if you think about Cisco as a company cisco wants to grow their business mid-teens that's what we tell the analysts right we want to grow our business annually mid-teens you know that was kind of easy to do when we were only doing a billion dollars worth of business I can remember a national sales meeting one time when Cisco was talking about because we were growing our business mid 40 percent year-over-year you talking about some surprise on your numbers when you got them the next year yeah they were going up forty percent they were taking good accounts away from me and next thing you know your numbers forty percent bigger but we killed it we killed it we were growing our company that big so guess what Cisco was doing all that as a small company but when you get bigger it's harder to do that so now we're telling analysts where are we trying to grow our business mid-teens at best low teens mid teens what's hard to do that when hack 50 for over 50% of our revenue is in routing and switching and we dominate those and the markets doesn't grow that fast we've got you know if you've got revenue you do revenue and market share here we're dominant players here we get great revenue it's way up here but how fast can that grow sometimes it shrinks which is the civillian shrinking 2% annually switch sales we've been holding our own with that overall switches are good let's switch the sales are actually shrinking so guess what we can't change our number with that we have to change our number with the thing that's in here the advanced technologies what we used to call the advanced technologies the thing that's a billion dollar business are better for Cisco so let's just think about Wireless land can we grow our business on wireless lamb - oh yes it's a great growing opportunity but it's probably not going to drive our overall number of over forty billion dollars that 10 percent 12 percent 14 percent we want to but it can help you know we are 50% of the market and if the market is growing we can grow abnormally more than that we want our unfair share of that so certainly we might be able to grow a little bit security video all these things are things that are in there that we we play really well it might be a growing market but overall we still can't grow our whole number the way the analyst might want us to be there to keep our stock price word is that so we have to launch new bubbles we have to get new things out there the MDS 9500 was a new bubble it was revenue that we never had an opportunity to get until we gave us new product out with that new product we launched out storage area networking San now hip you know we've invested 60 70 80 million dollars in it and eventually started growing after we got it out it started growing to them you know maybe a four hundred four hundred million dollar market effort comes so we just put it in there and say saying as a part of our advanced technologies and part of that bubble we can grow that and everything this is new money for us right data center switching and it was in a huge market huge fast market fast growing market and there was a great demand for it so we accelerated what we did so in a matter of it just seemed like every month we were coming up with something different next to 7,000 X's 5,000 X's 2,000 negatives 1,000 not even a switch at software that's bubbles up into here to grow our business because we have to eventually we said we need to attack another market there's another opportunity for us there there's a different type of server out here there's a different type of server out in the networks we have these standalone servers and everything and virtualization is doing really good but there's a different type of over out there this different type of servant I'm going to turn this this way so you folks can see can you see this over here can you see me folks in the different type of server is a blade server because I mean they made a liar out of me I kept saying a server as the server of servers commoditize who cares about servers I wouldn't sell them so wasn't putting money in my pockets anything I love talking about the Cisco said now that we're virtualizing blade servers are a little bit different the typical blade server chassis had what it had well had two power supplies for redundancy it had a management module in it to manage the servers inside of there it had Ethernet connectivity Cisco actually used to sell catalyst switched ports that would go in that for IBM and HP and Dell and everybody else it would have HBAs inside of it and then you would buy server blades that went in there you can maybe get 1 through 12 server blades in there the value of this is is I could get more servers per rack unit with the blade servers than I could if I bought standalone that was the benefit of it each one of these blades had processing and RAM on it he used the power supply and Ethernet NIC cards from all these other things so it was a chassis that's all it was the thing about it is is when virtualization came in this thing had some holes in it too you had the processing so now I start getting four cores here and four cores here I'm running one application on it it's only running 10 or 15 percent utilization I want to run VMware on it so guess what I take a couple of cores run one application on it the problem is this thing only had 96 gigabyte of RAM available on it because of the way they'd been building these things over and over been enough for just one application 96 gig is plenty but if I'm running multiple applications on it that goes away quick let's say my first application needs 48 my second application needs 24 my third application needs 3200 I'm already I even with simple math that's wrong Cisco said this is an opportunity for us to go back to zero and build a blade server like it was built for virtualization and that's what we did that's where the b-series chassis why did Cisco do this blade servers were a thirty seven billion dollar market opportunity new money that Cisco had