Silicom Ltd (SILC) 2020 Q4 法說會逐字稿

Ladies and gentlemen, thank you for standing by. Welcome to the Silicom Fourth Quarter and Full Year 2020 Results Conference Call. (Operator Instructions) As a reminder, this conference is being recorded.

You have all received by now the company's press release. If you have not received it, please contact Silicom's Investor Relations team at GK Investor & Public Relations at 1 (646) 688-3559 or view it in the News section of the company's website, www.silicom-usa.com. I would now like to hand over the call to Mr. Ehud Helft of GK Investor Relations. Mr. Helft, would you like to begin, please?

Thank you, operator. I would like to welcome all of you to Silicom's Fourth Quarter and Full Year 2020 Results Conference Call. Before we start, I'd like to draw your attention to the following safe harbor statement.

This conference call contains projections or other forward-looking statements regarding future events or the future performance of the company. These statements are only predictions and may change as time passes. Silicom does not assume any obligation to update that information. Actual events or results may differ materially from those projected including as a result of our increasing dependency for substantial revenue growth on a limited number of customers in the evolving cloud-based SD-WAN, NFV and Edge markets.

The speed and the extent to which solutions are adopted by these markets, the likelihood that we will rely on question on customers which provide solutions in these evolving markets resulting in an increasing dependency on a smaller number of larger customers, difficulties in commercializing and marketing Silicom's products and services, maintaining and protecting brand recognition, protection of intellectual properties, competition, disruptions to our manufacturing and development, along with general disruptions to the entire world's economy relating to the spread of the novel coronavirus, COVID-19, and other factors identified in the documents filed by the company with the SEC.

In addition, following the company's disclosure of certain non-GAAP financial measures in today's earnings release, such non-GAAP financial measures will be discussed during this call. Such non-GAAP measures are used by management to make strategic decisions, forecast future results and evaluate the company's current performance. Management believes the presentation of these non-GAAP financial measures is useful to investors' understanding and assessment of the company's ongoing cooperation and prospects for the future. Unless otherwise stated, it should be assumed that financials discussed in this conference call will be on a non-GAAP basis.

Non-GAAP financial measures discussed by management are provided as additional information to investors in order to provide them with an alternative method for assessing our financial conditions and operating results. These measures are not in accordance with or a substitute for GAAP. A full reconciliation of non-GAAP to GAAP financial measures is included in today's earnings release, which you can find on Silicom's website.

With us today on the call is Mr. Shaike Orbach, CEO; and Mr. Eran Gilad, CFO. Shaike will begin with an overview of the results followed by Eran, who will provide the analysis of the financials. We'll then turn over the call to the question-and-answer session.

And with that, I would now like to hand over the call to Shaike. Shaike, go ahead, please.

Thank you, Ehud. I would like to welcome all of you to our conference call to discuss our fourth quarter 2020 results.

We are very pleased with the solid improvement in our financial results in the fourth quarter. We reported revenue of $33.9 million ahead of our guidance expectations, a 33% improvement over the fourth quarter of the past year and also a strong 19% improvement over the previous quarter. Despite what has been a very tough year, 2020 still saw revenue growth versus 2019. Furthermore, we reported our 64th quarter of profitability with net income of $4 million. This represents a 29% increase over Q4 last year and a 36% increase versus the prior quarter.

We ended the quarter with over $76 million in net cash and through the quarter, used our strong cash position to further progress on our share buyback with the goal of bringing increased value for shareholders. Our results demonstrate that we have been able to move ahead with our strategy and more than maintain business continuity in a difficult environment. This shows the resilience of our business and our ability to plan and overcome the work challenges that the ongoing pandemic continues to bring to all of us.

During 2020, our strategy was made even more clear and more focused than ever before. Throughout 2020, we primarily ensured that we were ahead of and fully tapping into today's most important market trends, the shift to the cloud as well as disaggregation and decoupling. This is where we directed our R&D investments as well as our sales and marketing efforts over the past few years with increased focus and additional strategy fine-tuning in 2020. We already see the fruits of our investments with successful customer penetrations throughout the relevant product lines that address those trends.

