Attached files
| file | filename |
|---|---|
| 8-K - FORM 8-K - APPLIED MICRO CIRCUITS CORP | d620061d8k.htm |
| EX-99.1 - EX-99.1 - APPLIED MICRO CIRCUITS CORP | d620061dex991.htm |
Exhibit 99.2
Applied Micro Circuits’ CEO Discusses FQ2 2014 Results - Earnings Call Transcript
CORPORATE PARTICIPANTS
Michael Major – VP, Corporate Marketing and Human Resources
Paramesh Gopi – President & CEO
Douglas T. Ahrens – VP & CFO
Shiva K. Natarajan – VP, Corporate Controller and former Interim CFO
CONFERENCE CALL PARTICIPANTS
Ambrish Srivastava – BMO Capital Markets Analyst
Christopher Longiaru – Sidoti & Company Analyst
Hans Mosesmann – Raymond James Analyst
Jason Rechel – Oppenheimer and Company Analyst
Christopher Rolland – FBR Capital Markets & Co. Analyst
Patrick Wang – Evercore Partners Analyst
Krishna Shankar – ROTH Capital Analyst
Vijay Rakesh – Sterne Agee Analyst
Suji DeSilva – Topeka Capital Markets Inc. Analyst
Brian Thonn – Kingdom Ridge Capital Analyst
PRESENTATION
Operator
Good day, ladies and gentlemen, welcome to the Second Quarter Fiscal 2014 Earnings Call. My name is Tony and I’ll be your operator for today. At this time all participants are in listen-only mode. Later we will conduct a question-and-answer session. (Operator instructions) As a reminder this conference is being recorded for replay purposes.
I would now like to turn the conference over to, Mr. Mike Major, Vice President of Corporate Communications. Please proceed.
1
Michael Major - VP, Corporate Marketing and Human Resources
Good afternoon, everyone, and thank you for joining today’s conference call. On the call with me is Dr. Paramesh Gopi, our President and CEO; Doug Ahrens, our Vice President and CFO; and Shiva Natarajan, our Vice President and former Interim CFO.
Before we begin I would like to remind you that various remarks that we make on this call, including those about our future financial results, including revenues, gross margins, operating expenses, design wins, product plans, our competitive situation, market trends, statements about future developments and adoption of our X-Gene™ and other products and our anticipated growth and profitability, all constitute forward-looking statements for the purpose of the “safe harbor” provisions under the Private Securities Litigation Reform Act.
These forward-looking statements, and all other statements that may be made on this call that are not historical facts, are subject to a number of risks and uncertainties that may cause actual results to differ materially. We refer you to our most recent Form 10-K and Form 10-Q filed with the SEC, in particular to the section entitled “Risk Factors,” factors and to other reports that we may file from time to time with the SEC. For additional information on factors that could cause actual results to differ materially from our current expectations. These forward-looking statements speak only as of the date hereof and we disclaim any obligation to update them.
I want to point out that Applied Micro has several analysts that cover our stock, and this creates a range of variability relative to the Street financial models. When we say “Street estimates,” we mean the consensus of the major analyst models and not necessarily the guidance that was given by the company.
With that, I’m going to turn the call over to Paramesh. Paramesh?
Paramesh Gopi - President & CEO
Thanks, Mike. Before we start I’d like to introduce Doug Ahrens, our new Chief Financial Officer, now in his second week with Applied Micro. As we previously announced Doug comes to us with rich experience from Maxim Integrated and Intel® Corporation, and I’m delighted to welcome Doug to my executive leadership team. Welcome, Doug.
Douglas T. Ahrens - VP & CFO
Thanks, Paramesh. I’m very excited about joining the company. I was impressed with how Applied Micro defined a new category for the cloud using high performance ARM®-based servers. I was also impressed with how well the team executed over the last three years to take the clear leadership position with its X-Gene Server on a Chip™. I’m thrilled to be here right now. Now over to Paramesh.
Paramesh Gopi - President & CEO
Thanks, Doug. I would also like to take this opportunity to thank Shiva for the fine work he has done for the company as Interim CFO. Thanks, Shiva. Shiva will address financial results for our September quarter and help answer finance related questions during the upcoming Q&A. Doug Ahrens will take on these responsibilities on future earnings calls.
We had a solid September quarter with revenues of $55.4 million. Non-GAAP EPS was $0.03 and in line with our guidance. Excluding our R&D investment for our pre-revenue X-Gene development, our non-GAAP pro forma EPS would have been $0.23 per share. Our backlog as of the date of this call is greater than 70% for the December quarter, and book-to-bill ratio is slightly under one. We ended the quarter with $76 million of cash on the balance sheet with no debt.
On the X-Gene front, we continued to march towards commercialization of X-Gene based server platforms and we’re very excited to demonstrate further progress next week at ARM TechCon in Santa Clara. I’ll cover X-Gene in more detail in a few minutes.
2
I will now discuss our product line starting with the base business. Our Embedded Computing revenues were $24.8 million for the September quarter. Substantial revenue growth from our new products largely offset declining revenues from legacy products. In particular, embedded processor new product revenues increased 87% over the prior quarter and were driven by our key data center and enterprise printing customers.
We also continued to see strong design-win traction for our embedded processors fueled in part by adoption of 802.11ac WiFi access points. We believe our current Embedded Computing products will continue to generate solid revenue growth for us in the next 18 months to 24 months as we transition to X-Gene-based embedded products.
Our Connectivity revenues were $30.6 million, up 12% over the prior quarter. Steady data center growth for 10/ 40 and 100 gig converged Ethernet / OTN connectivity, as well as upgrades in the China optical networking service provider infrastructure, contributed to this increase.
Now to X-Gene. Having completed another quarter of solid progress, we’re now closer than ever to commercial deployment. We are excited to report that we’ve shipped meaningful quantities of pilot production parts to multiple tier-1 customers. This is tangible proof that we are starting to break into the $7 billion market expansion opportunity via our tier-1 customer engagements.
As a reminder, last quarter we talked about our customers’ five step server system development cycle. Number one, silicon evaluation; number two, system design; number three, system validation; number four, pilot production; and number five, commercial deployment.
