Morning sessions and yesterday's sessions, we had several interesting conversations, both myself and my colleagues hosted. Now we're excited to welcome Super Micro Computer to our session this afternoon. From Super Micro, we have Founder and CEO, Charles Liang. We have CFO, David Weigand, and Patrick Wang, who is President and SVP of East Coast as well as Corp Dev. Well, thank you all for joining us. Let me pass it on to David Weigand to make some introductory remarks.
Thank you, Wamsi. I wanted to say that certain information in this presentation contains forward-looking information, including future-oriented financial information and financial outlook. These statements are not guarantees of future performance, and such forward-looking statements necessarily involve known and unknown risks and uncertainties, which may cause actual performance and financial results in future periods to differ materially from any projections of future performance or results expressed or implied by such forward-looking statements. The company undertakes no obligation to update forward-looking statements if circumstances or management estimates or opinions should change, except as required by applicable securities laws. Thank you.
Thank you so much. Well, to kick it off, maybe, Charles, good to see you. I know it's been a few years, but you look great. I wanted to maybe first ask how Super Micro is different from a few years ago. Maybe you can help the investor base first, ground them with what's been happening.
Okay. Thank you, Wamsi, and thank you to all the investors and potential investors. I'm very happy to join this conference. First, what's the difference with Super Micro today and say three years ago or five years ago? The major difference are, number one, our scale, business scale is much bigger now. Before we are a $2 billion, $3 billion company in terms of revenue. Now our run rate is close to $4 billion. We have the capacity to support $6 billion-$7 billion any time. The capacity is ready. Second, before we were a Silicon Valley-based company, and now we are still a Silicon Valley-based company, but also an Asian-based company. We have a big campus in Asia now. Our capacity in Asia now is bigger than our capacity in U.S.A. We also have a capacity in Netherlands. In Netherlands, we continue to grow as well.
The scale is much bigger. Now we have a Taiwan campus, which can produce our U.S.A. technology in Taiwan with a lower cost, higher volume, and slightly better cost for our large-scale customers. The other difference is software and firmware. Before we say we are a hardware company. Now I would like to say we are a total solution company. That bases the hardware, and then very fancy, reliable, powerful firmware and software for BMC, for system management, and for other operating systems, as well as some applications . Also service. Again, before we are a manufacturer, engineering company, and manufacturing company. Now we are a service company as well. Super Micro have a global service function now. We are able to help customers from concept discussion, brainstorm, design, manufacture, and service, including deployment and other services.
Super Micro today is a total solution company instead of five years ago a hardware company.
Wow, that's a lot of changes that you have brought about. Congratulations on achieving such a larger revenue scale from a few years ago. Charles, can you discuss Super Micro's building block solution? Why is this important?
Okay. Thank you. Building block solution, that's a very unique design methodology we have. With a building block solution, we are able to design products faster than our competitors because we leverage all our existing subsystems. When customers discuss with us for certain application optimize, we select pretty much most of our subsystems from our existing on a shelf subsystem. In some cases, we do redesign some modules, fully optimized for customer's exact demand. In other cases, we don't have to. Just select a different module, and then we can make application a truly optimized , complete solution, including rack- scale plug- and play. Handle all existing subsystems, and we make rack- level plug- and- play. We ship to customers, and customers just power on, and then connect to a networking, and then it works. For example, recently we worked with Osaka University in Japan.
It's a completed HPC, including liquid cooling, including lots of GPU, for example. We optimized liquid cooling, PUE at 1.06. We ship a complete rack to them, it's ready to power on and ready to run. With a building block solution, it helps us offer a solution to customers much quicker. Specifically, customer workload and application optimize, and it's kind of including service. Building block solution helps us a lot, and also helps us lower our inventory, because across different product lines, even different generations of products, we are able to share the same subsystem. In reality, that help us lower our inventory, and also makes our spare parts customer on-site spare parts much easier. Building block solution do really help us a lot.
Yeah. No, that's great. Can you maybe, Charles, just talk about a little bit about the customer base, how that has changed, and how much of your revenues now are these custom boxes, the custom servers that you're making for maybe hyperscalers versus ex-hyperscalers, or any sense of how much is going towards satisfying cloud-based demand versus not?
Yes, very good question. We have some large customers since a long time ago. Because before our Silicon Valley operation base, too expensive, we are able to support Tier 2, Tier 3 enterprise and channel, but not really Tier 1 customers, because Silicon Valley base operation costs too much. That's why we spend five to seven years to develop our Taiwan campus. Now the campus is fully ready. We have about 3 million sq ft facility there in the same campus. We are able to efficiently service Tier 2, Tier 3, and Tier 1 as well now. It's our choice. Depending on our resource, Tier 2, Tier 3 is still our preferred. Enterprise, channel, still our focus. We started to support some Tier 1, because our capacity is much bigger now.
