Packet Pushers: All About Optics

Packet Pushers: All About Optics

Listen in as Tim Dixon opines about transceivers, the Shannon limit and the future of fiber optics in the data center with Ethan Banks and Greg Ferro, hosts of The Packet Pushers Weekly Show.

Audio Transcript

Ethan Banks: Today in the Packet Pushers weekly show, optics. You know those transceivers you buy from the same vendor who made the switch because someone told you if you buy non-branded vendor optics that God will kill a kitten?

What if you had a legitimate option for optics where you didn’t have to buy the overpriced vendor transceivers and then God leaves those kittens alone?

And more importantly that that tech doesn’t give you a hard time when you call in for support. Our sponsor today, InterOptic, loves the kittens very much and they believe they have some optics you’ll want to buy. Joining us today is Tim Dixon, CEO at InterOptic. Tim, InterOptic, you guys have been around quite a while – 12 years I believe. Would you give us a little background on the company back in 2005 when you started? What was going on that drove you to start this company?

Tim Dixon: Sure, we saw the same dichotomy that you have in terms of pricing, what the OEM is doing to that end user, that end customer. We saw that because our background is in telecom and years ago, in the mid-eighties we started working with the Finisars of the world, the HPs of the world helping them spec optics to IEEE standards and really knew what the cost structure was and still know a lot of those folks with that big gap there, we’ve focused on transceivers and said.. “Hey look, you’re not getting the value you should, and we can give you that value.” So, we worked on supply arrangements and got that going back in 2005. So, it’s a pretty simple business then, as we have transitioned it’s becoming a little more complex but we can talk about that.

Ethan Banks: So, is that the core of your business then, you guys are selling transceivers, ethernet transceivers specifically or is there a broader range in what you do?

Tim Dixon: Yeah, so we focus on the data center, so, what is consumed in the data center. We call it data interconnect. We have gone broader than transceivers at this point because we see what is happening in the data center. We can talk about that a little bit, but

layer zero is becoming more and more important.

You think wow that’s pretty boring stuff but at the end of the day it isn’t and we can talk about why.

Ethan Banks: From one standpoint I can say I am a network engineer buying a switch and I gotta throw transceivers in the thing so I can plug in my cables and get everything connected up. You’re saying this term data interconnect, so is there more to what that means to you than just you’re selling transceivers?

Tim Dixon: Let’s talk about that a little bit. So, the statement you made is exactly what the OEMs would like that statement to be in the minds of their customers. “Hey don’t worry about it, it’s just light, it’s just throwing the cables, it’s just like a cable, just put it in there.” The fact of the matter is, we see a lot of performance issues, a lot of range of specmanship that are going on in the transceiver world today and frankly, the volume of what is being deployed on the optic side is much greater and it continues to grow at a geometric pace.

You are really in a position where that could be a ticking time bomb if you are not doing the right type of optical qualification,

…the right type of optical specmanship, being able to have the highest quality supply chain in place, doing the logistics that you need to do, really focus on it as an expertise. That’s where we sit in terms of them… our customers needing to never, ever, have issues with anything at the layer zero. They want to forget about it, but they can’t forget about it if they are not buying the highest quality product.

Greg Ferro: One of the things vendors say is. “You have to buy our optics or they’re not supported.” So what you’re sort of saying is that your business stands behind the optics that you sell.

Tim Dixon: Yes, absolutely! So, not only do they get piece parts from us, they get full engineering that back ends both the.. we can help them with the optic architecture, we can help them with the differing interfaces because we actually sell more transceiver SKUs for Cisco equipment than Cisco does themselves and that goes the same way for of all the OEMs. We support such a broad range of platforms, so there might be some certain applications and in fact, we’ve gone through and designed some specific applications for some of our customers.

Ethan Banks: I want to just pick up on something that maybe a lot of engineers don’t understand is that SFP modules are actually computerized in their own right and what I mean by that is they have code and processes in them that need to be maintained and updated.