never had a chance to get we needed a new bubble the UCS bubble if we could just in a couple of years get 10% of that market how much new money is that in the Cisco 3.7 billion dollars of new money could we use that to grow our business a significant amount this make sense this is why Cisco does all the stuff man is it's we've got to get new money we've got to find someplace where we can't grow the markets where the inter new ones where we can make a difference and this is a place where we could make a difference because we had a way of doing it so let's look and see what we did with the B Series yes it's we incubate more on this campus than anybody ever knows most of the incubation is there here yes not to say all of it is but most of it is so let's just think about this Cisco said let's think about we need a lot of RAM we need a lot of RAM in our blades we're going to have we'd like to think about it one management point for 12 servers if we could do better than that that'd be cool too so our requirements of the differentiate ourselves in this in this market came out so what Cisco came out with was kind of a lot different a big chassis a big solid metal chassis the 50 108 chassis it's just a big metal housing with connectors in it inside of that we're going to put the ability to have eight half wide servers in here so this is where our servers will go into that's why they call it a 50 108 because there's eight slots here maximum of eight servers we're going to put some you know power supplies in this thing in there and then this is just bare metal our smarts that we want to use we're going to actually put them outside of this box we're actually going to have two boxes on top here called fabric interconnects the fabric interconnects have our management platform it's got our management platform in there we'll talk a little bit more about that in a second but there's that's what we're going to do we're going to have it's going to be a little different than what traditionally was out there so I can put my server inside of here now on the backside of this I'll do it in like this we're going to have connections from here to there we're going to have 10 gigabit connections we're going to have the ability to have 8 10 gigabit connections small beginning 8 10 gigabit connections up here so we'll take a the possibility of 8 10 here and up to this guy so I'm going to have the ability to have a hundred and 60 gigabyte of connectivity into the in and out of this one chassis out of this one chassis and inside of each one of these connectors here is going to be Ethernet that will go out this way to the user so the user can get in and out and it's going to have Fibre Channel over Ethernet that will come up and go out to the storage area network so we're going to have unified cabling one cable is going to carry all of that stuff for our servers yes yes absolutely yeah you can actually do fiber Chema straight out yep yeah does this make sense this is what cisco came up with and the cool part of this is is so when you buy it you or buy your servers here now again the problem with these servers these blade servers ended up being the memory issue so cisco went to Intel and said hey listen we can build some Asics here and we got some ideas about this typical memory into a processor let me do this in green so if you've got a processor here it's going to be able to see memory coming RAM memory coming in in these these DIMMs writes memory you put in there and it can only see it see four of those which is limiting us over there cisco said can i cheat this and will you support it they said cheat this how do you mean what I want to do is I want to build an ASIC here that makes this guy see four dims but I want to put four times that on here I want to be able to represent 384 gigabytes of RAM back into your processors if I build this chip until so that's kind of cool and we'll support it this is called a extended memory technology and it's the first thing that we did kind of cool to get our name out there it's the very first thing that we did what does this mean it means you could virtualize more applications on our blades that required a lot of memory than anybody else at the time now it was a short win short term win but it got our name out there and customers started looking at this at least they were considering us at least they were considering us but we had other benefits that I think are even more more attractive remember I told you that we've been working with VMware we were actually part owners of VMware and we've been in there looking in their labs of things they were doing and Cisco said you know what there are some really cool things that you'd like to do VMware did you go VMware used to be in the is in the switching business even before they require this other company they actually if you think about it if I run multiple servers on here I'm going to have multiple NIC cards multiple VLANs all this that any other so VMware figured out that they needed a software switch on their hypervisor software in order to be able to do that to do all the QoS and things like that well when Cisco went in there and said what are you guys doing they said well we're trying to work out this software switch thing and we're not really good at it cisco looked at them and said won't you let me take care of that I'll do that for you I'm good at switching I know how to do that so cisco actually came up with software called the 1000 v the nexus 1000v which is a software switch that does all of that for you matter of fact if you buy the enterprise the large highest antic comes in all the VMware inside of that why so we can do QoS on these multiple machines we can do all kinds of cool things that I'll show you in there we our core competency with switching even in software we could do it better VMware could because they that's not their core competency so cisco actually created the 1000 v software switch they also said our interface here we're not our management's not going