Since the last call we have multiple customers that have moved into step four, the pilot production phase, after completing grueling systems testing and validation. At the top of the pipeline, more than 30 companies have signed our X-Gene limited evaluation and used license agreements. This is significant because it clearly signals customer and ecosystem partner financial commitment towards X-Gene commercialization. We expect to double the number of X-Gene license agreements in effect during the December quarter.
I’m pleased to announce that we’ve entered into a formal arrangement with a tier-1 software partner to produce the world’s first commercial grade enterprise distribution for ARM 64-bit servers. X-Gene is now poised to drive hyper- scale and cloud data center deployments within the next two quarters, with more engagements on the horizon.
Last quarter I talked about X-Gene’s differentiated features, including:
| • | a unique grounds-up high performance ARM 64-bit core, operating at up to 2.4 gigahertz specifically designed for Cloud Servers™; |
| • | enterprise class features including ECC and RAS; |
| • | highly integrated mixed signal I/O including 40 gigabits per second of Ethernet, generation-3 SATA; Generation-3 PCI Express and various workload accelerators; and |
| • | a high-performance on-chip fabric with software defined networking capability. |
These features enable our customers to provide a truly disruptive system platform to the end server market.
We’re excited to share a very important benchmarking result on this front. To provide context, most of today’s hyper scale and Cloud Servers use Kernel-based Virtual Machine, or KVM — the open source equivalent of VMware. Our customers have now shared with us that X-Gene exhibits comparable performance to Xeon® X86 based Cloud Server platforms, at a fraction of the TCO. This proves X-Gene is a very compelling value proposition for virtualized cloud applications.
While our first generation X-Gene product moves into production, I’m pleased to inform you that we’re making excellent progress on our second generation product and expect to have X-Gene 2 – our 28 nanometer solution – sampling in customer platforms in spring of 2014. X-Gene 2 provides a seamless migration path to our existing customers to extend their X-Gene 1 platforms into massive scale out server systems. In the context of massive scale out data center platforms, connectivity and I/O become critical.
3
This brings me to X-Weave™, our exciting data center connectivity product line. APM has had a proven track record of shipping high performance connectivity mixed signal products into the service provider space for over a decade. We shipped more than 400,000 converged OTN Ethernet mixed signal 10 gig ports during our past two fiscal quarters. This was primarily driven by our strategic move to combine leading mixed signal analog IO along with high speed DSP technology — thereby letting our customers offer the highest density, lowest TCO blades for data centers and service provider central offices.
X-Weave represents the evolution of our base connectivity business to serve an expanded data center TAM and adds a billion dollar market expansion opportunity to our base service provider connectivity business.
To contextualize, X-Weave customers are looking for the same key benefits for their connectivity solutions as they did for their server platforms. These are high performance; high density; low TCO – with server node uplinks moving to 10 gigabits per second, server cluster connections moving to 100 gigabit per second, and connections between data centers moving to multi terabit rates. Customers require low latency, carrier class reliability using converged OTN and Ethernet, providing for a real alternative to InfiniBand.
X-Weave was conceived to address all of these changing requirements at the lowest TCO. I’m very pleased to report to you that we’re now firmly anchored with tier-1 customers on X-Weave, and have secured three high profile tier-1 design wins for X12, the first member of the X-Weave family. We won because of our ability to offer unparalleled density and TCO for both optical and copper connectivity both within and between data centers.
With that let me turn the call over to Shiva. Shiva?
Shiva K. Natarajan - VP, Corporate Controller
Thanks, Paramesh.
September quarter revenues were $55.4 million and were slightly better than our guidance. Sales to North America accounted for approximately 44% of total revenue, sales to Europe contributed 21%, and sales to Asia contributed 35%. We have three customers that accounted for more than 10% of our business: global logistics support vendor, Wintec, and distributors Arrow and World Wide Avnet. Distributor revenues for the September quarter were approximately $28.2 million. Channel inventory days excluding non-cancelable, non-returnable orders were 57 days, down from 66 days the prior quarter.
Turning to the P&L, our non-GAAP net income for the quarter was $2 million, or $0.03 per share, compared to the non-GAAP net income of $1.4 million, or $0.02 per share, for the prior quarter. Our non-GAAP operating margin was a positive 4.7% of revenue and improved 3 points from the positive 1.7% achieved in the last quarter. The second quarter non-GAAP gross margin was 61.6%, which is in line with our guidance of 61.5% plus or minus 0.5%.
Looking forward to the December quarter, we except total revenues to be in the range of $53 million to $57 million and we expect gross margins to be approximately 60.5% plus or minus 0.5%. Second quarter non-GAAP operating expenses were $31.5 million, which was at the lower end of our guidance of $31.5 million to $32.5 million. This was at the lower end of our guidance due to the timing of certain expenses and tight expense management. For the December quarter we expect our non-GAAP operating expenses to be in the range of $31 million to $32 million.
Our non-GAAP interest and other income was a negative $400,000. This was negative as we liquidated some securities in our portfolio and realized losses. Interest and other income is expected to be approximately $300,000 for the December quarter. Our non-GAAP taxes were approximately $300,000 and we expect to see this continue at this level for the next several quarters.
4
The share count for diluted EPS purposes was 74.3 million shares. Looking forward to the December quarter we expect the share count to be approximately 75 million to 77 million shares. The increase is primarily due to the timing of the grant of shares related to Veloce post-closing earn-out milestones and the impact of stock option exercises in the September quarter.
Turning to the balance sheet, our cash and investments totaled $76 million, or approximately $1.03 per share, at the end of the September quarter, a decrease of approximately $13.8 million from the June quarter. Excluding the cash paid to Veloce during the September quarter, the cash flow from operations was a positive $1.7 million. The closing cash balance was better than our anticipated balance of approximately $70 million primarily due to proceeds from stock option exercises during the September quarter given our higher stock price.
Our inventory turns for the September quarter were 6.7x – in line with the 6.9x that we had for the June quarter.
