Arista Networks Inc (ANET) 2016 Q2 法說會逐字稿

Welcome to the second-quarter 2016 Arista Networks' financial results earnings conference call.

(Operator Instructions)

As a reminder this conference is being recorded and will be available for replay from the Investor Relations section at the Arista website following this call. I will now turn the call over to Mr. Chuck Elliott, Director of Investor Relations. Sir, you may begin.

Thank you, operator. Good afternoon, everyone, and thank you for joining us. With me on today's call are Jayshree Ullal, Arista Networks President and Chief Executive Officer, and Ita Brennan, Arista's Chief Financial Officer. This afternoon Arista Networks issued a press release announcing the results for its fiscal second quarter ended June 30, 2016.

If you would like a copy of the release, you can access it online at the Company's website. During the course of this conference call, Arista Networks' management will make forward-looking statements, including those relating to our financial outlook for the third quarter of FY16, industry innovation, our market opportunity and the impact of litigation, which are subject to the risks and uncertainties that we discuss in detail in our documents filed with the FCC, specifically in our most recent Form 10-Q and Form 10-K, and which could cause actual results to differ materially on those anticipated by these statements.

These forward-looking statements apply as of today, and you should not rely on them as representing our views in the future. We undertake no obligation to update these statements after this call. Also, please note that certain financial measures we use on this call are expressed on a non-GAAP basis and have been adjusted to exclude certain charges.

We have provided reconciliations of these non-GAAP financial measures to GAAP financial measures in our earnings press release.

With that I will turn the call over to Jayshree

Thank you, Chuck, and thank you everyone for joining us this afternoon for our second-quarter 2016 earnings call. I am pleased to report that we had another strong quarter. Our revenue exceeded consensus meaningfully as we grew 37.4% year over year to $268.7 million while earnings per share grew to $0.74.

New and differentiated EOS platforms fueled our growth across our top verticals, and services contributed over 12% of our overall sales. From a geographic perspective our customers in the Americas generated 75% of our total revenue while our international theaters progressed steadily through the quarter. We delivered non-GAAP gross margins of 64.1% in a highly dynamic and competitive industry.

We now have approximately 4,000 cumulative customers, and they are enthusiastic about the rate of customer adoption especially of CloudVision with over 100 customers in less than a year from its introduction. In Q2 we unveiled the next generation of Universal Leaf with its many roles, called the Arista 7280R series. Unlike other fixed-function devices, the 7280R is our value-driven leaf with multiple use cases for routing, storage and digital media.

When combined with our FlexRoute technology for the diverse protocols, the 7820R challenges the notion of separate routers by combining multipurpose routing and switching functions into a small one RU or two RU compact form factor. With its traffic companion that we launched in March, the 7500R we embark on our expansion into adjacent core and WAN routing markets.

It also aligns closely with our growing storage ecosystem of partners including EMC, Hewlett-Packard Enterprise, NetApp, Solid Fire, Nutanix and Pure Storage to bring reliable lossless IP storage as a compelling alternative to legacy fiber channel SANs. The 7280R also underscores Arista's media and entertainment initiative with key partnerships such as the Perry Corporation, Imagine Communications, Lavo and Levian and participation in many next generation organization including the AIMS Alliance, Society of Motion Picture and Television Engineers, Avenue Alliance and Video Services Forum.

Arista has been powering slow-motion replay video to real-time sports streaming, most recently at the 2016 European football championships and at the upcoming Rio Olympics. As I look back at our first half, and we move into our second half, we are executing well towards our goals, our product strategy and our dual pronged operation strategy for world-class automated manufacturing.

In light of the recent international Trade Commission decisions with the 944 investigation, Arista has released a new EOS 4.16 which contains design-arounds that we believe address the specific features that they have found to be infringing. Our major customers are actively qualifying or have already completed certification of the EOS 4.16 release.

There should to be no doubt about our commitment to our customers and to our compliance with law and ITC orders. Arista finds itself in the midst of a tornado market with migration from legacy IP silos to public, private and hybrid workloads in the cloud. We are expanding our inventory and operational capacity in anticipation of demand.

We are excited about our prospects and our acceptance of our software driven cloud networking vision.

Now I would like to turn it over to Ita for our quarterly financials and more specifics.

Thanks, Jayshree, and good afternoon. This analysis is of our Q2 results and our guidance for Q3 2016 is based on (technical difficulties) results is provided in our earnings release. Total GAAP revenues in Q2 were $268.7 million up 37% year over year and above our guidance of $259 million to $265 million.