Today, I want to spend time talking about the trends that we cater to that we see driving our business in the coming quarters. The first and obvious trend which I would like to discuss is the shift to cloud provided services, both public and private. The immediate outcome of the cloud shift is the move towards standardization, which is key for scalability. Standardization has led to 2 important trends, which are disaggregation and decoupling. Disaggregation means replacing the proprietary interfaces between the various parts of the network with standard interfaces. This has the benefit of allowing the various parts to be procured separately from independent vendors.

Decoupling is the separation of the software layer from the underlying hardware, allowing, again, independent procurement efforts. Disaggregation and decoupling trends, which started in the cloud, have created a process impacting all parts of the networking market. This process started with standard service built as a part of a generic infrastructure by cloud players. It then moved to telcos operating in the SD-WAN and NFV market segments, which are increasingly procuring the hardware and the software separately. And now telcos are adopting O-RAN, Open Radio Access Networks, which allows such disaggregation and decoupling in their 4G and 5G infrastructure deployments.

I would like to discuss how these trends are impacting us. Summarizing the impact, while at the most basic level, the cloud trend creates a gradual decrease in demand for our server adapters as our traditional appliance customers are increasing their software-only offering, which is what they sell to the cloud players. Overall, both the cloud as is and the disaggregation and decoupling trends have a dramatically positive impact on us.

Let me elaborate. Standardization and the existence of standard service in the cloud increases the demand for smart cards. As such, servers need acceleration offloading abilities to support the required performance. This obviously corresponds with our massive investments in smart cards, including FPGA-based cards and other types of offloading and acceleration cards. Furthermore, standardization in the cloud created a disaggregation and decoupling trend, which has caused significant demand by the telcos for software independent smart platforms for use as CPE devices for SD-WAN and NFV.

We, in turn, are seeing the increasing demand for our CPE devices, which we developed over the past years due to that trend. The potential for both smart cards in the cloud and CPE devices for SD-WAN and NFV is huge. These basic trends are now further penetrating the mobile 4G/5G infrastructure market through the O-RAN standard. It is, again, a huge opportunity for us. Within the mobile infrastructure, O-RAN creates opportunities for combinations of our products and expertise where we can take advantage of our unique and integrated capabilities in networking, acceleration, FPGA and smart platforms.

All these are needed, both in the front haul or Edge and at the backhaul, as all these segments of the network are now operating via the O-RAN definitions. Consequently, in the front haul, we are now able to propose to our customers a unique integrated distributed unit, which includes acceleration functionality, whether via dedicated accelerators or through FPGA-based solutions and time synchronization functionality, both and must for such a deployment.

We are also proposing our acceleration in time synchronization solutions separately for other DU, distributed unit manufacturers, which by itself is a huge market. And also in the backhaul, our FPGA-based offloading cards are generating interest for user plane functionality offload moving towards the telco cloud as well.

As you can see, moving into 2021 and beyond, Silicom is perfectly positioned. The markets that we address are all performing strongly and are expected to continue with their significant growth in CapEx investments. The trends within these markets support the areas that we have invested in over the past few years. They support our cloud penetration efforts through smart cards, our SD-WAN and NFV related efforts through smart platforms and now they also support our smart platforms and smart cards together and separately for the huge 4G/5G O-RAN based mobile infrastructure market.

In fact, we have correlated our investments with these trends, and we have been able to achieve initial success across all our product lines developed to address these trends. Our focus became very clear to us, allowing us to abort efforts which did not match our focused strategy. However, even with all the trends which are converging in our favor, I stress that the sales cycles in our current markets, especially with telcos, are longer than what we have traditionally seen.

On the other hand, many of such new potential telco design wins have much greater scaling potential than what we have traditionally experienced. Furthermore, each design win we have already achieved and continue to achieve represent an opportunity for sustained long-term revenues once we establish a relationship with the customer. And this is also why our recent success and wins are indeed so important. I would like to focus on those recent successes.