Tim Dixon: That is correct. You hit it on the head. Our background is building large switching systems deployed in Verizon, AT&T, at one time 80% of the traffic in the world went across one of the platforms we designed back in Tellabs. So high availability, large switching systems, embedded software. That is our background. We focus on how do those transceivers fit – not just, hey it’s a widget, or hey it’s a chip. A lot of people might know it kind of from the chip level. A lot of our competition are just distributors trying to skim a little bit of money off the transaction. But in fact, our background allows us to not only know the optical architecture, the exact application it is going into, but also, be able to black box reverse engineer what’s happening inside that Nexus 9000 to make sure that our transceivers never skip a beat.

Ethan Banks: So, you are making the distinction there. You’re not just a reseller for something that has been manufactured, you stick a label on it and push it out the door.

You are actually involved in the engineering of the modules themselves


Tim Dixon: Absolutely. So, we do that in two ways. So, let’s be fair right? The way it works you spec tires if you are a GM, you know.. Impala tires. You go out and buy a bunch of Impalas, you’re going to put a medium grade tire on that because you are buying so many if you are GM. We instead spec Corvette tires, as appropriate, because we know how important it is so, we work off the same assembly lines, get higher quality product off of that because of our relationships with them. But, more importantly, we do all the coding here in the US in Naperville, IL. We don’t allow that to come out of China, Malaysia, anything like that because we feel that we are going to have more control end to end in what we place in our customers network.

Greg Ferro: And that’s the key to understanding why third-party fiber optics can be successful. You can have poor experience with third-party optics when you know, somebody has just taken a piece of code that was shipped by a factory who just wants to ship a product and not interested in the after-sale support. I think what you are trying to say, or indirectly leading up to is:

There is a difference in just buying the cheapest optics from company A in China to buying them from someone who actually is writing code and software and able to troubleshoot them and make them work


Tim Dixon: Yeah, exactly. So, you know, we have suffered over the years because third party optics actually got a black eye right? Because people would try these things, they wouldn’t work. Even today, some of who are considered our competitors, you have to suppress alarms for them to actually plug in to an OEM switch. You have to suppress alarms because they are giving off alarms saying “non-compatible, non-compatible, non-compatible.” That doesn’t happen with our transceivers. We know how to design and code those things. OEM’s, aren’t actually perfect in what they do. They actually make mistakes and we even have to implement those mistakes to get our transceivers to work exactly right in those boxes.

Ethan Banks: Ok, so there is a couple clarifying questions I’ve got for you Tim. One is, we have been talking about transceivers broadly but we have also been talking about optics so just to be very clear about what the type of transceivers we are talking about. Does that include copper media or are we strictly talking light fiber optic?

Tim Dixon: Yeah, so it’s dominated by optics. But, you are correct. We also sell TEs, we sell copper, we sell a broad range. In fact, a lot of our customers love that because

whereas the top OEMS are going to try to move you to a different platform and do that by forcing new interfaces, we continue to support older interfaces for a very long time and our customers like that


Ethan Banks: And then the other question there, the other clarifying question. You are mentioning OEM gear. So let’s say I am buying a Cisco switch and I could also, off that Cisco parts list use a Cisco SKU to go buy a Cisco-branded optic and that would be an OEM optic? But in the context of what we are talking about here?

Tim Dixon: Yes, that is correct.

Ethan Banks: So Arista does the same thing, Juniper does the same thing and so on.. You are saying

InterOptic manufactures a transceiver that when I plug it in, it looks to that Cisco box or Juniper box etc.. like an OEM optic.

Tim Dixon: Yes and in fact, our strongest vertical segment traditionally has been the federal government. We are deployed in the DISA backbone, the top tier background for the DOD. Also, in the largest Navy Marine Corps and MCI, ENJEN now. We are very regularly shipping into that. The thing about that is they spec’d these transceivers, we helped them re-write a contract so that it was no longer Cisco-bundled but we saw transceivers as a separate line item. They didn’t call it third party. They didn’t call it OEM compatible. They called it OEM equivalent. So, they used a special term to specify our optics in use in the federal government because they knew they had to have that kind of compatibility.