Our cloud Titan verticals demonstrate continued strength in the second quarter combined with healthy growth from the other key verticals. Service revenues grew with the business representing 12.3% of revenues consistent with last quarter. International revenues came in at $68 million or 25% of total revenue, up from 24% last quarter.

While we continue to focus on expanding our international footprint, you should expect our geographical revenue mix to fluctuate on a quarter-over-quarter basis depending on the timing of US and international deployment. Overall in gross margin in Q2 was 64.1% down slightly from last quarter of 64.4%, but favorable to the midpoint of our guidance of 62% to 65% and consistent with a healthy cloud-titan revenue mix.

Operating expenses for the quarter came in at $97.3 million up 13.5% from last quarter. R&D spending was 22.9% of revenue reflecting growth in headcount and some incremental prototype and NRE investments associated with new products and design around activities related to the 945 ITC case. Sales and marketing expense was 10.4% up from 9.9% last quarter reflecting increased headcount and related expenses.

Our operating income for the quarter was $75 million or 27.9% of revenue, other expense for the quarter was $0.3 million, and our effective tax rate was 28.1% resulting in net income from the quarter of $53.7 million or $0.[23]. Our diluted share number for the quarter was 72.8 million shares resulting a diluted earnings per share number of $0.74 up 37% from the prior year.

Legal expenses associated with the ongoing lawsuit came in at $7.6 million for the quarter below our outlook of $8 million on the last earnings call. As a reminder these expenses are excluded from the non-GAAP results discussed above.

Now turning to the balance sheet. Cash, cash equivalents and investments ended the quarter at $823.8 million. We generated $54.9 million of cash from operation in the June quarter. DSOs came in at 50 days down from 51 in Q1. We are pleased to see continued progress in this area during the quarter. Inventory turns was 3.5 times down slightly from 3.6 in Q1.

Inventory increased to $118.1 million in the quarter up from $84 million in the prior period. This increase reflects additional inventory associated with our new contract manufacturer as discussed on the prior call. Our deferred revenue balance was $230.3 million up from $219.2 million in Q1.

This balance continues to be largely made up of short- and long-term service contracts with some product deferrals related to acceptance terms and future deliverables. Accounts payable days were 56 days up from 26 in Q1 reflecting the timing of inventory receipts and payments. Our capital expenditures for the quarter were $4.1 million.

Now turning to our outlook for the third quarter and beyond. We are pleased with our financial performance for the first half of 2016. With revenues and earnings per share up more than 36% on a year-over-year basis. As we look forward the fundamentals of the business remain strong, we are seeing good adoption of our new products and we continue to execute well on the delivery and design-rounds as required in response to the ITC rulings.

As discussed on the March call we began the process of adding a new contract manufacturer in Q2. We made good progress in this regard having qualified most of our products on the new production lines. We expect to ramp volumes in Q3 and should see products from the new plant contributing meaningfully to revenue in the fourth quarter.

In terms of effects on the financials, we saw some increases in inventory in Q2 and would expect this growth to continue into the third quarter. Our current view is that we would increased investment in inventory by a further $100 million as compared to our Q2 balance-sheet position. As we bring our US manufacturing and supply chain capabilities online, we expect to see some increase in costs and as a result all other things being equal some reduction in gross margins.

We believe these impacts could result in gross margins that range from 60% to 65% depending on how heavily weighted our revenue mix is towards products manufactured out of our new US manufacturing facility and its associated supply chain activities. Our outlook for Q3 assumes that while some products sold will come from the new production plant, the majority of our sales will come from products sold overseas.

With this as a backdrop, our guidance for the third quarter which is based on non-GAAP results and excludes any non-cash stock-based compensation expenses and any legal expenses associated with the ongoing lawsuits is as follows. Revenues of approximately $279 million to $285 million, gross margin of approximately 62% to 65%, and operating margin of approximately 26%.

Our effective tax rate is expected to be 27% to 29% with diluted sales shares of approximately 73 million. [Three notes] are based on our current understanding. We expect costs associated with the on going lawsuit to be approximately $11 million for the quarter.

At this time we would like to open up the line up for questions. Operator.

(Operator Instructions)

Jess Lubert, Wells Fargo Securities.

Hi, guys. Thanks for taking my question and congrats on another nice quarter. Maybe first for Jayshree if you can touch on whether there are any 10% customers in the period, and outside of the cloud titans, where you saw the strength and how you're feeling about the visibility going into the second half.