Now turning to Veloce payments, during the September quarter we paid out a total of $38.3 million in post closing merger consideration, approximately $19 million in cash and another $19 million in common stock. Cumulatively to-date we have paid a total of $106.4 million out of an assumed maximum merger consideration of $178.5 million. During the December quarter we expect to pay a total of approximately $1.2 million. Assuming no other changes to our cash balance, this would leave our cash balance at the end of the December at an estimated $75 million.
To summarize, through December, we expect to have paid a total of approximately $108 million in Veloce consideration.
Now turning to GAAP, our non-GAAP financials exclude certain items required by GAAP. Our net loss on a GAAP basis was $32.4 million, or ($0.45) per share versus net income of $10.9 million or $0.16 per share last quarter. A complete reconciliation between the GAAP and non-GAAP financials is set forth in our second quarter earnings release which can be found in the Investor Relations section of our website. Please note that there is no reconciliation relating to forward-looking non-GAAP measures.
That concludes my remarks, let me turn the call back over to Paramesh. Paramesh?
Paramesh Gopi - President & CEO
Thanks, Shiva. As you all know we have been – we have shipped hundreds of real hardware server platforms to our customers and partners over the past few quarters. These platforms include real X-Gene server hardware with a full suite of software including BIOS, operating systems and application development tools. We are extremely excited to announce that we are in the process of making these platforms generally available to a large populace of developers and enterprises. Stay tuned for more information at next week’s ARM TechCon, ARM’s technology conference.
With that let me turn the call back over to Shiva. Shiva?
Shiva K. Natarajan - VP, Corporate Controller and former Interim CFO
Thank you, Paramesh. Just before going to Q&A, let me address our guidance for the December quarter. Total revenues are expected to be in the range of $53 million to $57 million. Total gross margin is expected to be 60.5% plus or minus half a point. Operating expenses roughly $31 million to $32 million. Interest income and taxes each at $300,000.
Now let me turn the call back to Mike. Mike?
Michael Major - VP, Corporate Marketing and Human Resources
This concludes our formal remarks. Tony, please provide instructions to our listeners for the queuing process.
5
Question-and-Answer Session
Operator
(Operator Instructions) Your first question comes from the line of Ambrish Srivastava, please proceed.
Ambrish Srivastava - BMO Capital Markets
Thank you for taking my call. Paramesh, I just have a bunch of questions here, but I will restrict to just a few. Does the December quarter include any X-Gene revenues?
Paramesh Gopi - President & CEO
Go ahead, Shiva, you can take that.
Shiva K. Natarajan - VP, Corporate Controller and former Interim CFO
December doesn’t include any X-Gene revenues, as Paramesh mentioned in his prepared remarks, they’re pilot shipments that are happening and we are selling – we are shipping those for – it’s not free; we are shipping those for dollars, but they are not revenue.
Paramesh Gopi - President & CEO
We don’t consider – just to be clear, we don’t consider pilot shipments to be meaningful revenues and we wouldn’t recognize them as such.
Ambrish Srivastava - BMO Capital Markets
Got it. I just wanted to clarify. And then Shiva, back to you maybe, can you please just walk us through the OpEx and gross margins for the rest of the fiscal year? And then, also how should we be thinking about OpEx for next year? It sounds like X-Gene 2 is a little bit ahead of what you guys were planning for, I thought it was calendar Q3- Q4, but now it seems like it’s coming out Q2 and then I had one final follow up for you Paramesh: I am not sure I understood the metrics that you were alluding to when you said, take this as a measure of financial commitment from our customers, then you have 30, and then you said you were going to double that in the December quarter. If you can please expand on that a little bit please? Thank you.
Paramesh Gopi - President & CEO
Sure. Why don’t I let Shiva complete the first part of the question then I can give you color on the second.
Shiva K. Natarajan - VP, Corporate Controller and former Interim CFO
Yeah. So it maybe a little early to give you OpEx numbers for next fiscal year Ambrish, but I can tell you for next quarter, as we said in the prepared remarks, that will be about $31.5 million which is the same rate as we saw for Q2 here. And for Q4 again we don’t want to give guidance here but you should – you could expect – I don’t see any meaningful or huge significant changes to that run rate.
Ambrish Srivastava - BMO Capital Markets
It sounds like gross margin is trending a little bit higher. So for March, should we be thinking in the 60% range or should it come down?
6
Shiva K. Natarajan - VP, Corporate Controller and former Interim CFO
Again, it’s going to depend on the mix and I don’t want to preempt anything at this point. But you know, you should, for now you should assume 60% range.
Ambrish Srivastava - BMO Capital Markets
Okay, thank you.
Paramesh Gopi - President & CEO
So Ambrish, let me answer to those two questions. Let me start with the last question first. Whenever we have over the last six months roughly, we have had a lot of customer engagements where customers essentially either get platforms from us, get chips from us and/or devote resources to building platforms with us. Given the fact that we are on the edge of actually making a category happen in the server market, one of the things that we have instituted is a very, very strict legal process by which we bind our customer and/or our partner to a set of legal conditions that involve indemnity and financial relative to taking on an X-Gene platform and evaluating it or developing with it. So what we decided to do this quarter was to – we’ve talked about the customer pipeline last quarter, we’ve talked about the types of customers and the types of ecosystem partners, just to give you the rough scale of the types of companies and the number of engagements that we are today dealing with: the 30, what we call, end user license and use agreements are encompassing in that they’re only signed when they are significant financial commitment from a resource perspective and a hardware design perspective from the other side, whether it’s a customer or a partner.
So that shows you the makings of a nice funnel for our business. That is the proxy, and that was the point of talking about that. Let me also now talk to you about – you asked another question, can you repeat that I forgot.
Ambrish Srivastava - BMO Capital Markets
Now I think Paramesh that was – I just wanted to get a sense for and you said this would double in December. So I just wanted to clarify what it meant? I think that helps.
Paramesh Gopi - President & CEO
Okay, thank you.
Operator
You next question comes from the line of Mr. Christopher Longiaru of Sidoti & Company. Please proceed.
Christopher Longiaru - Sidoti & Company
Hello! Can you hear me?