Ethan Banks: Interesting, ok, this is happening in the show sooner than I anticipated. But this is a great conversation to have, Now, I have been involved in a lot of these over the years. So, for example, I used to run a small data center backbone that was based on Arista and if you plug any non-Arista transceivers in there is a, if you complain loudly enough to Arista support, they will say “Alright, to suppress the alarm and allow these transceivers to work, here is a command.” And they give you this undocumented command, then it quits whining at you and you can go, but they always give you that caveat, “But you know, if we have any problems, we could point to those transceivers and say hey we are not going to support you because you are not using Arista brand transceivers.” Ok so, now as big as you guys with the sorts of customers you have, you have some very high profile, well known customers. It can’t be any secret to Arista or Cisco or the rest of the people that InterOptic exists and a lot of people are using your transceivers and your products. What’s the stance on all of that? How is all of that handled? I mean, obviously you are eating into some pretty mighty profits that these guys make off of these products.

Tim Dixon: You just have to look at it from a business perspective. At the end of the day, this is a commodity market, right? And we embrace that.

You know, you used to think you had to buy your Dell memory from Dell if you had a Dell laptop. You didn’t. Kingston could make it better, cheaper, faster, higher performance so you went to Kingston.

This is going to happen. I mean there is just no… look at 150, 200 years of tradition in technology and any kind of industrial revolution when you see commodity market, it’s going to disenfranchise OEMs. So, I’m going to say that as a blanket statement up front. But, at the end of the day, if you plug in an InterOptic transceiver, whoever is providing your service contract… they can’t tell that it’s and InterOptic transceiver. From an OS perspective, it looks like a Cisco transceiver. Now, if they go down into the line item, and do a query at the very lowest level, they will see InterOptic in one of the bites instead of Cisco. But up to that point they cannot see it.

Also, practically think of how they use SmartNet. You are going through, you see something, you’re troubleshooting, you start doing some isolation, you say ok. It’s not the admin, it’s not the motherboard, it’s you know, it’s a plug-in card. Oh, by the way, it’s a particular interface. Oh, we are doing loop backs on both sides. You know, we can’t exactly tell but it looks like it’s on this link. Well, even then, they cannot specifically say it’s a transceiver. They are going to look at the DOM and try to see if there are level issues. That kind of thing. But, at the end of the day, it’s either this side of the transceiver, the far side of the transceiver or the link. In fact, most of the time it’s the link. They’ve gone down to the level of troubleshooting that they are going to go to and they are going to stop. So, the person that’s on the site is just going to look at say, “hey, is there a light on that? Is it blinking or is it not? And they are going to flip the transceiver out whether it’s Cisco or IO, the exact same thing happens. And you know, there is nothing that is going to change in terms of the operational, the practical operational experience that happens in these data centers. It just doesn’t.

Ethan Banks: Practically speaking, I would agree with that. I’ve heard the threat come from the vendor about not using their optic or using their transceiver. But, I have never actually been in a position where I had to call in for support and they refused to support me because – hey we took the inventory of your chassis and saw these optics and so therefore we are not going to talk to you. That’s never ever happened to me because — my opinion here — is that they want your business. They are going to support you no matter what unless there is some extremely obvious reason why it would be.. You know, that optic that’s got a problem, they’re going to ignore that. That’s just not going to be an issue.

Tim Dixon:

There is only one thing in the OEM business model, their P&L, and their profit stream that makes more money than optics. And that’s their service contracts

. So, are they piss someone off and cut off a service contract when it’s one of the more profitable, actually most profitable thing in their business besides optics? No. They’re not going to do that.

Greg Ferro: Well I think there are a few things. I’ve certainly have seen in the lower end of the market – so, when you’re not a big player – the technical support people will blow you off. Especially since most of the vendors have outsourced their first level of tech support and it’s in that, they actually have financial incentives not to escalate your case. So, literally they are financially rewarded in gold on not escalating cases up to the vendor who owns this contract. So, if they can get away with saying “Oh you haven’t got vendor optics, and then you say “oh well ok” then, they will do that to you. They will literally give you a bad day.

Ethan Banks: Have you actually seen that? I have never actually experienced that.