And then for Ita, I was hoping you could touch on domestic supply with respect to the components and in the event the workarounds are not approved for Q4, how confident are you in your ability to purchase all of the components to build all of your product domestically, and is everything available here in the volumes you would need?

Thank you, Jeff. We do not comment on 10% customers on a quarterly basis, but as I look at the trends for the year, and also verticals are contributing double-digit percentages to the Company. The strongest contribution is coming from our Cloud Titans, which is not just one of them, but many of them. I think as a category they are spending very strongly in the first half of the year.

This would explain why the strength of Arista and the macro economy has been less of a contributor to Arista's results. So feeling pretty good about the cloud titans but feeling quite good about the financials, the service providers and the web and high-tech enterprise as well.

Ita, you want to answer the second half?

I think in terms of thinking about our plans as we move through the rest of the year, we have been pretty consistent in saying that we've taken some steps from a supply chain inventory perspective to give us some flexibility as we work through getting customs approval on the design arounds strategy. I think that, that's still the strategy.

We've seen us grow the inventory significantly. We will continue to do that in Q3. All of that is taken to give us that flexibility. We're obviously very working hard to get the customs approval in a timely manner, and that is obviously a big part of the strategy, but we have taken those other steps as well.

Any update on where customs is on the approval process and when we might hear something there?

No I think that it's a process that obviously there's been some communication back and forth, but it is an ongoing process, and will have to wait to see when they are done with their efforts.

(Operator Instructions)

Alex Kurtz, Pacific Crest Securities.

Thanks for taking the questions. Jayshree, on your web scale business if you could comment on the second tier accounts and how that install base has grown and whether or not going into 2017 maybe the business is a little bit more balanced out of that vertical as opposed to the prior years where a couple of really large accounts drove a big chunk of that segment.

No, I think -- thanks, Alex. I think that is an excellent question and one that we do not spend enough time on. We tend to think of our cloud business as just th Titans, but we are indeed seeing a very nice uptick both domestically and internationally in our web scale business of the next tier of cloud providers. I think of a lot of these cloud providers our natural experts in building scale both in their own enterprise and being able to offer hybrid cloud services. What is interesting about that is that you would have assumed that to be true in the domestic regions, but we are also seeing it internationally.

Thank you.

Stanley Kovler, Citi Research.

Thanks for taking the question. I was hoping -- you spoke a little bit about all of the different product extensions and storage, and other segments that are driving some of the growth. Can you help us understand what portion of the revenue is coming from these or at least the growth coming from these this quarter or next and for the balance of the year.

And I also have a quick follow up for the international side of the business if you're seeing any impact, especially in Europe or Great Britain and given your financial exposure to the Brexit vote.

Thanks, Stanley. Is your question you want to understand our contribution of revenue by product, is that what you're asking?

Vertical.

As I said before there are four verticals that contribute a large percentage which is the cloud Titans, service providers, high tech enterprise which also includes the web scale providers and the financial. There is no question that the cloud Titans are spending and contributing and their increase in budget while networking is a small component has an influence on us.

However, the other three businesses are contributing and growing as well. The same is true internationally. It is the cloud Titans spend and budgets are lot more. We are -- unlike some of the macro situations we're seeing spend in service providers, financials, in the cloud hosting and Internet hosting, in the high tech enterprise and some of the media entertainment verticals and the government. All of these end customers are contributing greatly both internationally and domestically. In dollars and numbers the cloud Titan budgets tend to be bigger.

Just to clarify. I was open to get a sense of the contribution to growth from application. The storage and traditional (multiple speakers) and routing. Thank you.

That is a good question. I will answer broadly but I think there are more specifics you are looking for. One of the things we see particularly with the three out of the four verticals is that they tend to start with a project. And the project like you pointed out will be the compute or storage, or security or data monitoring.

And the application schedule can be four or five of course all of them are centered on a specific data center project. The typical four a five are the ones I listed, the first one the have dev ops or SDN or containerization of virtualization application. The five top projects are these.

They quickly in a few months or sometimes a year moving to what I call mainstream where they are then looking to build out a whole scale of service or data centers. It depends on the state of the customer but there's almost always start out as a project.

The next question comes from the line of Mark Moskowitz, Barclays.

Good afternoon. I want to get a sense here if I could around the gross margin beyond the quarter. How should investors prepare, Ita, for the impact from the US manufacturing operation. The question comes up almost every other day or every day for us if you could address that more in terms of should we expect a step down a couple of hundred basis points and then a step backup and is it temporary or can it be more of a permanent impact?