Paramesh Gopi - President & CEO
Yes.
Christopher Longiaru - Sidoti & Company
Okay. So my question has to do with – you said you had three major X-Weave design wins. I was wondering if you could give us an idea of how those progress and what the timing on those is and what are some milestones that we can kind of look forward with respect to X-Weave?
7
Paramesh Gopi - President & CEO
So we announced X-Weave as a product family about two and a half months ago and when we – and X-Weave had been in development for over two years to-date — and we will start to sample first X-Weave parts in the spring of next year, and we will start to see meaningful platforms that get designed, that get shipped to trials with X-Weave probably towards the end of next calendar year. So X-Weave, by all accounts right now, you will start to see revenue end of our fiscal ‘15 and end of calendar ‘14 towards beginning of calendar ‘15.
Christopher Longiaru - Sidoti & Company
And then, just in terms of your progression here with X-Gene 2, is there any significant a ramp in R&D associated with that or is there any tape-out cost associated with that in the next 12 months that we should be modeling?
Paramesh Gopi - President & CEO
I can tell you one thing, all of the development costs have been contemplated into the current consensus model and all of the tape-out and/or that includes both the tape-out cost as well as the labor cost as well as the technology, inherent technology tool cost, are all in; that we expect no surprises to the expense profile going forward.
Christopher Longiaru - Sidoti & Company
Right and then just one last question, just in terms of your current business outside of X-Gene, ah, there’s been a big ramp in just OTN sales in general, can you make any comment – just some visibility there and how you expect that to progress over the next couple of quarters?
Paramesh Gopi - President & CEO
Yeah, I want to start by saying that most of our OTN sales in the last two quarters have come as a result of data center connectivity platforms, not necessarily service provider connectivity platforms. And the biggest difference between those two being the platform, such as the Cisco ASR box or the ALU 1800 box, they are very different. The router and packet optical platforms are usually the higher growth enterprise class data center platforms that are data center oriented, while the telecom platforms are more service provider oriented. I think we have always maintained that our business mix has changed dramatically in the past three years, we don’t consider OTN as a singular technology anymore, it’s basically combined at 10, 40 and 100 with Ethernet. So, I want to be very clear again, we believe that the data center vector is the growth vector for us, and in the service provider vector, over time we will absolutely be the recipient of all the tailwinds in that sector, but it really, if you were to depend on service provider we wouldn’t be able to perform the way we are today.
Christopher Longiaru - Sidoti & Company
That’s really helpful, thank you guys, I’ll jump back.
Operator
The next question comes from the line of Mr. Hans Mosesmann of Raymond James, please proceed.
Hans Mosesmann - Raymond James
Thanks. Good show there on the X-Gene progress, hey Paramesh, can you provide some more clarity or granularity on what you described in terms of performance or the benchmarking aspect of KVM and Xeon, I didn’t quite catch that you’re equivalent in performance – what Xeon class you’re talking about and where can we get this benchmarking information?
8
Paramesh Gopi - President & CEO
It’s essentially, I wanted to illustrate two things with my comment: One is that the customer base has now moved from – treating this as a, I call it a research category into more of a development and deployment category which means that when you have a – most of our customers whether it’s in the cloud or in the public cloud or the private cloud in the enterprise, run virtualized machines. Part of what we look at and they are all using either E5s or E3s just to be very clear. And these are 2P E3s or E5, 2P E5s or E3 platforms, dense platforms. And what we look at is Kernel Virtual Machine, which is the open source equivalent of VMware, because that’s what’s used in the cloud today and over 25% of all servers shipped run KVM today and they are all Xeon in the cloud, essentially looked at – essentially what is the virtual machine ah, running metrics, for instance they would look at things like utilization, they would look at things like network bandwidth. And with all of the benchmarking that’s been done by our customers and their deployment targets, they have concluded that they see no, in fact, they see no difference between our implementation of an X-Gene based Cloud Server and an E5 based server, which means that we are able to get E5 class performance at 50% to 70% of the TCO that they talk about from the virtual machine perspective. Again, remember there are factors like a 10 gigabit pipe. There are factors like large memory bandwidth and channels. So those are the factors that are used to make sure that you can do an apples-to-apples comparison from a virtual machine per gigabit per TCO.
Hans Mosesmann - Raymond James
Okay that’s helpful. And so is that information published some place or is there a white paper where we could find that information, or is it more anecdotal at this point?
Paramesh Gopi - President & CEO
We actually have hard metrics that our customers have provided us so that’s why we said this on the call today. We wanted to make sure that there is at least some color that we can provide in terms of the class of processor, the class of platforms that we have been able to build and pilot ship. Thanks.
Hans Mosesmann - Raymond James
Okay. And then as I have two follow ups. The three designs that you have with X-Weave, does that mean – are those correlated to X-Gene wins as well or are they independent?
Paramesh Gopi - President & CEO
Really, really good question. Wow. No, they are completely independent of X-Gene and I can give you some more color on those. If you look at our OTN port shipments, the reason we are able to enjoy a lion’s share of the OTN market – or I will call it the data center 10 gig and 40 gig and 100 gig port market — is because we did not confine ourselves to small low density telecom type devices. So what we have done is, we have enabled folks like the data center providers, who are like at Amazon and Facebook and Cisco and a bunch of other folks, to be able to supersize the 10 gig density in the rack, right? So what X-Weave does is, it essentially becomes the front-end to a packet optical router, to an optical switch or to a really large aggregation router platform. So to answer your question directly, none of the three anchor customers that we have landed and have gotten design wins with, are shipping today, a server. They are all – they are shipping complete full data center aggregation either switch or router platforms.
Hans Mosesmann - Raymond James
Okay. Okay that’s very helpful. And then the last question Paramesh, what is the view in general here in terms of carrier spend over the past quarter, has the visibility changed and if so in what way?
