheaper than Sonet. The optics are the same and after that it's just one silicon chip versus another silicon chip of roughly the same size and complexity."

Heh, you believe the chips are nearly identical? We are talking an asynchronous vs a synchronous interface with completely different precision requirements for things such as jitter. Go look up the 10GE WAN PHY spec and see if you can figure out why it was created. Ill give you a hint, it has to do with economical reasons. Your post proves that you are a fool. Please refrain from blasting these message boards with your incompetent babblings if you are simply a spectator to this sport.

You have been clownstamped.

-------------
Well I seemed to have really pushed your button Clownhammer. Probably something that isnGt very hard to do, so IGll do it again. Since you are obviously much smarter than I am maybe you can explain something to me. If you look at all of the elements that make up the cost of a 10GE or OC-192 system; the optics, chips, PCBs, sheet metal, labor, etc. how much would you say is common between the two? 10%? 50%? 90%? So the difference in cost of the two is determined by difference in cost of the non-common elements in one versus the other, right? Can you do the analysis and the math and show me how a 10GE switch will be SO MUCH cheaper than an OC-192 Sonet switch to manufacture? IGm really looking forward to your analysis.
Iipoed 12/4/2012 | 10:14:38 PM
re: 10-GigE Vendors Get Cold Feet Yeah is does make sense to buy cisco LAN switches for one fifth the throughput, old bus, blocking architecture. But they do stress no forklift upgrade. Try any of Cisco new modules into a one year old Cat 6500, yeah they plug in but to get them to work requires new power supplies and new supervisor modules and still you get 1/3 the performance. Remember Cisco count packet throughput going in and going out as two packets. Foundry only countes one packet and even with Cisco's crazy math they come out less than 1/2 the through put performance.

The only Cisco product that comes close to Foundry's BigIron and NetIron products is the GSR12000 which is 3x the size and 3x as expensive and still on delivers48 gig of throughput versus 96 in the Foundry 8000.

As to 8gig per module this is true. BigIron/NetIron provide an 8 gig connection to the backplane but on all their chassis products which are up to 4 years old. Plug a 10 Gig module into a Cat 6500 and if you are lucky you might see 3 gig. On a 12000 maybe 6 gig.

This information you can get from Foundry's web site.
skeptic 12/4/2012 | 10:14:38 PM
re: 10-GigE Vendors Get Cold Feet Since you're probably not an engineer or smart enough to interpret an IEEE spec, your next exercise in idiocy is to simply research the COSTS of 10GE and OC-192 modules... not ports... modules.
---------

And if your going to look at costs, look at the
overall vendor cost (not price) to deliver the
port at 10GE from the backplane to the outside
fiber. Thats the number that matters.

The differences between 10GE and OC-192 are
not as significant as you think.
ntwkeng 12/4/2012 | 10:14:39 PM
re: 10-GigE Vendors Get Cold Feet I would never compare the 15454 and the Foundry BigIron. They are completely different products.

The 15454 is an optical transport and SONET grooming device for large enterprise and service provider networks.

The Foundry BigIron is a chassis-based modular Ethernet switch for campus LAN deployments.

The Foundry delivers 8gb/s per slot (yet somehow has a true 10GbE???)

Okay- funny story. At the N+I show this year Foundry was showing offits 10GbE at 'wire rate' IT had all the traffic from its IXIA going into and BACK OUT OF THE SAME INTERFACE. So its local switching capability is good, but once you carry it across the backplane you will experience loss because of the oversubscription.

If you wanted to compare products of a like flavor between Cisco and Foundry I would use Cisco's Catalyst series of LAN switches against Foundry's BigIron/NetIron/FAstIron offerings.
tasmanian 12/4/2012 | 10:14:39 PM
re: 10-GigE Vendors Get Cold Feet "SONET is less expensive, more reliable and provides bandwidth guarantees."

OK, Stu. You're on the idiot stick again.

Since you're probably not an engineer or smart enough to interpret an IEEE spec, your next exercise in idiocy is to simply research the COSTS of 10GE and OC-192 modules... not ports... modules. While you're at it, check out costs of OC-48, OC-12 and 1GE modules (fiber and copper, different reaches).

Then come back to this forum and post your findings.
GW Pearson 12/4/2012 | 10:14:40 PM
re: 10-GigE Vendors Get Cold Feet The whole point of making the 10GigE PHY standard was so that SONET/SDH and 10GigE would be compatible. Thus, you can have both running on the same network. I don't think it's an either or question for carriers. You'd like to be able to carry both to serve the broadest customer base possible. It's not a religious issue.
clownhammer 12/4/2012 | 10:14:40 PM
re: 10-GigE Vendors Get Cold Feet "There is no inherent reason that 10G ethernet is/or will be cheaper than Sonet. The optics are the same and after that it's just one silicon chip versus another silicon chip of roughly the same size and complexity."