Tim Dixon: Yeah so, maybe there is an outlier where that happens. I can tell you from experience they will always, you know, it’s their job. They are going to lay it out there, they are going to use some FUD, they are going to try to rattle the sword. But at the end of the day, at a pratical matter once that transition is made and they are very clear, the customer is very clear with their vendors, to say this is what we’re doing. It’s never affected their service. By the way,

we wrap our equipment with the same and probably better logistics about getting product into the hands of our customers if something goes wrong, which it hardly ever does

. But, we also have sparing programs like you should. We can do depoting if we need to. We can log onto their equipment and look at optical power levels and everything just like a third-tier tech for an OEM if that’s what they would like to have. So, we work with the people that are in the datacenters. We contract with them, we work with them hand-in-hand on that optics part of the network and we feel that is one of the advantages of having us at the vendor table. We can talk about this and I think we should, but the amount of optics that is going in, this is becoming a more and more important part of what a large Fortune 100 company should look at. Cisco isn’t going to provide you the expertise in optics that you need. We were at a global 100 yesterday and that’s what they said. They said “We are so upset, in the past year we have had significant performance problems. We try to get this worked out with the OEM. They have no answers. They can’t tell us what’s going on.” And so, you know, you need to have the right people in the right seats around the vendor table.

Greg Ferro: So, let’s talk a little bit about the technology that’s inside an SFP. Is it just literally a chip of a certain type and anybody could, you know, program one of these things? Can I reprogram the firmware on that myself?

Tim Dixon: So, look, some people even have business models where they’ll sell you a programmer. We can’t get around that. So, yeah. That’s what we do. We buy blanks. We just happen to buy blanks from the top tier optical suppliers and you know, coming off the same physical line as the OEMs and though spec’d to our standards. So yeah. We get blanks. That’s what we do. But, we also have spent tons and tons of time reverse engineering and figuring out how these work in every platform, under every condition and like I said we know that optical architecture and what’s happening there.

Ethan Banks: So, let’s say here is a situation that I’m standing up a new datacenter and it’s a greenfield. I’m assuming I can build out my switch parts list and so on from the vendor.. line cards, power supplies, and the chassis and all of those things and then go “Ok InterOptic, here’s my build sheet for this data center that I am doing, and I need optics for this. I am going to need some 10gb, some 25, 50, some 40 and some 100 and then you would what? Flip around a parts list that is going to meet those needs?

Tim Dixon: Absolutely. And you know, where do they want it, you know, drop shipped? How are the logistics? We pride ourselves on logistics and getting… we’ve never had a situation that they were waiting on optics. Let’s put it that way.

Ethan Banks: So, I am buying from InterOptic directly then?

Tim Dixon: We have had in the past relationships with CDW, Ingram Micro, all the big boys. At the end of the day, our customers tend to be better served, they buy directly from us. They have direct service engineering support, I mean, I could get down to some specifics, right? We do serial number tracking of every individual part. We do interim testing of, what does the laser look like? What does it look like when it goes on the board? What does it look like when the can goes on? What does is look like when It leaves the Asia? What does it look like when it comes into Naperville? We do individual testing on every transceiver. There is no OEM that does that. There’s no… I mean I don’t know of any third party that does that. Although, they might claim it, but we actually do that physical testing of every part and the view of what we are trying to do, the traceability is from our telecom days. I mean you don’t see that in the data world, but we can give you a report and we have had people call in the middle of the night, pick up the call, we go log on to their systems, we say “hey no, this transceiver hasn’t deviated at all. The power levels are the same, there’s nothing wrong with the eye diagram. This is what we have seen over the past 6 months. So, it must be something else in your system.”

Ethan Banks: Yes. So.. Yes you are delivering a top quality product and level of service but, the big sell here is that you are also delivering at a cost savings. In other words, one of the things that you spend an enormous amount of money on is the data center build is populating that switch with optics that comes without them and then adding the optics ends up driving up the cost. In some orders, that’s roughly half the cost of the build is specing out all the optics.

If I go to InterOptic, one of the big reasons I am doing that is, is not just what the optic is and how you track it. But also because I am saving a significant amount of money. It gives me a lot of dollars back into my budget that I can put into other things.

Is that fair?