9
Paramesh Gopi - President & CEO
I can tell you for a fact that, you know, we don’t see the – we don’t see the large – we don’t see a large kind of wave of carrier spend. From our perspective, it’s more lumpy. I would characterize it as lumpy. You get spurts. I can tell you for a fact that our last quarter was heavily influenced, at least, I would say, maybe some small double digit percentage, maybe 20% of our growth on the OTN side was influenced by folks like ZTE building out key wire line optical aggregation in the router devices that terminate both 10 gig and 40 gig OTN over Ethernet. But I can tell you that if we were to just – I will draw another analogy here – if this was the AMCC of three years ago, we wouldn’t be able to produce the type of results that we did last quarter, and we wouldn’t be able to tell you what we were going to do – we wouldn’t be able to tell you with confidence what we were talking about next quarter. So because we are not dependent on the service provider ebb and flow, it’s purely opportunistic advantage for us.
Hans Mosesmann - Raymond James
Okay, very well. Very helpful. Thanks again.
Operator
The next question comes from the line of Mr. Rick Schafer of Oppenheimer, please proceed.
Jason Rechel - Oppenheimer and Company
Yeah, hey guys this is Jason Rechel calling in for Rick. I guess just looking ahead to margin and when we should expect initial X-Gene revenues. Do you guys have, have you gotten any type of visibility from your customers in terms of what the actual revenue or ramp is going to look like with X-Gene into next year? And I know we have kind of seen a pretty wide variety or wide range of revenue estimates for X-Gene next year and just wondering if you have any ability to quantify what it may look like at this point? Thanks.
Paramesh Gopi - President & CEO
So let me take that in two steps. First of all, I think the ARM server category establishment is clearly underway. And I think by all factual data we pretty much lead the charge. I think the one thing that we have talked about in the past is talking about a revenue funnel and the funnel has to do with people who own their own data centers are one category of the funnel. People like the Dells and HPs are the other category of the funnel and the third are people are networking OEMs who build blades and servers into networking gear, that is the third part of the funnel.
If you were to look at it, we have said very openly, that the X-Gene revenue for us, meaningful X-Gene revenue – we don’t like to talk about trivial numbers – meaningful X-Gene revenue would occur towards the second half of our fiscal year next year. In fact, I think we have maintained that all along and we have also maintained that many thousands of servers would be in deployment over the next two quarters and leading to many tens of thousands of servers two quarters after that.
So I think if you were to look at it, we are doing everything in our power to make sure that every piece of the ecosystem is ready. The revenue ramp for something like this is very hard to accurately estimate, but I can tell you that when we look at modeling this, we look at pretty much doubling our revenue in three years – that’s what we talk about and keeping expenses relatively flat. So that is kind of the thought process that we have applied to this. Having said that, we have all of the pieces, we have all the visibility that our customers have given us and I am happy to share with you based on the benchmarking color that I gave you, we have also now seen what our customers’ customer in the case of the OEMs and the server space are looking to do.
So we are very, very comfortable with the fact that we understand and have pipelined the actual revenue funnel to understanding the key metrics towards making it materialize.
10
Jason Rechel - Oppenheimer and Company
Okay. That’s really helpful. Thanks Paramesh. I guess just maybe on the core business, you talked about new products and the embedded business being up 87% sequentially. Can you give us some color in terms of what the split is now in your core business between new products versus legacy products? And then, I think, I missed what your backlog coverage is on the quarter at this point? Thanks.
Paramesh Gopi - President & CEO
So Shiva can talk about the backlog coverage.
Shiva K. Natarajan - VP, Corporate Controller and former Interim CFO
Yeah. So new products represented about 40% of the total revenue this quarter, the way we define new products, and backlog coverage was (is) over 70% at this time.
Jason Rechel - Oppenheimer and Company
Okay, thank you guys.
Shiva K. Natarajan - VP, Corporate Controller and former Interim CFO
Thanks.
Operator
The next question comes from Mr. Christopher Rolland of FBR. Please proceed.
Christopher Rolland - FBR Capital Markets & Co.
Hi guys! How are you? So on the licensee side, basically could you describe maybe what that initial vetting process is like and what the makeup of those, I guess 30 some odd soon to double licensees look like? Are they hyper scale versus server OEM versus networking OEMs, still sort of that 50:50 split between hyper scale and server and then also geographically, maybe if you could describe that as well where that interest is coming from?
Paramesh Gopi - President & CEO
Well, Chris you really know how to dig deep into the stuff, okay. Let me start by saying that the license pool includes three categories: one is, ecosystem partners for instance people who have built software and/or applications, I want to make a fine point on this, people who actually build the applications on top of these systems. They also include folks in the hyper scale/supercomputing space, they include folks in the mainstream server space and the mainstream server space includes server OEMs in all geographies, and finally major networking players both in the telecom infrastructure space and the packet infrastructure space, right? So I wanted to actually maybe walk you through a few of the qualification metrics that they use and we use mutually to hand them over I’ll call it at this level of disruptive modern technology. So when you take on developing something like this it involves you getting exposed to not only our entire hardware and software development platform, but also getting exposed to things like base hardware specifications and roadmaps for the next two generations, right. And the investment that’s made on the customers’ part is between 10 and probably 20 engineers at a stretch for at least six months. They start to do meaningful software work and/or meaningful hardware work to make a customer driven platform for internal use. So that’s the level of actual work that has to happen to do that.
Some of the folks who had signed these types of agreements include the folks that have been partnered with us for the longest time like the HPs and the Dells of the world. And some of the more recent folks who have signed it are more application development houses, are more folks who are the commercial edge of receiving an end customer platform who run their own cloud service, for instance. So there is a very large difference in the type of company
11
that is willing to make the commitment, and I think if you were to go back and tie it to the benchmark that I talked to you about, in their mind, they now see the equivalence between this class of ARM server platform and X86 and Xeon class platform.
Christopher Rolland - FBR Capital Markets & Co.
And just geographically, are they all North American or are there Some Taiwanese, Chinese, European?
Paramesh Gopi - President & CEO
You should assume that it’s worldwide, you would say preponderance in Pan America, I would say, and Asia PAC, being China and Taiwan, being very critical.
Christopher Rolland - FBR Capital Markets & Co.