As you know, we're selling one of our FPGA-based cards to an important cloud vendor, which is using the card to accelerate the functionality of its network nodes. And indeed, what we were planning with this customer is now actually happening. Quantities are growing gradually and opportunity for another smart card, which is offered for a different part of the customers' data center, is now being discussed. And we have also begun discussions about the next-generation solution for the current smart card.

All of this is just one example of the fact that, once again, with our new cloud and telco customers, we're building the same kind of long-term multiyear relationships we have built in the past with our OEM customers. Early in the fourth quarter, we announced 3 new telco uCPE wins. These wins, which are expected to reach $10 million per year altogether, demonstrate that indeed, the strategy of providing smart platforms to telco while in the background continuing to work together with software vendors, exactly the purpose of the decoupling trend, indeed, works for us, contributing to our portfolio of customers as well as to our revenues.

Later in October, we announced that a Tier 1 mobile operator selected our O-RAN ready architecture for next-generation distributed units for its field trials. Should the solution be selected for mass production, the revenue potential could be tens of millions of dollars from this customer alone. The architecture selected by this customer for its field trials reflects all the principle of our disaggregation and decoupling strategy. It is based on an architecture of smart platforms that we imported from our SD-WAN NFV solutions. It is over uncompliant, disaggregated and decoupled. It includes our smart cards for both SEC acceleration and time synchronization and the customer valued our capability of proposing a fully integrated solution of all the new components and decided to do its field trials with this architecture.

Furthermore, this selection was also enhanced by our very close cooperation with Intel and a leading 5G software vendor. I know that such cooperation and relationship building is an important part of our strategy. About 2 weeks ago, we announced that one of the world's largest telcos awarded us with a design win for our V-RAN, Virtual Radio Access Network, SEC accelerator cards, for their new 5G network distributed units.

The importance of this win is far beyond the numbers. Our customer in this case is a beacon, paving the way for many other customers, each of which is a giant in their own right. We are already feeling the dynamics which have been created by this win through continuous request for evaluations, relationships created with other industry giants and also through additional variations of the solution that we provide.

Furthermore, as a part of the deployment process, we will work closely with 2 of the telcos key distributed unit suppliers, both of which are world-leading server manufacturers. This will open the door for further significant sales of our SEC accelerated cards to additional end users on top of the telcos which are approaching us directly.

This win, facilitated by the disaggregation and decoupling trend, demonstrates the unique value that we bring in all our products for this market, value-based on our deep 5G system-level understanding, our high-performance to cost ratio, our extremely close working relationship with Intel and a close and intimate relationships we have built with most of the key players in the 5G space.

As you can see, our investments, as well as our recent success and announcements, are all in line with our strategy of aiming at addressing these market trends.

I would like to spend a few moments discussing our guidance. For the first quarter of 2021, we expect revenues at between $28 million and $29 million, which represent year-over-year revenue growth of 29% at the midpoint. Given our very long and growing list of design wins, generating ongoing orders, our solid baseline of activities and strong market fundamentals with our focus on some of the latest -- of the fastest and latest growing markets in the networking space, we are exceptionally well positioned for 2021 and beyond.

Thus, we feel comfortable with providing more color on our 2021 annual revenues and believe it will be in the range of $120 million and $130 million. At the midpoint, it represents 17% year-over-year growth. Please note, our expectations assume that the COVID-19 crisis will not worsen. It also assumes we will continue to be able to overcome the significant challenges presented by long lead times of critical components from many chip vendors.

Components lead times are increasing and scarcity is increasing pricing -- prices. And in some cases, we see lead times of 12 months or even longer. While we have done our best to ensure more than sufficient inventory, the effect of longer lead times could have an impact when or if the mix of actually ordered products differ significantly from the forecasted mix. It could also bring -- it could also impact the possible upside we would gain from additional potential wins in our pipeline or from faster than forecasted ramp-ups of existing design wins.

In summary, as I've shared with you, the disaggregation and decoupling trends have had the impact of significantly increasing Silicom's potential. Our long list of design wins, our partnership with leading software vendors and telcos, our extensive collaboration with Intel, and our current long and deep pipeline provides us with much optimism going forward, which continues to grow.