Heh, you believe the chips are nearly identical? We are talking an asynchronous vs a synchronous interface with completely different precision requirements for things such as jitter. Go look up the 10GE WAN PHY spec and see if you can figure out why it was created. Ill give you a hint, it has to do with economical reasons. Your post proves that you are a fool. Please refrain from blasting these message boards with your incompetent babblings if you are simply a spectator to this sport.

You have been clownstamped.
analogworm 12/4/2012 | 10:14:40 PM
re: 10-GigE Vendors Get Cold Feet Hey,
Sonet: Category II jitter
Jitter generation no greater than 0.1 UIp-p in 50kHz to 80MHz band
Jitter tolerance: it's an input jitter vs. frequency plot where the device should tolerate large jitter at low frequencies (15 UIpp between 10 and 2000 Hz) and low jitter at high frequencies (0.15 UIp-p above 4 MHz).
[A UI is a unit interval or 1 bit time]
Jitter Transfer: Transfer no more than 0.1 dB of the input jitter to output jitter from 120 kHz and down in frequency.

10 Gig Ethernet is just different. One must construct the "stressed eye" test, where the effect of jitter is incorporated into "stressed eye." In a sense 10 Gigabit Ethernet lumps receiver sensitivity tests and jitter tolerance together and adds a dash of offset modulation.
Jitter generation is comparatively measured with the "Transmitter and Dispersion Penalty Test." First, one forms the stressed eye with a very good reference Tx and rams it into a reference receiver. Then one takes the DUT, connects 10 km of fiber, and rams that signal into the reference receiver at different attenuation settings. The comparison of those two sensitivity results is the "Transmitter and Dispersion Penalty" which in a sense tests jitter generation.

Cheers,
aw
Marguerite Reardon 12/4/2012 | 10:14:41 PM
re: 10-GigE Vendors Get Cold Feet The story has been updated to reflect some of this new information. Please check it out. I appreciate your feedback.
CRC_Check 12/4/2012 | 10:14:42 PM
re: 10-GigE Vendors Get Cold Feet ""Is the jitter spec difference making a signficant difference in the overall cost of the port? ""

My view is that in the SONET world you really have GR253 transport at 15xx nm and GR253 access and CO interconnect at 13xx nm.

Good jitter specs were originally needed for transport with all those long spans of non reclocking repeaters. Today good jitter specs are needed for LH DWDM transport for all those high power non reclocking optical repeaters.

In the access and interconnect case at 13xx nm I can't really see the point of keeping such stringent specs, and most datacom and LAN gear is really just GR253 compatible meeting a subset of the specs. 13xx nm won't interoperate directly with 15xx DWDM optical amps and I don't think anybody still really runs 15 spans of nonregenerated SONET with the kinds of long reach technology we have today.

So where does this leave 10Ge? In the WAN case, my view is if it has 15xx nm or DWDM laser its a carrier transport port and it better measure up.
For the LAN case or for 13xx nm SR or IR, there full jitter conformance is probably a waste of money when you look at practical the network topology that results.

my 2 cents.
stuartb 12/4/2012 | 10:14:43 PM
re: 10-GigE Vendors Get Cold Feet "What are you smoking, Stu?"

I'm smoking some primo stuff man, it's called reality, and I think you need a hit. In fact, take another.

There's not a single carrier out there who will choose to use Ethernet as their transport technology in the near future. SONET is less expensive, more reliable and provides bandwidth guarantees.

-Stu
Iipoed 12/4/2012 | 10:14:44 PM
re: 10-GigE Vendors Get Cold Feet Foundry 4 slot 10 Gig chassis is 8.75" high by 17.75" wide, 60lbs fully loaded. What about 16 gig ports from redundant management modules, plus 24 10/100 providing numberous configurations to support pushing traffic out the 10 gig port. Oh year full BGP4, MPLS, ISIS and true wirespeed. No way Cisco can come close in price or performance. And for only 5k more you can have 4 additional ports.

These chassis have been 10 Gig ready since 1998 so all the legacy customers need only purchase the 10g blade.

This is all off of Foundry's web page.
skeptic 12/4/2012 | 10:14:44 PM
re: 10-GigE Vendors Get Cold Feet "The optics are the same and after that it's just one silicon chip versus another silicon chip of roughly the same size and complexity."