to just manage one of these chassis x' you can manage more so if a customer buys one chassis one these fabric interconnects and wants to expand and have more they buy another chassis here of 50 108 chassis they put the servers inside of it and then they run that up here to these guys what if one management point for 20 of these chassis x' for 20 of these chassis x' 20 x 8 that's a hundred and sixty you can manage a hundred and sixty servers instead of twelve at one management interface that's what cisco did customers like that too less management interface is easier to manage and all that we also said found out and this is i'm sure it was hard for cisco to do this to say when we design our management we do not need we do not need we can't be the lead you know when you go dancing a lot a bit very good dancer but you know what i've learned enough to people can't lead when you're on the dance floor you step on each other's toes that looks pretty stupid when you're out there by the way guess what we know that vmware is going to be the lead in the data center because they know what's going on with the application they actually know the performance of the application all the important stuff about the application so we're going to design our management to be subservient to vmware but give them a lot of information and by the way we'd like to make it to where instead of having all these management points you could use our management's for servers and actually make a server profile inside of our management that has three distinct values I'll have my IT organization put all the MAC addresses and VLAN information all the networking stuff in this software and then somebody from the storage the sand group can come in into the same software the same application can put their information in and then the application guys can come in here and tell us all about their application characteristics we can put all everything that used it's like 61 different characteristics 61 or 62 characteristics they used to take three separate management platforms to find out everything about in one we can make something called a server profile that is one pane of glass that has everything about that server that we need to be able to move it anywhere we want to and customers love that what's that mean that means with within 20 of these chassis x' if i have a virtual machine running right there i can move it anywhere on any of these other 20 chasis in the same management plane anytime i want any way I want just like that and who tells us to do it not Cisco VMware does we let VMware tell us word of it they say we you know it might be a time of day thing you know up and you know from 12 o'clock at night to 6:00 in the morning we're running our application on to core processors and 24 gig of ram but guess what set 8 o'clock in the morning or 6 o'clock no more so a lot of people start coming in my utilization goes up by use my application says I need more help we can move it over here into this because we've got four you know eight cores and 64 gig of ram the same Fang put it there do you think customers would like that yes sir how's that unique to Cisco it was designed for designed to do this customers love this this is still unique to Cisco this is still unique to Cisco this management interface sorry machine anywhere and you that's a that's what VMware does but it's how easy it is to manage this and this management plane of are totally unique but yeah if you are to a clear into CPU machine which is gardening service propriety and it is supposed to let see the city from need more computing capacity reduces your machine comes with a service provide that most of your machines in the sea this typically takes about hours or days yeah in a traditional one it might take hours or days you know this we can do it in a matter of sex matter of seconds yes I mean what what I mean this is but I think I can add something yeah just want about it yep it's still that you can take a stripper pole five and applied to a slough not just for server-side yeah you can set a slot to have that particular profile and then once you buy the sorry there's budget that would inherit what the start time we can do that hydrated I think yeah just as beautiful longer it's been doing it for a long time I've ended all dogs in here what should I wish to understand the different charity about because you know if we tell people the first thought is they saying this my shoes not cool like things yeah walking on one of the differentiators you have one match one layer without a modulus at night instead of having one per chassis that if you would have so that would be kind of the major differentiator okay I think the scalability is another one terms of how much you know p.m. you can support on a certain yeah sorry yeah guys want one one as if you see it here is a three characteristics that as your droid and what's a differentiator maybe you add to this but then you move the virtual machine there is no not you yeah as we Americans before but when you move it we move you change the characteristics for networking and the storage area network so you do not need to you know if you just move it with HP yeah yeah to the you new hardware you've got a new MAC address and you listen others it's physically to the most adapters are different so you should now we can reprogram all three elements go at the same time so your MAC addresses are still the same your IP addresses are still safe everything is exactly the same yeah much much much again to this way that IBM's doing it they do through that virtual MAC addresses and they are a whole table you can do across a set of chassis okay it may have been new voices for us well that's what I'm saying again we're driving distance but we drove them to do this right and again it took somebody from the outside going in so if I was talking to a customer about this I'd say well guess what Cisco saw innovation here innovation that wasn't happening with your traditional vendors because