Consequently, we expect that the coming few years for Silicom will see performance well ahead of what we have achieved over the past few years and that we will continue to achieve ongoing revenue growth at a double-digit compound annual growth rate for several years ahead.

With that, I will now hand over the call to Eran for a detailed review of the quarter's results. Eran, please go ahead.

Thank you, Shaike, and hello, everyone. Revenues for the fourth quarter of 2020 were $33.9 million. This is a year-over-year increase of 33% compared with revenues of $25.5 million as reported in the fourth quarter of last year and a sequential increase of 19% over the $28.4 million reported in the prior quarter.

Our geographical revenue breakdown over the last 12 months were as follows: North America, 61%; Europe and Israel, 33%; Far East and Rest of the World, 6%. During the last 12 months, our top 3 10% customers together accounted for about 35% of our revenues.

I will be presenting the rest of the financial results on a non-GAAP basis, which excludes the noncash compensation expenses in respect of options and RSUs granted to directors, officers and employees, acquisition-related adjustments, lease liabilities, financial expenses, as well as impairment of intangible assets.

As Shaike mentioned, in 2020, we correlated our investments with the disaggregated and decoupling trends. And following our initial success across all our product lines addressing these trends, we aborted some efforts which did not match our focus strategy. As a result, we took a onetime $1.7 million impairment charge to our cost of sales. For the full reconciliation from GAAP to non-GAAP numbers, please refer to the press release we issued earlier today.

Gross profit for the fourth quarter of 2020 was $11.4 million, representing a gross margin of 33.6% compared to a gross profit of $9 million or gross margin of 35.1% in the fourth quarter of 2019. The variance in the gross margin is a function of the specific product mix sold in the quarter.

Operating expenses in the fourth quarter of 2020 were $7.5 million compared with $6.1 million in the fourth quarter of 2019. All of the increase is R&D related and represents our continued significant investment in developing new products, IP and technologies.

Operating income for the fourth quarter of 2020 was $3.9 million, an increase of 37% compared to operating income of $2.9 million as reported in the fourth quarter of 2019. Net income for the quarter was $4 million, an increase of 90 -- of 29% compared to $3.5 million in the fourth quarter of 2019. Earnings per diluted share in the quarter were $0.56. This is a year-over-year increase of 37% compared with EPS of $0.41 as reported in the fourth quarter of last year, and a sequential increase of 37% over the $0.41 reported in the prior quarter.

Now turning to the balance sheet. As of December 31, 2020, the company's cash, cash equivalents, bank deposits and marketable securities totaled $76.1 million with no debt or $11.02 per outstanding share. During the quarter, we further executed on our second $15 million share buyback plan, which we started on May 4, 2020. During the fourth quarter, we purchased approximately 111,000 shares at a total cost of $4.2 million.

That ends my summary, and we would be happy to take any questions. Operator?

(Operator Instructions) The first question is from Alex Henderson of Needham & Company.

First off, congratulations, outstanding quarter. And obviously, your strategies are really paying off and quite impressive. I did want to talk a little bit about the issues around the supply constraints, 12 months or longer lead times is really unusual. I get it, this is an unusual period relative to COVID and the like. But how do you think that, that gets resolved? What are the suppliers telling you in terms of flexibility, how are you approaching it relative to giving up some margin to procure additional supplies if you need it to keep your customers satisfied. It seems like that's an unusually large constraint.

Well, first of all, yes, I mean, this is an unusually large issue or constraint this year. This year is 2021, obviously, is really going to be a challenging year from that perspective. And we are doing several things in order to mitigate that. First of all, I think we are -- and this is not something that has started right now. We are preparing ourselves. We did have some indications that it's going to be a challenging year. We have approached our customers, and we ask them to provide as accurate, with a certain potential for upside, a forecast as possible. We have had internal meetings estimating the forecast and we have placed orders for components ahead of time.