I think there is a number of different drivers obviously. We set the range of 60% to 65% thinking about 60% of the world where we are heavily dependent on the US manufacturing plant plus supply chain. And 65% obviously being with there's no impact from any of these effects.

I think in Q3 we're pretty much in a normal situation using offshore supply. I think as we move into Q4 we will see more products from the manufacturing plant that we will sell in that quarter, but we will probably still be leveraging an oversee supply chain. That will be somewhat of a muted impact but there will be some impact.

Beyond that honestly it depends on what happens with the customs approval. What is the timing of that? Where do we source components and finished goods as we move through Q1? The best way to think about it beyond that that is if we are heavily dependent we are at the lower end of the range, and if not we are somewhere in the middle to the upper end of that range.

The follow-up for Q3 it is more of a philosophical question but also strategic. There seems to be a lot of noise over the past month or so around the risk of affinity toward the channel. Could you maybe just address your view on the channel and any importance on the channel, it seems like it is a lot more important than your giving it credit for. Just want to get a sense if you could address that as people think about -- the rest of investors think about over the next couple of years as you do growth beyond your core business. How that channel helps you out.

Absolutely, Mark. Arista has always been unique in its approach for go to market and sales including the channel. As you know just to step back a little when Mark Smith was here on the last call, he highlighted the fact that we're going after three categories, cloud class which is our top four verticals; cloud converged which is more of a turnkey solution for enterprise, and this is really where our partners both our channel partners and our technology partners come in; and then cloud scale which tends to be the direct cloud providers.

If you zoom in on where our partners are meaningful, it is really technology partners and channel partners, and channel partners especially internationally have always been a huge element of this. The rest of go to market strategy is really three-pronged and one prong of that is very channel driven.

That is very much the enterprise and turnkey solutions that we have to provide both with our technology partners and our channel partners. And another prong of that is outside of the domestic area we are greatly dependent on all our categories being fulfilled and driven through channels more. Hopefully our commitment to the channels is clear, but our focus on quality channels rather than just signing up a whole bunch of them is also clear.

(Operator Instructions)

Rod Hall, JPMorgan.

Thanks for taking the question. I guess I had one clarification which is on your gross margin comment, Ita, and that is could you help us understand could you help us understand what the timing on whatever impacts we will have will be and by we and you think most of this will have affected gross margins and will to be toward a steady-state. Could you kind of restate that give us any further color you could give.

And my question on seasonality, the H2 and H1 seasonality we had anticipated would be relatively high considering availability 25 gigabit mix and availability of 100 gigabit networking. I think most people on this call that are tracking that expect second half of year to be accelerating from a hyperscale deployment point of view.

I wonder if you guys could comment on what you think H2 over H1 seasonality this year is going to look like. Do you think it will be roughly similar to last year? Do you think at least color wise we should see accelerating investment by these guys, just give us some color on what is going out there on the hyperscale guys, thank you.

I will take the second question first and then Ita can clarify the gross margin again. When we looked at second half seasonality now both last year and this year, we see a similar pattern which is some of the cloud guys their year ends in Q2, and so they have seasonality in Q3 and then they pickup in Q4.

I don't think you will see a big change in seasonality second half of last year the second half of this year. The trends will be the same. The numbers maybe the different because it depends on the level of commitment and investment and from where we're sitting, almost every one of the major cloud providers, in terms of numbers, is looking to spend at or higher from last year to this year.

Seasonality patterns aren't varying, but the numbers themselves may vary.

Going back to the gross margin discussion again. I think the best way to think about this is that as you think about how we're transitioning through time, clearly, we are awaiting customs approval and will be awaiting customs approval in that time frame. We will be leveraging our US manufacturing and our inventories that we have built up. Right?

Once we have approval, we will have access to offshore manufacturing and components again and that will be the second case at some point in the future. You should expect that we are going to navigate through this. There will be some volatility, but at the end of the day, nothing has changed in the overall class structure of the product other than what we've talked about or in the competitive environment, right?

So I think about it as being temporary and something that is going to be somewhat volatile as we go through the next number of quarters. Alright I tried to bound this with the 60% and 65%, saying that if we ended up being totally dependent on the US, we would be at the lower end and if we're in some hybrid model which is where we probably will be for a lot of the time, we would be somewhere in the middle of that range. I can't give much more precision to that right now because it's dependent on the timing of some of the legal decisions and so on.

Vijay Bhagavath, Deutsche Bank.