Is the jitter spec the same? Not even close.
--------------
Is the jitter spec difference making a signficant
difference in the overall cost of the port? Not
in the parts that I've seen.
beowulf888 12/4/2012 | 10:14:44 PM
re: 10-GigE Vendors Get Cold Feet Hi, taz:
Could you provide a URL or two that would compare the jitter specs of OC-192 vs 10GigE?

And another point that will support your position -- even if GigE is substantially cheaper than OC-192 (and I'm not fully persuaded that it is), the big carriers just LOVE that APS that SONET offers. They'll pay a premium for the warm fuzzies that APS gives 'em.

I'm not saying 10GigE won't find a place in carrier networks -- but I don't think it will kick SONET's butt, at least not yet.

--Beo
beowulf888 12/4/2012 | 10:14:45 PM
re: 10-GigE Vendors Get Cold Feet Iipoed asks:
"What does a 15454 platform run?"

Answer: I haven't worked with the 15454 in a couple of years, so I can't give you current figures. But Cisco's list price is always reduced very very substantially to its best customers. I'd be willing to bet an Cisco account team would match Foundry's pricing if it would make or break a multi-million^10 sale. Also, last I heard, the ONS 15454 had one of the smallest footprints of any metro optical switch (feel free to correct me if my info is out of date!).

--Beo
tasmanian 12/4/2012 | 10:14:45 PM
re: 10-GigE Vendors Get Cold Feet "The optics are the same and after that it's just one silicon chip versus another silicon chip of roughly the same size and complexity."

Is the jitter spec the same? Not even close.
tasmanian 12/4/2012 | 10:14:46 PM
re: 10-GigE Vendors Get Cold Feet <in 10gbe="" are="" bet="" cheaper="" fact,="" i'll="" of="" on="" one="" ports="" systems.="" than="" the="" these="" they="">

What are you smoking, Stu?

First, 10GBE will dominate on switching systems.

Second, the cost per port of 10GBE IS one-fifth the cost of OC-192.

Third, the cost delta is due to jitter specs.</in>
Iipoed 12/4/2012 | 10:14:48 PM
re: 10-GigE Vendors Get Cold Feet Foundry's 10 gig blade fits in a 4 slot NetIron Chassis which costs about $15,000. including redundant power supplies and an 8 gig management module. Can migrate to an 8 slot or 15 slot chassis, supports 1 gig, 10/100 ports, Oc 12,48 in various combinations and with the normal industry discounting of 25% it is a steal. Oh by the way you also might want to compare the footprints, weight, power requirements. Also remember these chassis have been shipping for 3+ years and do not require a forklift upgrade to support the 10 G module
What does a 15454 platform run?


Check with Univ. of So. Calif Information Sciences Institute as to their thought on Foundry's 10 Gig (not bundled 1gig, true 10) solution.

Oh one last thing the 1 port 10 Gig blade's list price is around 55k.
wdog 12/4/2012 | 10:14:49 PM
re: 10-GigE Vendors Get Cold Feet "An OC192c Sonet interface, which also runs at 10 Gbit/s, costs approximately $300,000 per port."


$300K is no where close, not even on a router. Somebody is feeding your a line of marketing hype. There is no inherent reason that 10G ethernet is/or will be cheaper than Sonet. The optics are the same and after that it's just one silicon chip versus another silicon chip of roughly the same size and complexity. In fact, a Sonet TDM switch is very likely to be much cheaper per port than an Ethernet switch because much less buffer memory is needed. You can't over subscribe a Sonet port like and ethernet port. The rule of thumb is; circuit switches are the cheapest, next are layer two packet/cell switches, and last are layer 3 routers. Look at the cost of a 10G port on a Ciena Coredirector, versus a 1OG ethernet switch, versus a 10G anything port on a router and it is pretty clear.
stuartb 12/4/2012 | 10:14:50 PM
re: 10-GigE Vendors Get Cold Feet Cost of Foundry's recently announced 1-port 10GbE card (LR optics) for their NetIron platform: $84,995

Cost of Ciscos's recently released 1-port 0C-192 card (LR optics) for their ONS 15454 platform: $60,000

Showing that 10Gbe technology lacks any true value proposition for service providers with an actual cost comparison, priceless.

-Stu

stuartb 12/4/2012 | 10:14:51 PM
re: 10-GigE Vendors Get Cold Feet "An OC192c Sonet interface, which also runs at 10 Gbit/s, costs approximately $300,000 per port."

Margaret, you need to research the SONET cost figures given to you by the Ethernet vendors much more thoroughly. NG-SONET systems have driven the price point of an OC-192c per port to below 50k. In fact, I'll bet they are cheaper than the 10Gbe ports on one of these systems.

-Stu
HOME
Sign In
SEARCH
CLOSE
MORE
CLOSE