In terms of application of vertical, AI, 5G, Telco, embedded, appliance, or just a regular data center, we support all of them.
That's great. With this expansion that you're talking about in Taiwan, would you say that your cost basis and cost structure can match some of what the Taiwanese ODMs are doing?
Yes, we design in U.S.A., now indeed we have a big design team in Taiwan as well. Take the background, the foundation we have here developed in Silicon Valley, we extend the design team to Taiwan as well, then manufacture in Taiwan. Our cost is equal to our other competitor, especially our scale, have been moving very smoothly, very well now. We know we sell $3 billion per year and $6 billion per year, the cost will be different, our operation in Taiwan will indeed be quite lower than Silicon Valley. We take both advantage, Silicon Valley-based technology, quality, service, but Taiwan manufacture cost.
No, that's fantastic. I remember that you always had a very strong both technological focus and focus on green computing. How has that differentiated versus competition? It's funny, as I was thinking about this, I remember that there was a green tie that I still have from Super Micro from years ago.
Yes. I like green, simply, because I grew up in a former country surrounded by forests. I like green. Most important is that green computing helps customers to save energy costs. From our calculation in detail, if global IT all goes for green solutions deeply, together, we can save about $7 billion yearly of energy costs. That's a big saving for customer. Second, because we have been focused on green computing for almost 15 years now, so we have a very solid foundation, very optimized, very cost-optimized green solution. Indeed, the green solution won't cost the customer too much. We're talking about ROI. If customers go for a deeper green computing solution, how much do they have to pay for extra, and how long can they return. Indeed, the return time can be very, very fast now. Before, people said two years, one year.
Recently, my announcement in Computex. I said ROI can be almost real- time, kind of immediately. Once you decide to go green, the return already happened. When we approach green computing, when we offer a really good green solution, we save customers, for example, their data center air condition capacity from, for example, 500 tons- 100 tons. To save 80% of the air condition capacity and shrink their data center size, because higher computing density, and also save their UPS capacity. Lots of savings can be realized from the beginning. Because we have been focused on green computing for so long, our green computing design
The cost, even the material cost, the delta is very, very minimal. Customers don't pay extra, but they have an energy- saving advantage, and they reduce carbon footprint, and they feel happier because they contribute to a more beautiful, less polluted planet. Without green computing, indeed, I don't know how I can find so much energy every day.
No, that's great. I wanted to ask you guys about supply chain shortages. It feels as though we hear this from every company on their earnings calls, that they're dealing with supply chain shortages, as it pertains to many different components. How is Super Micro navigating these shortages?
Maybe you want to start, Patrick?
Certainly. Wamsi, we have good relationships with all of our suppliers, and those relationships don't just span one or two years. They go back, in many cases, over 28 years. Therefore, we rely on our suppliers as partners during this process. Therefore, having good relationships helps. We also have maintained a really high level of inventory to service our customers. Are we facing shortage situations? Absolutely. We're navigating through those, and we had a midpoint last quarter in Q3 of $830 million, and our final revenues were $896 million. We're managing through.
What components would you say, from both an inventory perspective and from a procurement perspective, are somewhat challenging?
I would rather say IC and a chip because of the global semiconductor capacity limitation. Now with COVID-19 in Taiwan, become very serious as well. We are very carefully watch all of those. I believe with our long-term partnership, and everything planned in the long-term, our impact, yes, some, but hopefully will be continued under control.
When you say ICs, are you talking about processors, memory? Are you talking about anything more specific that you can share, Charles?
Yeah, in a system, multiple backplanes and other control circuits. Every company uses lots of different kinds of IC, from CPU, memory, and FPGA, to small controllers, you name it. Each system including more than 100 different ICs or components. It is a complicated matrix, but we manage, I would rather say, pretty well.
Okay. Can you maybe talk a little bit about, you mentioned sort of some of the impact, that you have to be very careful now, even in Taiwan, from a COVID perspective? We've heard from a lot of companies, not only just escalated costs because of freight and other logistics reasons. What are you seeing in real time? Are you seeing things improving? Are you seeing things deteriorating? Just from a supply chain and component availability perspective, and also from a cost, freight, and shipment logistics cost perspective, can you give us some sense on what's actually happening on the ground?