Hi, Jayshree. Solid results here. So one of the things we have been talking to clients in the service provider opportunity, routing opportunity with merchants for the -- I'd like to hear any color you can give us on how that is coming along, and if you would give us that merchant silicon routing development would be disruptive to the service provider routing market.

Thank you, Vijay. Yes, it's -- you're always right on some of these technology and trends and you're absolutely right. We see that the standalone router market is migrating more and more to switching and routing on merchant silicon with a really good software stack. I know this is sitting in the middle of all of those trends.

As I said in my prior call, we are very enthusiastic about both the 7500R Series and the new companion we just announced June, the 7280R, which is the fixed form factor or the Leaf, the value Leaf as we call it. All of these products have been -- are in trials and they are being received well on the routing side. And they are being received very well on the switching side.

The common theme that I would see, we see in both cases is the need to adopt 100 gigabit spines, and the value that's bringing in. While it's easy to look at 100 gigabit spines as a performance metric, it really is a flexible port speed, where you can have [10x10s, 4x25s, 2x50s, 2.5x40s]. So customers are really adopting that as their strategic spine, both for routing and switching and that's very much favoring the trends that we are seeing.

In terms of material effect as I've often said to all of you, we'll be in a lot of trials and a lot of customer acceptance in the second half of this year. We think the real material effect is 2017.

Ryan Hutchinson, Guggenheim.

Good afternoon. So my question is around lead times relative to last quarter. Maybe you could touch on that? And then as part of that, we get a lot of questions around the potential for accelerated purchasing decisions by potentially some of your large customers ahead of this new manufacturing capabilities. So if you could speak to that and in addition, if you witness any challenges securing components; it's all sort of tied together.

That is a loaded part A, B, C, Ryan. You're very creative, but thank you. I think the lead times on our products are higher than the standard leads that I would like to see. A lot of it has to do with the fact that as we just introduced tremendous amount of new product.

The number one reason is some of our suppliers are not able to meet the demand that we have put in front of them, and their lead times reflect our lead times. Our lead times are usually, as you know, in the 4 to 10 weeks, depending on the type of product, but our suppliers' lead times can be as much as 22 weeks.

So we've had to plan. We've had to forecast. We've had to budget. We've had to do inventory management. It is been a non-trivial challenge and particularly, as we've had to do this across multiple manufacturing entities outside of another layer of forecasting to this.

So I would just go on record saying that we could do better and we need to improve our lead times and our on-time shipments is an area of focus for us. I think that everything else after that is really tied to that. Our lead times are affected and our product availability is affected by the component lead times, but also, some the planning we've had to do across multiple contract manufacturers in multiple locations.

Just as a follow-up, though, Jayshree, have you (multiple speakers) -- any accelerated purchasing or request by customers?

We're seeing purchasing tied to demand; we're not seeing any different or unusual behaviors, besides the normal demand related purchase.

I think that we said, Ryan, is we'll call that out clearly if we do see that at some point, but for now --

We're not (multiple speakers) that.

(Operator Instructions)

Kulbinder Garcha, Credit Suisse.

Thanks. Ita, I just want to be clear on one thing on the customs department on the approval; isn't there like a relatively fixed timeline on -- it sounds like --

Kulbinder, we -- you are very soft-spoken. We can't hear anything.

Can you hear me now?

Yes, that is much better. Please can you repeat the question?

So my question is that the customs department approval process, isn't their fixed timeline on right? It sounds like they may not be. I'm just curious as to what their schedule of events is.

And then the other thing that leads to that, this whole the process is a -- in the event that you have to be very US-orientated in manufacturing, couldn't you be redesigning your products to cost in the next generation so you could take out some of that excess cost that you're facing? So gross margins go back up even if you have to do a very heavily reliant US-manufacturing business for some time or is that not in the cards or not possible? Thanks.

I will take the second question first and then I'd like Marc Taxay, our Senior Vice President and Lead General Counsel to comment on the customs process.

I think in terms of cost reductions, there always many components to that both in new product and in more mature products at high volume, Arista is aggressively taking out costs. But we're also aggressively bringing volume-driven pricing. So I think the two kind of balance each other, whether it's new product or not and it has less to do with US manufacturing it's just got to do with being competitive.

The things that can contribute to higher margin is if we start to increase the software content, things like CloudVision, FlexRoute, and software licenses could become greater components and are relatively small components now. Also, if the mix of our product changes to more Spine and less Leaf or more software-driven Leaf, that could also help. So I think product mix can have a -- and customer types have a greater contribution than purely US-driven optimization. Hope that answers your question.