Okay, great and then, a large competitor came out announced a 64 bit architecture seems like it’s really the first step at least is more of an NFV product even though they call it server class. I wanted to know if you had any opinions there, also IBM announced and ARM announced them today it’s not – it doesn’t look like it’s a micro server announcement or something like that but I am just wanted to know if you have any thoughts there as well?
Paramesh Gopi - President & CEO
Yeah. We are very happy that more people are joining the ARM ecosystem. It makes the ecosystem stronger and gives lots of differentiated and large market acceptance to the ARM architecture. I can tell you that as far as the announcement from the major competitors, major – I’ll call it larger companies concerns – I think the way we view it is, it is a network and security processor oriented architecture and it comes from a MIPS legacy base and it talks about things that are two years out. I remind you that by the time that is announced, we have publicly said that we probably have two plus one generations in the marketplace and if you were to go back and look at the software presence that we built in the ecosystem and the lock that we have built for the server class devices it’s going to give us a very nice catapulted advantage.
So I think it’s a great – it’s great to see them, see a larger company having to be ARM space, we view it more of an embedded networking kind of continuation of an existing MIPS product line that is more oriented towards security and networking.
Christopher Rolland - FBR Capital Markets & Co.
Thank you, guys. Thanks for taking the questions.
Operator
The next question comes from Mr. Patrick Wang of Evercore. Please proceed.
Patrick Wang - Evercore Partners
Hey guys! Thanks for letting me ask the question. You know the first one I had was, I want to go back to your optical, the way you guys see the optical market over the next couple of quarters here. Just give me your exposure to the data center side. Can you give us a sense of what and you think we’re in here and you know, what the type of confidence you have got in the ramp next year?
12
Paramesh Gopi - President & CEO
Let me take, let me again come back to you. I may sound like a broken record, but let me say it again. We are designed into – 50% of all 100-gig links in the world are designed in with our silicon. I can tell you that. You can – that’s really neat. And I would say that more than 50% of all packet optical links in the world that are bridging data centers are designed in with our silicon. I will also tell you that we have a unique pole position in the tier-1 telecom networking space both with our legacy products and with our new products pre X-Weave, right. If I were to take those three moving parts, the part that I see having the most robust trajectory from a growth vector point of view that is predictable is the middle, which is packet optical data center platforms.
The most lumpy is the telecom service provider platforms and it goes – it ebbs and tides, ebbs and flows with basically service provider spending. The one thing that’s really surprised us and continues to drive for us is that 100-gig piece is catching a lot of traction relative to the data center and dense data center both internal and external kind of platform land. So I can tell you for a fact that the margin that you saw last quarter was a lot driven by the 100-gig stuff and we are very, very, very pleased with the 100-gig stuff. And we believe that long term with X-Weave and the reason X-Weave is so important is, all of that mixed signal IO at 100-gig that we are shipping today including the DSP and the modulator and the files will all get integrated onto one die. So that’s the whole vision going forward.
So to characterize it, we are very, very – we are predictably confident on the data center vector in our optical business. We remain exposed favorably to lumpy service providers spending on the telecom infrastructure side. And we remain disposed favorably to 100-gig deployment.
Patrick Wang - Evercore Partners
Okay, great, that’s all right. So it’s fair to say that you feel pretty good about the ramp next year. Okay, and then the other question I want to ask was, well going back to X-Gene, there is lots of thought out there in terms of what X-Gene is capable of, what the competitors are at the timing, this and that. You know, there is a part out there today that you guys will be competing with and it Intel’s Avoton® part. Have you – do you have a sense of kind of how X-Gene compares on a performance side of things versus Avoton, and I know that you guys are using E5 as your kind of competitor, but given the fact that Avoton is their Atom based part, can you help us understand how you expect to outperform them and by how much?
Paramesh Gopi - President & CEO
Patrick, it’s a really good question, and I want to take this opportunity to say two things. I think the industry is confused with this whole notion of a micro server. There is no such thing as a micro server. All of the servers shipping today in the world are cloud servers, because cloud is characterized by Linux, Apache, MYSQL, PHP, 64-bit X86. I would say almost, 90% of those are all E3 and E5 based. What I want – the other characteristic in the cloud is large addressable memory for big sparse data, for big data, which means two to four channels of very high performance DDR3 or 4 and many 10-gig pipes. So the struggle that we have, and we talked to people and making them – to help them understand the market is, Avoton does not have the 10-gig or the memory support of what is required for the mainstream cloud, which is why we keep talking about Xeon, Xeon and Xeon – that is the only existing platform that we need, that we are in the process of basically either replacing or augmenting with our ARM portfolio. So I want to stop there, because anything else that people talk about in the micro server space is the term that was coined, I think, with no basis. It’s Cloud Servers, it’s what ships today, and fundamentally you have to talk about 10-gig, multiple 10-gig, multiple two to four channels of high speed DDR, a lot of the integrated storage and lots and lots of PCIe Gen 3. We are talking about between 12 and 24 lanes, you can’t find that on an Avoton part today. So it’s very hard for us to draw that comparison.
Patrick Wang - Evercore Partners
I see. Okay. So basically apples and oranges.
13
Paramesh Gopi - President & CEO
Precisely, you got it. You nailed it.
Patrick Wang - Evercore Partners
I look forward to. Thanks so much.
Paramesh Gopi - President & CEO
Thanks.
Operator
The next question comes from Krishna Shankar of ROTH Capital. Please proceed.
Krishna Shankar - ROTH Capital
Yes. Congratulations on your X-Gene momentum. What needs to happen between now and I guess early 2014 where you mentioned commercial deployment? Does there need to be any changes to the silicon driver platform? Can you talk about the process between now and commercial deployment?
Paramesh Gopi - President & CEO
Yeah. So fundamentally we are – we have said this earlier on, we are preparing for essentially running production wafers towards the end of the year for our first generation platform, which means that from a validation perspective, from a system and application benchmarking perspective, all of the work on the first generation is complete. So at this point in time, it is making sure that operating systems, all of the compiler tweaks, all of the pieces relative to making sure that you have a robust platform ecosystem that is shrink wrappable and shippable, exists. So from a hardware perspective, you know, as I said, the production silicon in the offing, we are doing all the appropriate things to start production wafers which is why we talked about pre-production parts that we shipped last quarter.