they were certainly happy to just recycle the server's over and over and over who wouldn't be who wouldn't be but Cisco's came in and said if it we had why and it goes back to the same thing with IP telephony Nortel anivia had an opportunity to do IP telephony before cisco did why didn't they do it before any of this happened why didn't they do it didn't we eat their own young and when they when they it was a cash cow yeah they've had they if they wanted to do a change they had to drag a whole bunch of already installed stuff why did Cisco do IP telephony they said if we start if you started this and wanted to do it as an application in 1999 if you could wipe the slate clean and start it how would you deploy it exactly like Cisco did that's the exact same thing we did there if you because the design of the HP and IBM all those servers up to that point they had been designed for this and the traditional vist when VMware came it changed the game Cisco was clean they had a clean slate it's easier to draw a pretty picture on Nelson on this piece of paper then on this piece of paper right here's this card is all this about my guy I'm saying this cisco innovated in a place where it was it would be a lot more difficult for HP to innovate yeah we did yes Plus is this because if I said to someone Cisco did this was just going to dip ago so long look now first of all and I'll show you where we're doing that I'm not finished yet I'm not finished yet but but I've got to show you what were the innovations in order I can't show them all to you at once right but we all agree that extended memory technology for about six to eight months when nobody else had it do we all agree that on a single blade Cisco was the only person to give you four sockets on a blade on a blade server that nobody else could nobody else could do that for sockets know all these were to Cisco could give you four yeah and then again I'm not saying you didn't care they didn't catch up or changing everything but did they lead the way who do you mr. customer want to go with somebody who is sitting there happy to just take your money or somebody who will enter the market and innovate it for you that's the conversation I would have with historical value of this that's what I would say but Cisco entered this Mart and why did Cisco do it again we did it for own good our own new bubble but we came into a market that was ripe for innovation ripe for innovation so again all of this stuff worked really well we start doing it and again I think our server profiles are still a very very valuable asset because you move everything everywhere it's one pane of glass to put everything in and it's designed specifically to be subservient with VMware and all this than the other yes it gives us a very very good story but we want to do more cisco wants to do more remember that nexus those nexus switches and stuff again let's go back to the network where we're best at in theory what because if customers were wanting to do that and had multiple data centers in places maybe we wanted to start consolidating data centers or actually using data centers think about it if i've got a data center over here a server running an application in this side in this data center let's say this data center is in LA and I had another data center in Phoenix Arizona out of desert nice place to be actually and I had some hardware here would it be easy to move that same let's see its IP address is 10 10 10 6 would it be easy to move 10 10 10 6 all the way across and make it pop out over here absolutely not think about my rack and by the way there's another server setting right beside of it that's 10 10 10 seven running the same application and they have to talk to each other on the same subnet whoo when when this server starts grabbing some database he actually sends a broadcast out says I'm grabbing this database so they know how to share that database so it stays valid Wow that was a challenge but Cisco wanted this accept that challenge and do that so what Cisco did you know we would connect these guys up and we'll talk more about into the two case we created a fan-out switch called a 2k in answers to the 5k in which answers to a 7k switch and what we did is we create they created a whole new technology a whole new way of doing things called OTV of late transfer transport virtualization to where you can actually move that server all the way over here with the same IP address and when he sends a broadcast to that server over there in a totally different area we will encapsulate it and know where it's at and release it over here so this guy still thinks it's over here we can take 16 data centers and make them look like one this way do you think that customers would like that that means that you're not only saying I can move a server in any of this architecture infrastructure here limited to only 20 of these chassis but now I'm saying I could put it in any of 16 data centers and move it over here and still have them work yes expertise yeah position ahead of you in any what happens to the application well it's going to be replicated in both places right it's going to be replicated in both places that's what EMC and all those guys did you're going to have an array here and you're going to have a right here and it's the software that will make the date of work everybody knew to be savory one physical sedative hmm how do you tell that operates this instance that yeah using the other stories writing each one we could actually lose we can actually use the same storage over here if we wanted to well yeah probably wouldn't you'd probably want to do it here yeah it's the software with the server vendors that deal with that it's not cisco than this i'll know the deep parts of it so that'd be a good question for your PSS yeah I'm about as deep as I'll go right there does this make sense oh there certainly would be delay and distance and all this this is going to have to be high-speed fiber optic connections and all this that the other