re: Who Makes What: GPON EquipmentTherein lies the crux. It's hard to imagine anyone making money selling GPON products right now. The market is a shallow revenue pool with too many swimmers. It's like sperm attacking the egg....
re: "Can you elaborate on the comment on "oversized, bloated designs" of ONT ?? what part is unnecessary, eg., triplexer is not needed etc., ??" --
I can address your question on two levels:
- sheer tonnage - features and functionality
Try to imagine the standard customer premises versions of ONTs and ONUs being situated behind the desks of 8000 knowledge users in an office building. They won't fit.
From the standpoint of features and funcionality, the basic enterprise worker requires a transparent connection to the data center, devoid of triplexers (video is satisfied by IPTV), or battery backup provisions (which is also satisified locally through other means), nor do they need to be housed in ruggedized enclosures befitting a Tempest endurance test.
At the same time, it can be argued that certain features now available through Category 5/6 UTP, e.g., Power over Ethernet, are lacking, although I believe that this topic has now become the tail that wags the dog in many respects.
In short, the ONT needs to fit into the standard form factors of NICs and today's UTP and Wireless PC Cards, and priced accordingly, if not also made integral to PCs and other work-area appliances as well.
Scott's view is situational, I believe, and especially apropos of residential solutions. WDM-PONs in the enterprise and institutional networks may follow a significantly different path. And of course, service providers themselves will find uses for it in the delivery of their routine services.
The cost-benefit tradeoffs differ once you move from residential to business, since businesses must take into account additional external factors such as power consumption, real estate (space requirements) and cooling.
Prior to WDM-PONs being leveraged to address these external issues, however, I'd agree that several lower-ordered considerations must be addressed first, not the least being the need for acceptable price points for desktop-level optical (0-E) interfacing. Here we go again, though, w.r.t. chicken & egg economics.
That being said, I'll go out on a limb and state that the situation may in fact be changing at the very moment I am typing this, primarily due to the heightened state of exigency and awareness on the part of managers of the need to radically cut back on carbon emissions and the escalating (in some cases, near-debilitating) costs associated with energy consumption.
re: Who Makes What: GPON EquipmentI should add that, this is not a trivial matter of one-for-one substitution, or a matter that is handled with a drop-ship approach.
Once you've gone from a hierarchically-switched/routed unshielded twisted pair (Cat5/6 UTP) in-building environment to one that is based on a collapsed optical switch (which could include OLTs), virtually every aspect of network administration and configuration, and even tuning, changes. This is because hundreds of billions of dollars worth of embedded provisions in most areas of enterprise networking still depend on the primary and secondary attributes of the status quo, and this is mainly due to overcoming the distance constraint of UTP.
One cannot claim that the optical solutions singularly makes sense for all situations, since many potential benefits can also be attained now through the hybridization of fiber and a growing list of new and equally-disruptive wireless solutions as well.
re: Who Makes What: GPON Equipmentre: "Try to imagine the standard customer premises versions of ONTs and ONUs being situated behind the desks of 8000 knowledge users in an office building. They won't fit."
This is a new application of G-PON that I've never heard of. I can only imagine all the uses of an asymmetrical 2.5 & 1.5Gbit/s optical link to each office worker. Office productivity will soar. Bits will be flying around the office at rates previously reserved for the NSA computers.
Apparently now with W-PON, the bits will fly even faster and take less desk space! I can't wait to discuss this new technology application with IT for inclusion in next year's budget.
re: Who Makes What: GPON EquipmentHas anybody considered the operational limitation in GPON due to the fact that if one ONT fails all the 32 other ONT are out of service due to lack og syncronization. This happens in practice.
re: Who Makes What: GPON Equipment First off on Novera...a quote from their website:
"To address these future requirements, Novera has developed a WDM-PON FTTH solution that supports a dedicated and symmetric 100 Mb/s connection to each home at an affordable cost."
Now, that is their discussion of their product. You don't believe me look it up. They use resonant cavity lasers to create this (and they are custom and Novera only). This is their limiting factor at the moment.
Most other WDM-PON schemes are actually 1/2 duplex and based around Reflective SOAs. Before everybody gets all excited - go price one out. Note current indoor single modem EPON units are going for about $75 in volume and then lets talk about when these are going to get built for $20 (about the cost of an EPON triplexer). If you think that is happening anytime soon, then we can start talking about sourcing all these wavelengths at low costs at the OLT.