And that means that it's not as if these challenges are impacting everything that we have planned for, just the other way around. In fact, the impact would be or I wouldn't say the impact because we would try to mitigate the impact as well. But the challenges would be on everything which is not planned. So everything which is a part of our plan, we took care of that, we placed orders even a while ago, a certain while ago, in order to be able to do that on time. So that's one part of that.

Another part of that is, I would say, applying pressure. I mean, it may not be that nice, but at the bottom line, we do see that pressure helps. So if you would say that you would look at how I split my time, for example, then due to these challenges, I am allocating more and more of my time to talk to these suppliers to their management to explain to them about our business potential and so on and so forth. And I think that we're seeing some success in that area because we invest more in getting the components that we need, so we get more. So that's the second part of that.

A third part of that is making our customers a part of that process. And when we have big customers, they're helping us with this process as well. Now on top of that, because some of these components, you could also buy in the free market rather than from the authorized -- rather from the, I would say, the manufacturers. Obviously, only once confirming that from a quality perspective this is okay. But in the free market, if you want to buy these components, you need to pay more. And we are talking to our customers to participate in that and to undertake for themselves at least when they come up with an upside or an unfocused demand so they should take the expediting fees, which we call PPVs, in order for us to be able to deliver to them their goods. And this is also working for us.

Now I would say that this is not always working. And sometimes, we would need to pay somewhat more for components in order to meet that -- to get them in time. And some other parts, we would not be able to achieve what we want to achieve. But overall, I still feel that 2021, that we are going to be successful. There would be challenges. This is how we're trying to mitigate them, but the challenges are still there.

A couple of questions. Thanks for the guidance on '21 topline. Can you talk a little bit about what you think the impact of the exchange rates are? Obviously, the shekel has been quite strong on operating margins and what your plans are relative to investing in the R&D line given the magnitude of the opportunities in front of you? And then finally, can you talk about what you think is going to happen with gross margins as you continue to shift to larger and larger deals? These trends obviously have an impact.

Yes. Well, I would respond, Eran would later elaborate on the shekel versus the dollar. I would say that, first of all, yes, we are investing in R&D. We are, I would say, increasing our R&D efforts, but it's not going to be anything dramatic overall. But what helps us a lot, I think, is that our strategy during 2020, we have been able to clarify our strategy to a level that we haven't been able to do before. And that's why I feel that our investments right now are much more focused and that's why even though -- that's why I can tell you that I feel as if we are increasing our investment in those areas, which we believe in even more than the increase in R&D expenses because we're able to take away from our table, take off the table, other things that we consider to invest in, which are not a part of the strategy right now.

So overall, I would say, yes, I mean, our R&D expenses in terms of men and labor hours because, indeed, the shekel versus the dollar can change the picture significantly. Just like it has this year. But in terms of how much we invest from the perspective of people that we engage in these efforts, sometimes subcontractors, et cetera, that would increase, not dramatically, but that would increase to a certain level, but it is much more focused and I feel much better with this investment because I really believe in our focus. Eran, you can add a little bit more color to the shekel versus the dollar.

About the exchange rate. First of all, it's the effect of the dollar compared both the shekel and the Danish krone. We have a subsidiary in Denmark. So the effect in the quarter was actually very significant. The negative effect on our operating expenses was approximately $400,000 and the negative effect on our financial expenses was also approximately $400,000.

Yes. I was really looking forward to, obviously, the move over the last couple of months to new highs on the shekel. I did not look at the krone, that I will admit. But the obvious move. As we look forward, could you just remind us what your hedging policy is and whether there's any hedging in there or whether we should be (inaudible) the negative impact in forward periods.

Our hedging policy is very simple. We do not hedge.

That's right. Okay. I couldn't remember it. Just too many Israeli companies under coverage to remember. And then can you give us some tax guidance for the year? I think obviously that...

Yes. As you could see, the tax expenses in the quarter was actually 0 or close to 0. It is a result of a few onetime positive effect. Generally speaking, I repeat the guideline that we say every quarter, we anticipate, we forecast approximately a 15% effective tax rate for 2021.