Tim Dixon: Yeah exactly. Let’s get down to the nuts and bolts right? They probably wouldn’t look at us unless they had an economic advantage. So, we definitely provide that. But, you could do that if you went and bought them off Ebay too, right? Maybe even better economic advantage. Our argument is, that’s not where it stops, and especially as we see the data centers growing. We see this kind of demand from the edge with 5g, 4kHD, IoT. This is going to swamp the data centers that we see today. So, if you aren’t doing optics correctly now — that layer zero correctly now — you are just going to bury yourself in problems and frankly, we are in a situation that we’ve reached the physics limit of what we can do in optics. So, everything is going parallel. So, we are going to have to have a more nimble layer zero inside the data center in the years to come, and we’re here to do that.

Ethan Banks: You mentioned things that are driving more traffic in the data center. 5g eventually is going to get there. IoT is starting to pick that up. 4K, right, higher data streams going around. A friend of mine who does audio work and is talking about the data streams he is dealing with. 1.2G of a single data stream. So what optics are you seeing pickup in sales? Is it 100g? Is it the 25/50 market beginning to take off from InterOptic’s perspective?

Tim Dixon: So, 100g for sure. We are shipping that today. So, 25 drops for sure. 50g not so much. But you never know, things always change. But we support the full range of all of those future products. So, we are looking at 200g right now. We don’t deliver it right now but those are the kind of things that we look at all the time. But it’s more than speed….

Greg Ferro: Mellanox has announced some switches for 200g. So, we are starting to see that and none of the other vendors has actually announced shipping products around 200g like Mellanox has. That’s an emerging market there.

Tim Dixon: Yeah, exactly. But my argument would be that

it’s really more than speeds and feeds. Just like it is more than just cost.

We fundamentally think there are going to be some pressures inside the data center that are going to say, “Hey, is it better for us to do wavelength manipulation, wavelength switching, instead of terminating every one of these things on a big router box or big ethernet switch. So, I think you are going to see kind of a migration to a smarter optics layer in the data center over the next 5 to 10 years.

Greg Ferro: So, what you are saying there, I think is you are actually saying the optics themselves are going to become networking devices. So, we have seen SFPs, of course wave division, or dense wave division. But, most of the manipulation of the bandwidth has been done in the switch.

Tim Dixon: Yeah, after you terminate that line.

Greg Ferro: So, the SFP module has literally just been like an adapter. Optical to electronic sort of that a serial signal comes in and plugs it into a parallel box kind of thing.

Time Dixon: Yeah, point-to-point.

Greg Ferro: And what you’re saying is that we are seeing the SFPs become intelligent in their own right. They can start bonding together multiple fiber optics or multiple channels into whatever you need. Is that what I hear you say?

Tim Dixon: Well, so you could say it is in an SFP format. I would argue that

there might be something in the interim where you are actually doing wavelength manipulation at a very low cost, at a very compact size, that if needed up top a rack or kind of peppered around that data center to help you more efficiently work in the optics realm


Greg Ferro: So what does that look like then. If I’m an engineer listening in and I am going.. ooh, that sounds interesting. What does that mean? Is that longer haul like longer distance 2km over standard single mode? Is it bonding together moldable fibers in a single SFP?

Tim Dixon: Yeah, I mean the thing about the data center which is different than the long haul. I just want to look at some contrast here. If you go and talk to Verizon and they put forth the fact that, “Yeah, we will never switch from single mode. That is what we have installed. Any solution has to be in that realm.” Well, you know what, the data center isn’t the same way right? Yeah, you have a lot of single mode. You do have some multi-mode. But, you could actually lay in a completely separate layer – let’s call it multi-core fiber where you have multiple cores, be they single mode or multi-mode. 6, 8, 10, 12 cores inside one physical fiber, housing, or sheath, and so you can then add that more complex layer at the top of your hierarchy and get some bang for the buck in manipulating wavelengths in kind of that realm before you drop it off into one of your large routers or in a blind switch.

Greg Ferro: So what you are saying is that you are able to mark say.. normally a 100gb which is 4 by 25gb lanes, you can mark that onto a single pair of single modes for some distance. You are sort of suggesting that there’s SFPs coming which are for more than 4 channels like 8 channels or 16 channels on a single pair?