Notes on EPON in NTT. NTT is NOT allowed to do video by the regulators. That is why they don't do it. There is not a choice in the matter today, but may be in the future. Verizon ONTs (the other volume user) have triplexers and are mounted outside (note this allows them to be capitalized and not expensed). They also support direct POTS interfaces, which the NTT ones do not (well the $75 ones don't). When GPON based diplexer ONTs are built (and they do exist), they reach cost approximation with EPON. The optics is a bit more expensive, but they offer more bandwidth to the PON so can be deployed at a higher split ratio (if desired).
re: Who Makes What: GPON EquipmentThere are eo physical limitations to the data rate of the "single color" sources in WDM PON schemes. Costs will be very attractive hoever, and ONx/CPE size is a standard 1-in. GBIC or 1/2in. wide transceiver (SFP or pin-though-hole). It's pretty nice and likely to pop in the nimble markets everywhere that can easily reference what Korea has done well.
re: Who Makes What: GPON EquipmentWe seem to have veered off of several different tracks during the progress of this thread. The first notable exception came with the introduction of WDM-PON, then leading to a discussion of GPON's vs. EPON's attributes, and how those fit into the enterprise space. I'd like to pursue this last point for a while, if my learned colleagues here will permit.
On the surface, at least, one might get the general impression that using GPON in the enterprise would be tantamount to deploying ATM (actually Ethernet over ATM) to the desk, while others may perceive the use of EPON in quite the opposite fashion, i.e., exposing end users to the perils of collision-like domains. In actuality, neither supports a collision domain in the classical sense, nor is either of these entirely deterministic in the sense of being fully-isochronous, although that possibility, or an approximation thereof, has certainly been entertained and possibly included as an option by some vendors, through the use encapsulation or pseudowire-like techniques.
So, to state the question in plain terms, which of these two frms of PON, i.e., GPON or EPON, is best suited as a FTTD(esktop) solution, and why?
On the matter of WDM-PONs, in the early going, at least, it is not a desk-level solution in my estimation, except for the most voraciously demanding power-user's needs, and even then it assumes that a product is used that exceeds 100 Mbps. It will more likely be used in enterprise- networks as a substitute for existing backbones and arterial links between buildings, and/or very likely for shunting aggregation and distribution switches to data centers. I also think it will find a home with ILECs and independent fiber operators, alike, for the purpose of tying multiple customer locations together in virtual campus configurations.
I think we'll come to see WDM-PONs ascendence in the enterprise and in carrier's metro services for businesses becoming apparent a lot sooner than in the residence, since the economic drivers for these segments are significantly different to warrant the higher premium on the enterprise side.
For residential purposes, WDM-PON will very likely serve in upgrades at first, aggregating multiple, lower-order FTTP/N/x deployments over one or two fiber strands, where some of those lower-order networks themselves will be single wavelength-based xPONs and/or pt-2-pt switched Ethernets as well.
re: "Can you elaborate on the comment on "oversized, bloated designs" of ONT ?? what part is unnecessary, eg., triplexer is not needed etc., ??"
--
I can address your question on two levels:
- sheer tonnage
- features and functionality
Try to imagine the standard customer premises versions of ONTs and ONUs being situated behind the desks of 8000 knowledge users in an office building. They won't fit.
From the standpoint of features and funcionality, the basic enterprise worker requires a transparent connection to the data center, devoid of triplexers (video is satisfied by IPTV), or battery backup provisions (which is also satisified locally through other means), nor do they need to be housed in ruggedized enclosures befitting a Tempest endurance test.
At the same time, it can be argued that certain features now available through Category 5/6 UTP, e.g., Power over Ethernet, are lacking, although I believe that this topic has now become the tail that wags the dog in many respects.
In short, the ONT needs to fit into the standard form factors of NICs and today's UTP and Wireless PC Cards, and priced accordingly, if not also made integral to PCs and other work-area appliances as well.
Posted By: brookseven
You guys do know the Novera system is less bandwidth than a GPON right?
seven
...
And my minivan is faster than a Lamborghini :
http://tinyurl.com/5lwps7
http://www.pennwellblogs.com/l...
FAC said...
Scott's view is situational, I believe, and especially apropos of residential solutions. WDM-PONs in the enterprise and institutional networks may follow a significantly different path. And of course, service providers themselves will find uses for it in the delivery of their routine services.
The cost-benefit tradeoffs differ once you move from residential to business, since businesses must take into account additional external factors such as power consumption, real estate (space requirements) and cooling.