but while the customers want to do it let me give you an example hurricane sandy hits New Jersey how would you like to have your data center in the middle of the country instead on a coast where it gets destroyed how fast could I get it there really quick same IP addresses same everything right and I guess what I could start doing this I could say well I don't know if Sandy's really going to hit us let me just move half of my servers over there and wait and see I can prove half of my servers there wait and see and then when it does happen I can move the other servers out because I might have to actually you know if I don't have a hundred percent of the server utilization here that I need or server capacity I need I move half of it here and then then I have to start prioritizing which server blade do I use for what application but at least it gives you what you need it's making sixteen yeah and I don't know how they do that in the backend I really don't and again I'm not a PSS with that and is Dortmund typically doesn't just and let's see that this area wants to see works I mean the other is whether the utilization is proper there around engaged by the utilization is not up to you Rock you can seamlessly solutions here so no management perspective this gives our facility to be administrators yeah absolutely I mean it it takes away limitations and I know that people are using this I don't know exactly that what you're you're doing done it so you'll have P SSE's who can tell you everything about that I've gone to my limit of my expertise here and you know man's got a nose limitations Clint Eastwood said a so this makes sense cisco has done this in a very very very short period of time think about it when did we introduce switches I mean servers yeah 2009 Wow 2009 we did this in a very short period of time is cisco been an innovator here whoa we've been leading the pack yes sir it is this number it's number two in North America so it is number one in Australia mathematics the customer that you require is some 17,000 customers to use it constantly when you invest melibea how many 2121 customers our diamonds are a buyer stuff well Wow well so all in all the Cisco infrastructure the architecture that Cisco is starting to put in place here again is evolved pretty quickly to be honest with you for in a very very short period of time siskins put a lot of money where their mouth is an R&D and acquisition and this that in the other so in essence we have we've got a the B Series chassis we have the C Series servers what's a C series server so you may know the C series it's a it's a standalone server right again built for virtualization and everything but again it's got its own power supplies and again it can have a lot of different cores and all this that in the other but it's really a standalone solution now can it be managed from that fabric interconnect yes absolutely can so you could actually manage more you get to manage your standalone servers with virtualization as well as your blade servers all in one interface which is a pretty he think it's something different as well this Cisco does again so we so we have these different services so let's look it's kind of start from the ground up and work with some of the things that we're going to have in this and I'm going to try to draw this on the fly boy are we all in for an experience so we're going to have our servers we're going to have blade servers in racks here with our fabric interconnects our fabric interconnects here have multiple racks of blade servers these guys fitting in to these fabric interconnects like this like that so we're going to have racks of servers and using the fabric indicates to manage up to 20 of these chezy's each one of these having up to eight servers in it so a lot of servers a lot of density there for our rack-mounted servers whether they're traditional servers or other servers we're going to have to connect these guys and I'm going to put these some top-of-rack we're going to have to connect those in and out we're going to have the fair fabric interconnect or the fabric interconnectors here the 2200 series now what they are is fan-out remember we talked about switches like this remember I said you have to have a supervisor which is the brains and then you have all these line cards so you've got the supervisor and the line cards well cisco has actually had to create some virtual chassis x' inside of the network to be able to do deliver all of these ports that we want so what we do is on these racks we're going to put the 2200 series fabric interconnects on the top not fabric unit of Effie X's got them they're the fabric fef the fabric interconnects in here that's another 2200 series of fabric interconnects they're actually the blades so these guys will connect multiple servers and all this we can put them in these are standalone servers we can have them here let's say I have standalone servers somebody else's servers whoever they are our servers rack-mounted servers whatever we'll put some 22 hundreds up here depending on what they need all of these are top of rack switches they look like top of rack switches they're actually blades that need a supervisor so you'll have you know several rows of these servers in here and you'll have these top-of-rack switches the 20200 s and they all will fit to the end of row and the supervisor for all of these 22 hundreds will be the 5500 it's the supervisor so you actually have a virtual switch with blades out here so these guys will all be interconnected these guys like that these are the brains this is where actually where all the smarts are these two sir password they just are passed through there's an actual standard coming out for all the the tagging in that they have in here for this I think it's be something I think the limit stays 16 is it 16 16 on this yeah and on the 5500 it for the 7000 these can also answer to a bigger switch the 7000 and I think they can do 20 or 24 on the 7000 a blade on a 7000 yeah but in