Tim Dixon: Or even fiber that has multiple cores inside it. So, you have 4 or 8 inside a single mode and then you have 8 of those single modes inside a sheath, right?

Greg Ferro: Ok so you wouldn’t be using a standard LC then. You would be using like an MBO, which has 8 cores presented in a standard RJ-sized socket, you plug it in, there’s 8 cores so why didn’t you send 400gb down out of one SFP?

Tim Dixon: Right, yeah, and start manipulating chunks of that before you drop it off.

Greg Ferro: Alright I got it.

Tim Dixon: Right so, there are a number of… And again, this is early days, we need to work on the standard side of things to get things moving there. But really, at the end of the day, we have reached the Shannon limit on fiber. We have to be able to do more inside kind of at that layer zero realm so that we can be efficient as we build out these very large data centers because of the demands they are going to see with IoT, just the number of end points and yeah you can work on fog layer and all kinds of things to try to keep it localized but at the end of the day, you are still going to have a huge demand inside the data center.

Ethan Banks: You have mentioned the physical limits a couple of times now, Tim. I think there are a couple of things worth mentioning. One is, that I’m curious about is, what is the practical fiber optic limit for how much data we can push down that link over what distance? Then the other is, copper, not copper going in between switches let’s say, but the issue of dealing with an electrical trace on a board. So, you go from the optical input on front of the panel and then you got to convert that signal to electricity on the board with what we have currently because we haven’t made the leap yet for switch internals to run on photonics and so we have a challenge there, electrically as well. You have comments on those things?

Tim Dixon: Let’s step back. So first on the limits, I should probably send you a chart. It’s something that is readily available right? It’s from Alcatel and some things that have been done in the past. So basically,

you saw a 78% year over year increase in optics capacity from back in the 1960s through 2010. So, for 50 years you saw that. Well until we reached the Shannon limit on a single fiber and now since 2010 we have essentially zero growth. Zero growth in the amount of traffic that you can ship across a single fiber. And, it’s a physics limit, you’re not going to beat it, ok?

No wonder you have seen parallel usage of lanes on fiber, right?

Ethan Banks: Ok and that Shannon limit, so this Is new to me. I just Googled while you were talking about that so quickly to find out by Wikipedia that the Shannon limit is the capacity of a communications channel. Its theoretical max of information transfer rate of the channel for a particular noise level so as you are dealing with..

Tim Dixon: Yes, the noise level so it’s optical signal to noise ratio. Yeah. Right.

Ethan Banks: Ok got it.

Tim Dixon: And It changes. It changes for length, it changes for the standard kind of parameters, but at the end of the day, we have kind of reached that limit. So, for long haul, you get more headroom when it’s shorter distances, but at the end of the day, we are nipping at that limit. So, something has to be done. We have to go to parallel but the more we go to parallel, the more complex things are happening and the more cost that goes into the electrical boxes. So, do we want to look at ways to solve some of these things in the photonics realm? My argument would be that as we see these data centers growing we will need some of that.

Ethan Banks: Practically speaking though, with photonics that isn’t coming as quick as we thought it was going to. We’ve been hearing about photonics on this show for — well geez, Greg, for a few years anyways — but we still haven’t seen a commercial product come and last thing I heard was from Arista where apparently photonics is challenging enough to get right that mass producing a photonics base board in a switch is sort a big deal and we still got challenges that way.

Tim Dixon: So, there is a simple implementation if you look at my LinkedIn, we did this back in 1997, 1998, 1999. But, it’s used at Google. They use a photonics switch at Google, right? They see that as an advantage. Now this is a simple photonics switch. It doesn’t have any wavelength manipulation, it’s kind of dumb, kind of thing. But you know, it’s in use today. So, it’s a 256 x 256 photonic switch. You know, I worked with John Bowers and those guys at University at Santa Barbara way back in the late 90s to develop that product. We’re speculating and kind of pie in the sky stuff but what I am trying to say is, just look at the dynamics. So, if you had a growth rate of 78% year over year, you go to essentially zero growth rate now from 2010 to 2017 and in the foreseeable future, you look at a Cisco VNI, kind of their projections at overall capacity growth of 26% or so. You look at 86% growth for compute power at the highest level. You look at kind of averages of where other people think, it’s about a 50 to 55% growth rate,

you are going to get to a point where internet traffic, internet capacity, internet demands are going to cross over that zero line, that zero slope line of optics and it’s a ticking time bomb.