Kulbinder, the important thing is while obviously, we will have design rounds and we will have reroutes -- we will work to get those design rounds approved, and in that scenario, we will have access to domestic and overseas components, et cetera, again. So this is very much a temporary navigation as we go through this time period as the cases are heard and resolved. But the goal obviously is to design -- have designed around a product that will allow us to going access overseas, offshore component, et cetera, again, right?

Mark, can you comment on the customs for us?

Sure. So our customs does not provide for a specific time frame to get customs approval; however, we've been actively working with customs and they have been working very judiciously -- expeditiously to get the customs approval done as quickly as possible.

Erik Suppiger, JMP.

Congrats on a good quarter. On the inventory front, I was thinking you might see a bigger increase in the current quarter. Can you talk a little bit about how you are planning the inventory that you're going to build up? What kind of time frame are you going to be satisfying with what you are going to currently retain at the end of the September quarter? And how are you planning to add to that as to a timeline progresses?

I think what I said in my remarks was that we would expect to grow the inventory bounds at the end of Q2 by another $100 million by the end of Q3, right? So there is still a significant amount of inventory that we will add over the coming months to two months, right? So they got --

But you're right, Erik. We haven't added enough, but we have added some. It's a strong function of getting the inventory we want. We don't want to add the wrong inventory. We did make progress there in Q2 and we need to make more progress in Q3 and Q4.

But I think that is pretty much committed from a cost perspective as we look into Q3 so you should see a bigger uptick in Q3.

Will you be able to dial down your contract manufacturer once you get approval until the next lawsuit comes up or do you have to sustain certain volume levels to keep the contract manufacturing going?

I think the way to think about this is that the manufacturing in and of itself is not a big overhang from a cost perspective; it's really more the supply chain. We will maintain, obviously, a volume there that keeps that entity and we will -- it's the supply chain that will dial up and down.

(Operator Instructions)

Alex Henderson, Needham.

There's -- I heed to the hill of saying one question, it is kind of two related components to this one. One, your transition into 25 gigabit, 100 gigabit architecture, there's a lot of those components are heavily supply constrained as the manufacturers are just starting to ramp those production lines.

So does that create a situation where you have an inability to build enough of those components for the next-generation architectures and really, that plays into the second part of the question, which is really the most important piece. To the extent that you have a couple of months window before you are constrained from buying supplies internationally, and you can't buy any more supplies internationally until after the determination of that your workarounds are legitimate.

If the second tranche of patents is then put into play before you get through that window, in other words, the second tranche of patents gets this preliminary determination in, say, August and then final determination in December, and you haven't come out of the workaround on the first one, and don't get another window to build inventory. Will that mean that you would have enough capacity on the inventory that you have built to work through all the way to, say, July when that final determination on the second round of patents is determined? If these two overlap, can you get out to July without further damage to your business?

I think, Alex, it's really hard to speculate on exactly what's going to happen here. We've put a lot of thought and effort into multiple different strategies depending on what the outcomes are. Right? Some of those are inventory related, some of those are other plans, right, but we've been making plans for contingencies. I'm not going to guarantee anything, right, here but we've been very cognizant of the various outcomes and we've make plans for those. Marc, I don't know if you want to add anything to that?

The only thing I would add, I guess I would say is that, first, is that would include changes to our supply chain in order to be able to source in compliance with the ITC orders and the fact is that, I think Ita mentioned this earlier, is that ultimately, if we're manufacturing products here in the United States, any components whether they are sourced here or internationally can be used in those products. Again, as long as those products are not infringing.

I think that's the really important point, we get caught up on product infringement, but remember, the ITC found us to infringe two features out of thousands of features we have. So with the US4.16 that we've been diligently working on building a non-infringed product with two features, which is 0.2%. We've really designed an ultimate product. And therefore, we can build non-infringed products worldwide.

So the last -- the piece that you didn't answer was does supply availability of the 25 gigabit, 100 gigabit components given they're in early ramp?

They are, Alex, but we feel pretty good that over Q2, Q3 and Q4, we can address the lead time and component availability issue.

James Faucette, Morgan Stanley.

Just a follow-up to that and some of the other questions. When you talk -- look at the availability and lead times, how much of an improvement are you anticipating in building into your outlook for the September quarter and along those same lines, is availability at all hampering your opportunity to get test units of the new 7500 into service providers, et cetera, and has that changed your outlook for initial qualifications, et cetera.

Thanks, James. So first of all, availability has not changed our outlook and interest in service providers. We're very good about providing the requisite early few trials or seed units or testing units. So we haven't seen any change there.