So first of all, did I answer your question?
Krishna Shankar - ROTH Capital
Yes, you did Paramesh. Yes.
Paramesh Gopi - President & CEO
Okay.
Krishna Shankar - ROTH Capital
Secondly, who are working – you know, you cannot talk customers specifically, but can you give us a sense for who the first adopters will be between the server OEMs, the cloud data center operators and networking. Can you give us a sense for the type of OEM and also the application where it will be deployed, web servers or application servers, can you give us some sense for the application also?
Paramesh Gopi - President & CEO
Sure. So I think we can talk, I can give you some color on both. We have talked about kind of three classes; people who own their, vertically own their own data centers, they have their own software, hardware platforms. We have people like the server OEMs such as HP and Dell. We have networking OEMs like Huawei and Cisco and ZTE and so forth. And then we have the large tier-2 base which is all of the web scale and/or hyper scale guys who essentially buy from the server OEMs.
14
So I would say that you have pretty much multiple tracks from each of these segments driving X-Gene deployment and I want to be clear about one thing which is, in each case, there are different driving factors and ecosystem requirements, I am not talking about software but density, I am talking about 10-gig versus 40-gig versus essentially driving different form factors. So we believe that if you look at the workloads where X-Gene in the – workloads in the cloud, I am not going to – I said this 100 times, there are three major scopes of cloud workloads. One is all of the web scale dynamic web deployment. The other piece is all of the storage, I will call it everything from cold to I’ll call it Hadoop type warm and hot, and the third is I will call it caching, mem cache is one example. X-Gene is applicable across those three ecosystems. I would think that the volume and this is if you look at the tiered volume of servers that ship into those, the highest volume ships into the Web tier. The second highest volume ships into the storage tier and then distributed storage tier, and the third highest ships into the caching tier.
So I would – I can on the color wise, the first volume will all come from web tier guys and I would say that they will all come from either tier 1 or tier 2 server OEM,s and/or tier 1 or tier 2 vertically integrated data center OEMs.
Krishna Shankar - ROTH Capital
Great. Thank you. That’s very helpful. And Shiva for fiscal year ‘13 and for the first quarter of fiscal year ‘14, can you give us the revenue mix between legacy embedded and new embedded and connectivity and what it was for the first quarter?
Shiva K. Natarajan - VP, Corporate Controller and former Interim CFO
For first quarter, this is fiscal 14, are you asking about fiscal 13 or 14?
Krishna Shankar - ROTH Capital
Well yes, the first quarter of fiscal year 14 and then all of the last fiscal year?
Shiva K. Natarajan - VP, Corporate Controller and former Interim CFO
Well, I can give it to you offline because we have broken that out in previous calls, but I can give you exact numbers offline. Hope that works? Does it work Krishna?
Krishna Shankar - ROTH Capital
We can talk about this offline.
Shiva K. Natarajan - VP, Corporate Controller and former Interim CFO
Okay, thanks.
Operator
Your next question comes from the line of Vijay Rakesh of Sterne Agee. Please proceed.
15
Vijay Rakesh - Sterne Agee
Hi, Paramesh. Just looking at what you have said about the Xeon and the X-Gene, when you look at the TCO, let’s say if the Xeon has 130 watt power consumption and X-Gene is two watts per core and normally is it like eight cores, are you saying 16 and 20 watts on your X-Gene it can deliver similar performance to the Xeon?
Paramesh Gopi - President & CEO
Yeah, let me go back and talk you about this and we said this many times over. If you look at it, the comparison that we draw is very straightforward. Today’s cloud server usually consists of a 2P Xeon socket. The 2P Xeon socket has two processors, two CPUs and I/O chip, and some type of PCIe for 10-gig. That is the comparison framework that we used to talk about, a single chip Server on a Chip device which is X-Gene 1, which is basically equivalent of that for cloud workloads. So 10-gig, four channels of DDR, we talked about eight cores running at 2.4 gig, we talked about more off loads. We talked about integrated PCI, we have 17 PCI Gen 3 integrated. We talked about multiple, multiples SATA and size integrated. So fundamentally, that’s what we are comparing. So when we talk about it, we talk about it in terms of TCO. The reason we talk about in terms of TCO is customers don’t care about individual CPU power, they don’t care about individual NIC power, they care about system power. In the customers’ mind, TCO includes acquisition cost, and operating cost. And we are saying that if you were to look at it, acquisition cost wise, we are roughly half operating cost wise, we are roughly about a third to a fourth. So we say 50% to 70% in simple terms from a TCO perspective.
Vijay Rakesh - Sterne Agee
Got it. Good color. And also when you look at 2014, I know you talked anecdotally all the estimation of street vary between 50 and 100 million in X-Gene for next year, not to box in one of those boxes there, but how do (you) look at content and I know you talked about tens of thousands and slowly that number going up in terms of shipments. Can you help us kind of put a bracket around those numbers?
Paramesh Gopi - President & CEO
So I just want to be clear that we have never ever given guidance for X-Gene revenue, and we won’t until we get closer to, even if it’s actually materializing. So anything that you talk about didn’t come from us, specifically not from the company. So our perspective, we have always said we double our revenue in three years that’s been – this is when I said – I said this about four months ago, and we have kind of started the clock there. So I just want to be very clear and I don’t want have any type of mis-estimation going on from the model perspective.
To answer your second part of your question, please remember that a 2P socket, if you can, today you are fundamentally governed when you build a server by socket based economics, which means how many cores in a socket, how many 10-gig pipes per socket, how many memory channels per socket. In the cloud scale out server model, you move to – when you – what we have done is essentially for all semantic purposes take a socket based approach and make it a Server on a Chip. So we integrated a chipset onto one die and added way more in terms of mixed signal and I/O. From an I/O perspective, we are probably two to three years ahead in terms of I/O integration because today we are shipping 100-gig into platforms, I mean competition that we have doesn’t even ship that stuff. So all of the I/O integration is so important. So what are we doing, we are enabling our customers to have a rough three to four X density increase in one single unit of electricity. If you were to look at it, that’s what we are talking about. So if I were to look at it, it’s about two to four X density increase, therefore a two to sometimes six X information density session increase, in terms of session density, all for roughly half the power.