essence what does this give us it gives us more connectivity and all the intelligences here so our lower cost this is where the smarts are this is where the smarts are so think about it we're bringing all of this all this unified fabric all the Ethernet all the fiber channel everything all the way back to this guy and now this guy can actually feed it back out he can take fibre channel protocol and send it over to an MDS and actually into the sand proper if you wanted to right you can have the Nexus set Nexus 7 K up here which will you know these guys will connect into the Nexus 7 K 40 gig Ethernet whatever you want to put there these guys will support 40 gig Ethernet back into here so you can have several of these blocks of this over and over and over your cookie cutting the way you would design find it having this one module here this one module here this one module here so the top of rack end of rope kind of solution that you have these 22 hundreds are not smart they're not smart and you can get those to support 100 megabit gigabit or 10 gigabit ethernet depending on which the flavor that you have Ethernet coming out of your service yes sir well the fabric that's what that is a fabric extender that's what this is fabrics I'm sorry fabric extender that's what it's called FEX fabric extender right 22 hundreds back into the 5k does that make it makes sense the fabric interconnect can either go directly into one of these or back into this guy itself yeah yeah yes absolutely yeah depends on yeah you could have and again I won't do the design every design might be different but if I have racks of servers here two standalone servers and everything I'm probably just going to have the 2200 unless somebody still wants to use a sand traditional sand and fiber channel then I would have some fibre channel switches in here as well and I'd only be dealing with Ethernet I would be doing you know unified fabric so it makes sense so there's a lot again there's a lot of different ways there's no one datacenter it looks like every other data center yes it's actually shipping now yeah 40-gig is shipping in here 40 gig and hundred gigs on its way yes absolutely the Nexus 7 K is already ready for it to 5k do 55 hundreds or 100 degree yeah absolutely absolutely I mean I don't know where it will end I don't know where I'll end but it'll be faster than 100 hundred gig I guess because I've lived to see that yeah so that's kind of kind of how this solution would work now Cisco's made some relationships they've tore down some relationships what do you think you know HP and IBM thought about Cisco when they started selling servers boy did that sour of relationship absolutely call Cisco a lot again Cisco in today they've got this triangle of support now you know this love-hate relationship with EMC and VMware we've created these the Vblock well so I'll show you the Vblock and what the Vblock looks like the Vblock it that we created this organization called VCE which is VMware Cisco and EMC all together where they actually can sell you one solution it's a box that's got EMC software EMC hardware and software for storage it's got VMware for the virtualization of the servers and Cisco's servers and switching capabilities all in one in essence part number it's called a V block and I'll show it to you online when we get to the product section we've also got to a connection with the net app it's called flexpod and in essence the storage is flexpod vmware and cisco infrastructure for that and again this what we're selling with all this stuff and I've been to get into the subject is what what we're selling with that one product number is a cloud we're selling VM VirtualBox revisers if we were just doing standalone servers what a service provider be able to do application support for customers think about all the servers they'd have to have and all this that in the other two different operating so if we were just doing standalone service it'd be impossible for a service provider to actually think about offering a cloud solution but when you've got this technology now with this virtualization to be able to maximize the use of your resources and again if we think about service providers they want to sell the same thing over and over and now they can they can sell the same physical server over and over and use it over and over with virtualization so it's led to the cloud the opportunity that the cloud is it really has so from a service provider for a service provider it's helped them out to actually start developing a whole new revenue stream a whole new revenue stream of being offering services hosting services for customers that they probably couldn't Oh them doing is well absolutely it's it's it's not only this you know your service providers name it might be a lot different than your ever thought yeah when we talk about HCS tomorrow hosted collaboration solution the providers for that are a lot 90% of them are non-traditional service providers they're not the AT&T IBM's British Telecom all this of the world it's the other partners who are doing it and they're doing a very good job of it they're doing a very good job in a lot of places and they're it's a whole new revenue stream for them oh absolutely absolutely it is the relationship yeah exactly the barriers to entry that that market have came down have came down used to be if you wanted a host voice you had to have a big class 5 switch and all this infrastructure and all this you don't have to have that all anymore we're talking about servers and applications here so this virtualization all the things that we've came up with are certainly you know the Cisco's helped develop here and I think Cisco's had an instrumental role in this they didn't do it all themselves but they certainly took full advantage of it has led to a lot of money coming into Cisco that new bubble really nice for Cisco and it has soured relationships some relationships have healed back a little some relationships are still going to be not as