What was your first part of the question?

I was just asking, in your outlook for the third quarter, and then what your -- the control you're giving for the rest of the year. Is -- how much improvement and availability are you building into that?

No, I think we will see -- we will improve more in Q4 than we will in Q3. We might see some improvement in the tail end of Q3, but I think we're more likely to see it in the second half of the second half, if you will.

Steve Milunovich, UBS.

Regarding switching Cisco's reported fairly strong data center orders last quarter, Juniper's release the QFX10000 spine switch, are you seeing any change in the competitive environment or pricing as a result of this?

The short answer, Steve, is no.

What is the long answer?

No, twice. (laughter) No, kidding aside, I think that what was said to Steve and I will reiterate that. Our competitive landscape has always been extremely strong and dynamic, so -- because it is in such high fury and pace all the time, it is difficult to know any specific change because it is always a constant competitive battle.

Mitch Steves, RBC Capital Markets.

Great quarter. I just had a real quick one just in terms of the gross margin comments and all that. So if I were look out over the next, call it, let's call it five quarters, is there reason why EPS wouldn't essentially track the rate of growth of revenue? So what I mean by that is if you have any leverage to grow EPS faster than revenue or is it going to essentially be in line, just given that gross margins may come down?

I'm not going to try to go outside the quarters at this point, Mitch, honestly, right? I think we have talked about an overall business model in the past and I'd refer you back to that and I think we're not try to go out that far at this point.

I think the important thing to remember about Arista is we're still very much in investment mode and very much in growth mode. so Optimizing earnings per share perfectly is an end goal, but obviously, we're doing very well in earnings per share growth so we will continue to deliver to our shareholders.

But if we see investments, we're going to aggressively invest there, and you can see our R&D is still up ticking, and we want to be able to do that because we never want to be at a point where we tell our customers, sorry, we're not going to invest in here because we need to hold on until a certain earnings-per-share number.

Jeff Kvaal, Nomura.

Well, in the commentary from many of your web scale customers, they started to talk a little bit more perhaps than they have in the past about efficiency gains when they're spending on their networks. So I'm wondering if you can help us interpret what that may mean when it comes to how they are working with you, whether it is pricing pressure or more custom products or what have you? That would be helpful. Thank you.

That's a very good question, Jeff, and I think it's deeper than I may answer entirely on this call. I think one of the things that hasn't been sufficiently appreciated when Arista is with these large cloud titans is not only do we have a great product, but the level of effort we put into system upgrades, high availability, quality, automation, provisioning, programmability, what many of us call SDN telemetry, and two, has been what we've been doing not just for one or two quarters but for several years here.

Our customers have their own in-house developed tools, but we're working with them has been a very strong part of our initiative. And having that control for them is very, very important. So while it may be our product, we'll often work together jointly like it's their technology as well and make sure that, that their 20 automation tools and ours can work together.

So there's been a number of engagements that are not specific products on our side but specific aspects of our EOS at work better with them. I've alluded to some of those examples, but we continue to see that as an important differentiator in our partnership with these cloud providers.

How does that tailor to their efficiency, Jayshree, as opposed to just allowing you to win because you've got a better product than the person down the street?

A good example would be to look at a simple function like network rollback. Today, the way you do that is you would literally go switch by switch, port by port, and issue a command to each one and you're basically polling and checking the health of every device.

By virtue of what we're doing with them, they can automatic that process across their entire 100,000 servers or switches. They do some of it and we do some of it. CloudVision is our enterprise turnkey version but we implement CloudVision-like features with them, offering there DevOps tools so that's a tremendous efficiency and the number of people they need to manage their network, which is a lot fewer.

Okay. We will take this offline. Great, thank you.

Simon Leopold, Raymond James.

I'm going to ask one question with several parts. Actually I'm just kidding. I wanted to get a longer-term question in here.

Specifically, with the launch of the R-designated products, the routing products, I wanted to see if you could talk a little bit about your aspirations for how much of the market you take? And maybe if you could review your go-to-market strategy in that I think at the initial launch, you talked about targeting the web titans as sort of the first set of customers. If we could get a better understanding of how the go-to-market develops for the routing and then where you see the trajectory as in terms of material contributions to sales and market share in the sector? Thank you.

Sure. As you know, our primary market is data center switching. This is a $10 billion TAM. We're now, this quarter, approaching the first $1 billion run rate and we still see a lot of room in data center switching, so I want to be clear that while we're very enthusiastic on routing, that our job is far from done in switching.