(inaudible) that works. And then 1 more question on the modeling. So obviously, you've got tremendous cash generation capabilities. So what are you doing relative to share buybacks or other capital usage going forward?

As I think we mentioned it on the call itself, we progress as planned with the buyback plan and actually intend to meet the targets, which was set by our Board, buyback of $15 million within 1 year.

And any of that exercise at this point?

Any what?

Have you exercised any of that at this point? Or is it just $15 million still outstanding?

Yes, yes. Actually, during the fourth quarter, we purchased more than $4 million, which is even more than the...

So, $11 million outstanding then?

Sorry?

So there's $11 million still outstanding? In the buyback? There was $15 million, you've got $4 million in, it's $11 million outstanding. Is that right?

We still have in the pipeline approximately $4 million, $5 million to invest.

So there's $4 million to $5 million left of the $15 million?

Correct.

That's what I was looking for.

Yes, $4 million, $5 million. Correct.

The next question is from [Sergei Mascaro.]

So first of all, congratulations for the great quarter. It seems that we're finally building some kind of momentum. So that being said, my first question is about SD-WAN space. In the last quarter, there was a Tier 1 SD-WAN project that was not ramping up yet. Did you receive orders in Q4?

I think -- let me just make sure that I understood the question perfectly. You're asking about one SD-WAN customer that we have that in the last call we said it was not ramping up yet, and you're asking if it's ramping up right now? Was that the question?

Yes. Sure.

Okay. So I would say that this question was related to 2 telco wins that we had in the past, out of which, one -- we said last time that one of them was ramping up, the other one did not ramp up yet. This situation has not changed right now, even though there is some dynamics. Well, I'm a little bit hesitant to say that it would ramp because each time that we thought that something is happening, eventually we find out that something different is happening. So we did not.

But on the other side, I mean, during the quarter, we have announced, even during the last quarter, we have announced 3 design wins related to SD-WAN, which together would amount and are expected to ramp up to around $10 million per year and we did have some revenues from these, and I do hope that these will have a certain level of ramp up during 2021.

Great. Perfect. Thank you. Does the press release of earlier this month have any kind of relationship with the previous one, which was about O-RAN's field trials win?

I did not understand the question. Can you run that again by me, please?

Yes, sure. So does the press release of earlier this month has any kind of relationship with the previous press release, which was about the O-RAN field trials win?

No. The 2 press releases are separate, entirely separate with 2 different telcos.

Great. And these networks is working with Intel for the 5G national deployments in the USA. I believe these deployments are going to start in the second half of this year. If I am not wrong, you are the only vendor that's partnering with Intel for 1 product that is selected from Intel. So could you maybe speak about this project since there is public information available?

Well, I cannot speak about specific customers right now. We are working with Intel, I would say, across the board, and Intel is indeed helping us with a variety of customers. Intel is very significant in our engagements with all these customers. And what I can tell you, the one thing that I can tell you is that specifically with the SEC tech acceleration card that we're doing, we're currently the only, I would say, product in the market at this time at least that can support a non-FPGA and non-software, which means a hardware accelerator, which is needed in the 5G deployment and this is based on Intel's -- Silicom Intel's eASIC which is obviously why Intel is helping us in such an intensive way. And that goes to actually all customers.

Okay, perfect. I believe that Asia is expected to grow really fast in the 5G segment. Are you targeting the China and the India market?

I would say, yes, we are targeting the Asian market, obviously, not only the Asian market, but we are targeting the Asian market as well.

Yes. Perfect. Thank you. Do you sell any kind of software for your FPGA? And what does it mean in terms of gross margin?

You're saying if we sold FPGAs.

No, I mean, do you sell any kind of software with your FPGA?

Okay. I see. We definitely sell software with FPGA, but not as a stand-alone mostly. I mean, in most cases, we sell it as a part of the product that we sell, which is coupled with both software and IP, which is inside the FPGA, RTL or whichever way you want to call it. Now this has an impact on the margins, which is why our margins on these sales are higher. However, that being said, now that we're approaching the 5G and the major cloud players with FPGA proposals and the quantities they are going to be very, very significant, these huge player, once they grow up in quantities, they may expect the margins to go down to a level which is closer to our, I would say, general type of margins rather than to the higher level, which we are able to accomplish when the quantities are not yet that high.