Prior to WDM-PONs being leveraged to address these external issues, however, I'd agree that several lower-ordered considerations must be addressed first, not the least being the need for acceptable price points for desktop-level optical (0-E) interfacing. Here we go again, though, w.r.t. chicken & egg economics.
That being said, I'll go out on a limb and state that the situation may in fact be changing at the very moment I am typing this, primarily due to the heightened state of exigency and awareness on the part of managers of the need to radically cut back on carbon emissions and the escalating (in some cases, near-debilitating) costs associated with energy consumption.
Once you've gone from a hierarchically-switched/routed unshielded twisted pair (Cat5/6 UTP) in-building environment to one that is based on a collapsed optical switch (which could include OLTs), virtually every aspect of network administration and configuration, and even tuning, changes. This is because hundreds of billions of dollars worth of embedded provisions in most areas of enterprise networking still depend on the primary and secondary attributes of the status quo, and this is mainly due to overcoming the distance constraint of UTP.
One cannot claim that the optical solutions singularly makes sense for all situations, since many potential benefits can also be attained now through the hybridization of fiber and a growing list of new and equally-disruptive wireless solutions as well.
This is a new application of G-PON that I've never heard of. I can only imagine all the uses of an asymmetrical 2.5 & 1.5Gbit/s optical link to each office worker. Office productivity will soar. Bits will be flying around the office at rates previously reserved for the NSA computers.
Apparently now with W-PON, the bits will fly even faster and take less desk space! I can't wait to discuss this new technology application with IT for inclusion in next year's budget.
Thanks for the information!
First off on Novera...a quote from their website:
"To address these future requirements, Novera has developed a WDM-PON FTTH solution that supports a dedicated and symmetric 100 Mb/s connection to each home at an affordable cost."
Now, that is their discussion of their product. You don't believe me look it up. They use resonant cavity lasers to create this (and they are custom and Novera only). This is their limiting factor at the moment.
Most other WDM-PON schemes are actually 1/2 duplex and based around Reflective SOAs. Before everybody gets all excited - go price one out. Note current indoor single modem EPON units are going for about $75 in volume and then lets talk about when these are going to get built for $20 (about the cost of an EPON triplexer). If you think that is happening anytime soon, then we can start talking about sourcing all these wavelengths at low costs at the OLT.
Notes on EPON in NTT. NTT is NOT allowed to do video by the regulators. That is why they don't do it. There is not a choice in the matter today, but may be in the future. Verizon ONTs (the other volume user) have triplexers and are mounted outside (note this allows them to be capitalized and not expensed). They also support direct POTS interfaces, which the NTT ones do not (well the $75 ones don't). When GPON based diplexer ONTs are built (and they do exist), they reach cost approximation with EPON. The optics is a bit more expensive, but they offer more bandwidth to the PON so can be deployed at a higher split ratio (if desired).
seven
On the surface, at least, one might get the general impression that using GPON in the enterprise would be tantamount to deploying ATM (actually Ethernet over ATM) to the desk, while others may perceive the use of EPON in quite the opposite fashion, i.e., exposing end users to the perils of collision-like domains. In actuality, neither supports a collision domain in the classical sense, nor is either of these entirely deterministic in the sense of being fully-isochronous, although that possibility, or an approximation thereof, has certainly been entertained and possibly included as an option by some vendors, through the use encapsulation or pseudowire-like techniques.
So, to state the question in plain terms, which of these two frms of PON, i.e., GPON or EPON, is best suited as a FTTD(esktop) solution, and why?
On the matter of WDM-PONs, in the early going, at least, it is not a desk-level solution in my estimation, except for the most voraciously demanding power-user's needs, and even then it assumes that a product is used that exceeds 100 Mbps. It will more likely be used in enterprise- networks as a substitute for existing backbones and arterial links between buildings, and/or very likely for shunting aggregation and distribution switches to data centers. I also think it will find a home with ILECs and independent fiber operators, alike, for the purpose of tying multiple customer locations together in virtual campus configurations.
I think we'll come to see WDM-PONs ascendence in the enterprise and in carrier's metro services for businesses becoming apparent a lot sooner than in the residence, since the economic drivers for these segments are significantly different to warrant the higher premium on the enterprise side.
For residential purposes, WDM-PON will very likely serve in upgrades at first, aggregating multiple, lower-order FTTP/N/x deployments over one or two fiber strands, where some of those lower-order networks themselves will be single wavelength-based xPONs and/or pt-2-pt switched Ethernets as well.