It is, so as much as OEM vendors want to tell their customers, “Oh it’s just a transceiver, don’t worry about it.” Well, they are doing a disservice to everybody in this industry because of it.

Ethan Banks: Ok so let’s bring it back to InterOptic then. So, as we advance with the speeds of the and the technology to get these speeds of the new optics, when does InterOptic say, “Ok, we are going to offer a 200gb transceiver. We are going to offer eventually a 400gb or maybe an 800gb transceiver as these develops over a few years.” When do you guys jump into the market with that product?

Tim Dixon: Again, I think you dropped back in to the speeds and feeds. We will always, we will always be able to keep up with the market and enter at the right time because of our partnerships with the top suppliers. So, we are already working on those advance technologies and we will have those available when the need of the knee of the curve hits in terms of usage right? And actual product demand in the marketplace. Speeds and feeds, at the right value proposition. That’s always part of our nature. Now the question is, are there more complex products that are needed in the data center? We’re working that right now with some of our larger customers, trying to understand what their needs are, where we can add advantages, you know, as they become more and more comfortable with non-OEM solutions which are just rote solutions. Are there ways they could advance what they are doing in the data center, you know, I can’t give you a specific timeline on that. But we will see those.. that transition happening. We want to be at the core of that.

Ethan Banks: You answered the question though, which is, when the customer demand is there, we are going to have a product no question about it. In other words, you’re very aware of what the needs are and when that demand is there you will be able to meet that demand. It’s not about coming out with the latest and greatest thing that one customer might want some day, it’s about we are seeing significant adoption of whatever the product is and you will therefore be offering that product at that time.

Tim Dixon: Well, even to the point of working with customers to help define that product. To find those new layer zero platforms that they might need, or look at, or provide them advantages, yes. I mean from our point of view, we want customers to go out and pick best-of-breed, best-of-breed router, best-of-breed firewall, best-of-breed security, right? But, we think they should look at best-of-breed optics in everything that they do, and we hope to be there to do that for them.

Ethan Banks: So, companies that want to engage you guys, that want to talk to InterOptic and find out more, where do they go? What is the best way to get in touch with you guys?

Tim Dixon: The easiest way is to go to our website. We have some contacts, we have some super people who are industry veterans that know how to, not sell, but help. And, go to our website, do an inquiry. We will be right there to help you.

Ethan Banks: That’s

Tim Dixon: That’s correct.

Ethan Banks: You guys have a Twitter handle as well. That is, @GetInterconnect. So, if you are listening to this and you are interested in InterOptic, tell them that the Packet Pushers sent you. Now Tim, just curious are you personally someone that is engages in social media? Do you tweet or blog where people could follow you personally?

Tim Dixon: So, all I have is a LinkedIn page. They can definitely get me there. I do answer that often. I do look at what’s going on there. But, that’s about my limit. I just don’t have the time.

Ethan Banks: Yeah, that’s a pretty common story of people that are very busy getting things done often don’t have the time for social media, so we get that. Well that will bring us to the end of the show today. Thank you Tim, and thank you to InterOptic for sponsoring the show. Without our sponsors, hey we can’t do the research for the writing of the podcasting that all of you listening enjoy here on the Packet Pushers network. So, you can find this and many more of our fine, free technical podcasts along with our community blog at You can follow us on Twitter @packetpushers. You can find us on LinkedIn. Like us on Facebook and hey, take a minute to rate us on ITunes. We have seen some fresh comments come in lately and we really do appreciate that. Just take a minute and go into the ITunes interface, leave us a comment and that does help the show to grow. And hey, take a minute and tell a coworker that you really appreciate them. Wouldn’t that be a nice thing to do?

Last but not least, remember that too much technology would never be enough.