Okay. That helps. Could you maybe share with us which are your competitors in the O-RAN space? I can only find Ethernity in this segment, and it is a really small company.

Well, first of all, I should say that O-RAN is a big name. We are not covering the full space of O-RAN. For example, there are quite a few companies which are developing radios which will support O-RAN, and we are not there, obviously. So we are -- within the O-RAN, we operate -- well, rather than giving you a number, I would define where we are operating within O-RAN. So first of all, our major area of operation within O-RAN is what is called the distributed units. This is actually the unit, there are different architectures. So please, I may not be 100% accurate here because my description may not be in accordance with all architectures.

But anyhow, the distributed unit is a unit which is actually interfacing with the radios, sometimes with a switch, but typically in our way of -- in our architecture, it interfaces with the radios on the one side and with the central unit or compute unit on the other side. So within these distributed units, we are proposing to the market, first of all, what goes inside the distributed units, which are 2 types of cards. One is an accelerator, the other one is time synchronization. And the accelerator could be either a card, which is based on the Intel's eASIC or an FPGA card. And then the synchronization card.

We are also, as I said, importing from our SD-WAN activity. We're importing the platform itself and we are able to offer to customers not only the cards which go into the platform, but also the platform itself somewhat customized to address the distributed units requirements. And some customers, they do see value, and we understand why, because it's a relatively complex product, they do see value in buying the full unit, the distributed unit, the platform, the smart platform itself with the cards as a single unit because this reduces quite a few elements of integration, which otherwise they would have had to take care by themselves. So this is one area within O-RAN that we support.

The other area is with FPGA cards, which we are proposing to go into the central unit or the computer, everything is still within the Edge, but this one is closer to the core and to the telco cloud, I would say. So there, we are proposing, I would say, higher end FPGA solutions which is offered in order to support offloading of UPF, which is user plane functionality. So this is our offering within O-RAN. O-RAN includes not only that. It includes some other parts of that, as I said, which we are not participating in.

Yes. Yes. And maybe about your competitors in the O-RAN space?

Yes. So once again, I mean, you mentioned Ethernity and Ethernity is really is a small company. And Ethernity, the main area where Ethernity is trying to compete with us is in the UPF area, not only, but in the UPF, this is where they are operating. I believe that with Intel's assistance, we would be able to achieve most of our goals within this area. That being said, I would say that I think that we are really fast to the market with all these solutions, both SEC acceleration, time synchronization. Let me say one thing about time synchronization though and UPF offloading.

There is a competing architecture, I would say. The competing -- in terms of time synchronization. The competing architecture is using an external switch where the time synchronization is achieved in the switch. This is a much more expensive architecture and this is why many customers are interested in the architecture that we are promoting, whereby the time synchronization sits inside the distributed unit interfacing directly with the radios because this is saving a lot of money. So I think that this architecture will be the winning technology, but we indeed were, I believe, the first -- at least the first significant solution, which is proposed to the market. But it doesn't mean that we would stay like this.

I think that additional competition, both in terms of SEC acceleration and in terms of time synchronization will arrive what I'm hoping of is that because a lot of this business requires qualification of these cards within the platforms, we would be able to achieve sufficient, I would say, partners which would qualify our solutions, and that would make it more difficult for the competitors to plug themselves in once they're ready. But there will be, for sure, competition as well.

Perfect. Thank you. So I believe that your current buyback program is going to be completed by April. Can you speak about your capital allocation plans?

Our buyback program?

Yes, yes, yes. It's going to be completed by April, I believe. So could you maybe speak about your future plans about your capital allocation?

We will think about that once we get closer to the end of this program, and then we will decide and, obviously, notify our decision.

Okay. We will wait more. And my last question, can you speak about your expectations on the 5G and SD-WAN deployment for this year?

For this year?

Yes. Yes. The 5G and SD-WAN deployment?