But routing and switching are coming together. We don't see that they are distinctly different markets in the data center, and the core router in the data center is really lacking in dynamic functionality, port density, programmability, pricing, you name it.

As I said before, the first customers are likely to be our existing customers and we are already starting to see trials and traction with them. But they are also seeing trials and traction with new service providers, Tier 2 service providers, and some of the enterprise customers as well who use -- who have traditionally used core routers because we're really providing a disruptive price performance and functionality.

So if you ask me how I define success in this, and as I said before, I think 2016 is the year of trials, and 2017 and 2018 are the year of revenue. I would like to see several tens of customers, maybe even 100 customers this year adopt the R Series and that would be a measure of success.

Paul Silverstein, Cowen and Company.

If you've already answered any of the following, I apologize in advance, but some very quick ones. What did linearity look like in the quarter?

Nice. Pretty good, but we didn't answer that question, so no apology needed. Linearity was good through the quarter.

Well, I'm not done yet, Jayshree, so --

((laughter) Thanks, Paul, you get the prize for sticking to one question.

No, no, no, I'm not done, pricing. (laughter) Really quick, what did pricing look like?

Pricing was normal and competitive.

All right. Did you have any revenue from routing?

That was three questions. That's one too many, Paul.

Yes, but they are one-word answers. Any revenue --

The short and long answer is yes.

All right. Finally, was the cloud titan revenue growth rate greater than the rest -- I think I heard you already say this, but was it greater than the rest of your customer base's growth rate, as it has been for some time?

Our cloud titan contribution in dollars, I won't comment on growth rate until the end of the year whereas definitely greater than the other three verticals.

(Operator Instructions)

Hendi Susanto, Gabelli & Company.

I will try my best to get that one single question award.

(laughter) Thanks Hendi. That's the thing. I don't believe it any more but go ahead.

The latest Gartner reports that Arista's coverage is still limited outside of North America and Western Europe. I would like to understand more about the international markets. How should we view the progression, and the timeline, and whether use cases in the international market is similar or different than North America and Western Europe?

That is an excellent question, Hendi, and you do get the award if you don't ask me a second question for one question. I would say our investments internationally is similar to what we did in the US three to four years ago; we're just starting. And Mark Foss heads up a lot of our international region. Mark, do you add a few words to that?

Yes, sure. I think what I'd say is in Q2, the EMEA region represented 16% of the total revenue, 9% came from Asia-Pacific, and 1% came from other Americas. The Arista products are actually installed in over 70 countries globally and we're making good progress with new organic international customer acquisition, although we do benefit from global customer international build outs.

Our regional readership is working, using a strategy that mirrors what we did in the US three years ago. But since most of the cloud titans are domestic, we're beginning to have success in regional web and hosting providers, leveraging our land and expand model.

Simona Jankowski, Goldman Sachs.

Was there any change in the mix of business in the quarter between new customers and existing customers?

Thanks Simona. No particular change; existing customers suddenly drive our contribution and they are a bit -- new customers tend to start out small so there are always, in dollars, smaller contributors, but no major change.

Maybe just I'll ask it a slightly different way. Has there been any change in the pace of addition of new customers, in particular, since the ITC ruling?

I don't think I have seen any specific tied to the ITC ruling on new customer acquisitions. But I think because our data center purchases tend to be large, we are favoring not just acquiring new customers but acquiring million-dollar customers. So our metric is not just new customers but big customers.

George Notter, Jefferies.

I guess I wanted to ask about the three selling models that you guys talked to, I think on last quarter's earnings call. I know that the cloud-scale model allows you to sell hardware and software independently and I think you guys were letting your customers steer you in that regard. But are you seeing sales in that model where you're separating hardware and software right now?

Thanks, George. It's a good question. I think our strategy is definitely ahead of its execution, if that was your question. We are -- we have always been prepared to sell software standalone models, but it takes preparation not just from us, but also from the customer.

I would say even though we were ready last year and we've been discussing it a lot this year, we expect to see more execution of it in the second half of 2016 and 2017. Because there has been a level of preparation required by our customer. But we don't expect -- we have 4,000 customers. We would be surprised if we saw more than single-digit interest in customers, but they will be important names.

All right. This concludes the Arista Q2 2016 earnings call. I also want to mention that we have posted the presentation which provides additional perspective on our second quarter FY16 results, which you can access on the Investors section of our website. Thank you to everyone for joining us today.

Thank you for joining, ladies and gentlemen. This concludes today's call. You may now disconnect.