So that is the type of economics we are talking about here.
Vijay Rakesh - Sterne Agee
All right, thanks, very helpful.
16
Operator
Your next question comes from the line of Suji DeSilva of Topeka, please proceed. The next question comes from Suji DeSilva of Topeka, please proceed.
Suji DeSilva - Topeka Capital Markets Inc.
Hi guys! Can you hear me?
Paramesh Gopi - President & CEO
Yeah, we have been waiting. I think you were on mute.
Suji DeSilva - Topeka Capital Markets Inc.
Hi guys! Can you hear me now?
Paramesh Gopi - President & CEO
Yes. We can.
Suji DeSilva - Topeka Capital Markets Inc.
Okay. Can you hear me now?
Paramesh Gopi - President & CEO
Yes we can hear you.
Suji DeSilva - Topeka Capital Markets Inc.
Great, sorry. So on the phase 4 of the server rollout, the five phases you talked about, how long is that length on average; three months, six months, just kind of a rough time frame there?
Paramesh Gopi - President & CEO
Yeah, pilot phase will take between three months, sometimes it can go out to six months, but three to six months on average. And basically they will do all of the mission-critical, both system testing as well as application testing.
Suji DeSilva - Topeka Capital Markets Inc.
Okay great. And my second question is on the 30 customers you are going to 60, have you done the analysis of how much of your target audience that already covers or is there a lot more there and maybe even more important kind of percentage they spend those very 60 customers already covered, have you guys done that analysis, is there any 80-20 rule here?
Paramesh Gopi - President & CEO
I think we just use that as a proxy to tell you the pervasiveness of the estimable category and showing you how much involvement there is, our customers are talking about. So you remember that these include partners as well. They are not just pure customers. For instance, when we do an enterprise release with the major Tier 1 software partner, they are included in that. So I want to be very clear, it’s partners and customers.
17
Suji DeSilva - Topeka Capital Markets Inc.
Okay, it’s very helpful, thanks guys.
Operator
Your next question comes from Mr. Brian Thonn of Kingdom Ridge Capital. Please proceed.
Brian Thonn - Kingdom Ridge Capital
Hey guys! Thanks for squeezing me in. two questions for you. First off, I know you don’t want to give your explicit guidance, but if we look into 2014, the calendar year 2014, how should we be thinking about your new products in your base business? Obviously they were strong this quarter, but just trying to kind of get an idea how that can play out going forward and then second, I wonder, if you have any qualitative commentary you can provide on the grueling testing and validation. Do you think like what people were expecting to see and what they saw was basically in line or are there any variances there that would be notable?
Paramesh Gopi - President & CEO
Let me answer the first question. I think we have said all along that, even many months ago, we said we developed about nine new products that were going to generate between $150 million and $200 million of revenue over the next two years. This was about almost a year ago and we are tracking, right on track to that Brian. The other thing that we are really encouraged by, when you have a 87% increase in your embedded products, and they are in all of the leading… and these are, remember, these are in ASIC products for printers, these are in switches like the catalyst type switches at Cisco, we are extremely encouraged by the fact that our new products are actively driving the offset of the legacy falling off. So the point that we are trying to make there is that if you were to look at our new product ramps, they are in line with what we talked about. We also see strength in the 100-gig business next year and we also see the X12 coming on board in a really neat opportunity to drive our connectivity business to new heights because we now have the ability to provide LAN type densities into large routers, into large telecom switches.
So I think from a new product perspective, we have been right on track and we are very encouraged by signs we see with all of our embedded computing products and of our connectivity products. Okay. Does that give you – did that answer your question?
Brian Thonn - Kingdom Ridge Capital
That’s good.
Paramesh Gopi - President & CEO
Okay. The second piece, which is the grueling test, I mean you picked up on the word grueling, we have been kind of – we have been in – I would call it the grinder with our customers. Just to give context, when somebody decides to switch mission-critical infrastructure in the server space, it’s equivalent to forklift upgrading and writing an SLA in the telecommunication space. It’s like saying, I am going to switch your pipe over to 10-gig overnight and I am going to give you the same level of service which means that they have to do a whole bunch of application testing and by the way, just to be clear, today, there is only one architecture that drives the cloud. It’s X86. Therefore, it’s also been the same architecture that drove the enterprise for the last 25 years. So when our customers talk about testing and grueling validation, they are taking 15 to 20 years of X86 software code and rerunning it on X-Gene. And the reason we feel so good about it is every corner case that’s mission-critical, every I/O case, in fact what’s really neat about it is that we have now understood what it takes to build a mission-critical server device. And if you have not been in the space and you haven’t done it, you get a whole new appreciation for how important it is to be able to withstand that level of, I will call it, reliability, accessibility and manageability testing, in addition to the performance testing.
18
So it gives us great confidence that there is going to be no – there is no big holes at all that we haven’t talked through. The second piece that’s really neat about is that we have actually advanced the state of art of some of these customers because they never had the ability to test this much I/O, this many pipes going in and out of a large dense server. And it’s really neat because now we are also going to be in the process over the next year of setting a new bar for the Cloud Server market, because we are enabling our customers to run tests faster, to run tests in a more acute way, because of the integration we have achieved with X-Gene.
Brian Thonn - Kingdom Ridge Capital
Okay. Thank you very much guys.
Operator
There are no further questions in the queue gentlemen.
Michael Major - VP, Corporate Marketing and Human Resources
Okay, we would like to thank all of you for your participation today. There will be an audio replay of this call available on the Investor Relation sections of our website. You can also access the audio replay of this conference call by calling area code 888-286-8010 and entering the reservation number 48794252. We will also file a copy of the transcript of this call on an 8-K with the SEC in the next few days. Feel free to call if you have any additional questions. Again thank you all for your participation on the call today and have a nice evening.
Operator
Thank you for your participation in today’s conference. This concludes the presentation. You may now disconnect and have a great day.
19