happy as it used to be I personally you know I don't know what the numbers showed out because IBM and HP used to sell we sell a lot of Cisco stuff boy you don't think they don't want to sell anything Cisco now and vice versa Cisco would never yeah it's caused a lot of grief in the industry as far as it goes with that and some customers have actually mentioned it to me god I was just Cisco HP and IBM thing would quit well everybody's got to do their own business everybody's got to do their own business so that's the data center pert in particular and again these 7 KS the 7 KS you know they have an outward facing in the inward facing we actually have two different series of cards that go in these things you have F cards down here at the bottom that go here so when you look at a 7k it's a cards that go in it you'll have modules that go in here that start with an F they point downward that's their job you'll have M cards M cards point out they do routing tables and all this that in the other so there will be different cards that go in these and I'll show you that online when we get into products so Cisco's got a lot of things yes sir people using the nexxus you know five and seven caged without going to the ucs computer yeah yes you know there's a religion that's a religious world right this is not so much of religion this is a really good use case for this when you get in the servers you you enter a different layer of religious war so there's and again for our collaboration solutions if you think about we're selling a lot of servers there when we do we have you see on UCS right on our V Series that'd be great but our collaboration specialists are not going to fight this war for our data center people they'll say put it on a C series customers are much more familiar with putting it in a c series of rack-mounted server and it doesn't fight that religious war so yes yes yes sir with with the switches the nexus OS all right so listen first of all it's not a new operating system all right so this operating system has been around since 2002 it was the same operation operating system that was on the MDS 9500 that was released in December of 2002 and became the number one director class fibre channel switching in the world it's the same operating system it's different than iOS now why did we have to do it the iOS has been around for a long long long long time this is a Linux core with our iOS on top of it that's all it is what does it allow us to do it allows us to give not only Hardware redundancy but software redundancy because what we can do is we can start isolating things that we do in services services software services and if these individual services lock up so let's say that I have a nexus 7 K up here part of the services they'll be running up here might be a BGP a service of BGP if that service starts here do I want it to stop all my traffic that's going in here maybe not so putting a Linux core on this allows me to do software and hardware resiliency so I can have a software restart the service in the Linux box and if that doesn't work I can always use this Hardware of another interface another card so it actually gives us two types of redundancy it gives us the ability to add other services on top of it which I OS was designed back in the late in the mid-90s and it was closed and we couldn't add any other services so now we could actually put other people's applications in conjunction with ours and work very well so it's really opening it up now no old iOS is not based on Linux no it's not at all but all the new stuff is the new catalyst 6500 and the new 4500 is all running on Linux Linux core running a Linux core they're running iOS on a Linux server inside of a Linux server we've ported iOS the capabilities of I always to run on a Linux based server next this is the same thing yeah absolutely the last thing that goes in this I can use every iOS command matter of fact when I saw the MDS 9,500 for the first time they looked at me and said you're an old switch routing guy and I said yeah he said guess can you sit down here and tell me what commands we missed if an iOS guy can configure any of these boxes the exact same way they could do an old router back in the 90s but if they want but now if we want to have a GUI it's much easier to have a GUI because it's it's running a different version of them just right they just had to change the name of it yeah so so Iowa nexus is the operating system I can still use all the commands of iOS and do everything I ever wanted to an old router guy like me command sets the same just the inner interface is different so I can have a GUI interface like people like in a data center they certainly like GUI instead of command line or I could actually do command line I have the best of both worlds but the real reason under underlying reason for that is to put it in a core where individuals so versus don't stop other services so I could have an operation in Linux a thread in Linux that's my BGP tables and if that that starts having issues I can restart that service in this card try to restart it here and if not go to the next card so I have multiple redundancies restarted here it doesn't work go down here but does that affect any of my other stuff going on over here no it isn't affected at all I can compartmentalize now it's it's you know it's why don't we use COBOL and basic anymore that's really it it's the next generation and Cisco's move to it yes oh well I wouldn't care but I'll be honest with you purport 10 Gigabit Nexus is cheaper than 65 6500 if ten kegs not big to them and everything then yeah 6500 yeah yeah absolutely that makes it I mean I hope I've answered the address to questions and stuff and again since guy just moved on they had to with the with the operating system and they wanted to run it on a Linux because you can use multi multi processors now we can have you know this product this process using this process of this process using that processor so kind of much more flexible for us