The one thing that I can tell you, and I cannot provide details at this time, is that we have a very thick and long pipeline and we believe that we will be achieving additional design wins in both areas, both in 5G and in the SD-WAN.

The next question is a follow-up question from Alex Henderson.

So I was hoping we could talk a little bit about the lumpiness of the -- some of these contracts versus the alternative dynamic, which is the breadth of the number of contracts that you have. It seems pretty clear that when you had the IBM issue years ago that it was because you had a huge singular contract that eventually didn't work out. Where we are now, it sounds like you have a lot of fairly solid sized contracts and the number of them is diversifying the demand picture enough that it should be a smoother and more predictable ride from here. Is that a fair characterization of the environment?

Yes, I believe it is. I mean the contract that we had with IBM, just like you said, had its risks. And while everyone knows by now that it's IBM. I don't think it's important anymore. But even though we were not supposed to -- that's fine, I think. There were many unique parts of this contract. And definitely, what we're experiencing right now is more, I would say, smooth, just like you said.

So is there upside in terms of the pipeline that additional contracts are coming near in that will be constrained on supply side that, therefore, gives us visibility that the robustness of this demand will persist through '21 and into '22 as these supply constraints start to ease up. And if we actually see an improvement in COVID conditions, does that mean that the supply constraints might ease significantly, in which case, there would be additional upside. Is that -- I mean, is that the right way to think about the dynamics here?

Well, once again, in general, I would say yes. I would say that -- first of all, and this is still mostly related to your previous question. What we're seeing right now is that all these design wins and partners that we work with, including our pipeline, as a matter of fact, they -- and this is why the sales cycle is so long. They are taking a very cautious approach before they are actually launching massive deployments. This was not the case with the IBM contract, where they took a huge risk and started to deploy before they have resolved all their internal problems or issues or challenges.

In this case, what we're seeing from all our customers is a different approach. They are very carefully evaluating, testing, taking their time before they actually launch that. Now some of these, obviously, with different customers, we are in different time or different phase within this process. So some of them are closed and some of them, we may be able to announce the design win pretty soon and then hopefully experience the ramp-up pretty soon. And others, it will take longer.

Obviously, when the components challenges are eliminated and disappearing, then I think that we would be in a better situation because it does happen and it did happen to us in the past, that once something was delayed, eventually, we lost this part of the business due to the delay. And customers decided to use their previous generation or whatever. So overall, once the challenges on components are disappearing, this would represent a higher ramp up, I would say, capacity than what we're experiencing right now. So yes, I believe we would see continuous growth, and that's why we're talking about double-digit growth, not only that's what we're planning for, not only this year but in the years coming ahead.

So I wanted to talk a little bit about the OpEx numbers. Historically, your OpEx has been sub-20% of revenues. But in the most recent couple of years, it spiked up to 22%, 23% -- 23% on average for the last 2 years. Should we be anticipating that, that stays at that higher level? Or will that gradually trend back towards the 20% level?

I think -- well, first of all, it's obviously a little bit difficult to speak about the OpEx because they're so dramatically impacted by the dollar value. But I would say, to give you just some sort of a thought that if the exchange rate is where it is right now, then we could expect an annual OpEx of approximately $28 million.

Okay. That helps. Is it expected longer term to be able to bring that percentage down again? Or is that business model no longer a (inaudible).

Well I do believe (inaudible) really grow, the percentage will go down again. Because I do believe that once our plan is working, you would be able to see again the famous leverage, and that's where the OpEx in percentage would be less.

(Operator Instructions) There are no further questions at this time. Before I ask Mr. Orbach to go ahead with his concluding statement, I would like to remind participants that a replay of this call will be available by tomorrow on Silicom's website, www.silicom-usa.com. Mr. Orbach, would you like to make your concluding statement?

Yes. Thank you, operator. Thank you, everybody, for joining the call. We wish you all health, and we look forward to hosting you on our next call in 3 months' time. Good day.

Thank you. This concludes Silicom's Fourth quarter and Full Year 2020 Results Conference Call. Thank you for your participation. You may go ahead and disconnect.