Yeah, it's really complicated now. It's a dynamic change every day, including recently, we heard a couple of IC packaging manufacturers have some COVID-19 cases happening every day now. We are very carefully watching that. For ourselves, in our campus, in our office, we have very strict control. So far, we still have zero infections in our Taiwan campus. Even in the U.S.A., in our campus, in the last 15 months, we control zero infections from our campus in U.S.A. So far, we are very lucky, and we will continue controlling strictly.
Okay. Well, thank you for that. Can you talk about a little bit, Charles, you alluded to how you've seen so much growth over the last few years, and your capacity is there for supporting even more expansion? How would you characterize your market share today, and how would you say that is likely to change in the next few years? I don't know if you want to take it, Charles or David.
Maybe I can start a little bit. As you know, we have been suffering a lot because of Silicon Valley- based higher costs, and our revenue was small, $2 billion, $3 billion before. That's why in the last five years, we have spent a lot of effort to grow our capacity, including extending our capacity to Taiwan, and including developing a customer base to support our scale. We have been achieving very well in this area . Last quarter, I mentioned about our $10 billion plans, and now we are executing that plan very smoothly. I would like to say we should have a chance to pull in a schedule, like a day out last quarter. We should be able to reach $10 billion faster than what I promised last quarter, in other words. Last time I said five to six years.
I believe we can do it much quicker.
That's pretty impressive. When you think about your competitive landscape, how would you say that has changed over the last few years? Certainly, you're talking about your own cost structure improvements that probably put you much better cost structure-wise with some of the Taiwanese ODMs. Would you say that is maybe the main competitor that you think about? Or who are you thinking about as your key competitors here?
Cost is just one factor. Indeed, to myself, I truly believe our real advantage is still technology, better design, better application optimization, and also a stronger software firmware team. We are able to work with Tier 2, Tier 3, or even Tier 1 to improve their data center, to improve their infrastructure, and make sure their data center are able to outperform their competitor. That's the major advantage we have. Doesn't matter AI, GPU, or HPC, or kind of hybrid cloud. We have invested a lot in the last five years. Just as a software engineer alone, last two years, we doubled our software headcount. This is a kind of a very important resource. Now we are ready to work with customer for optimized total solution.
Okay. That's helpful. I remember in the past you used to work extremely closely with Intel primarily, and you were able to sort of give them the initial boards to test out their chips. Can you talk about now you're offering obviously AMD, NVIDIA, how has your sort of partnerships evolved? What sort of mix are you seeing across these customers? Do you think that there's also room for ARM-based servers in the future, and any thoughts on that?
Yeah, very good question. We work with Intel very closely since 27 years ago, and then in the last few years, we also worked with AMD very closely. NVIDIA now, we have a very higher GPU AI architectural market share, and we continue to improve our relationship because we are very close to them all three, and that's why we prepare for a $10 billion scale very soon. As to ARM, whenever our customers need, we will be ready for them, especially with our building block solution. The same building block solution, I feel, we are very lucky. The building block solution we developed is good for all our different solutions, whether it's for computing, for storage, for AI, or for Telco.
By the way, Telco recently, we just gained some very good territory, some really strong players. The relationship is building up very solid now.
No, that's great. When you think about this goal to get to $10 billion, where do you think that growth is going to come from? Can you talk about what that plan is where you get to that sort of revenue scale, and what is going to be driving that?
There are a couple of areas we can grow. One is continuing to grow today's customer base in this same segment. Even enterprise, there are still lots of new enterprise customer like to work with us, and we are happy to embrace them even in especially Tier 2, Tier 3. There's still a lot of opportunity for us. We also started to work with a Telco 5G partner, and we found it's really a good vertical for us. Tier 1 data center and OEM, yes, with our Taiwan capacity now, at least some of the Tier 1, before we had to say no, now we have embraced them as well. The only limitation is our capacity still. Although we have a much bigger capacity, if we support Tier 1 aggressively, I think the capacity will be used very soon.
Okay. That's helpful. Maybe one for David. David, can you talk about sort of your overall balance sheet, what your cash position looks like? What are your expectations around the growth of free cash flow?
Certainly. We ended last quarter with about $180 million of cash, and our business has no problems generating cash. In fact, in the last three quarters, we returned to shareholders $118 million through share repurchases. We took four million shares off the market. In addition to that, we also increased our inventory levels, and we did that in response to the supply shortages as well as to increasing demand. When you look at our balance sheet, we've grown up to a $2 billion balance sheet. Not through debt, not through merger, but just through organic growth. That's been our legacy. Our balance sheet has very little debt. In fact, we use debt just for working capital requirements, and so we think we have a very strong and healthy balance sheet.
Would you say that in some ways, when I hear Charles talk about his aspirations here to grow the business as much as he's talking about, from that perspective, it looks like the equity is quite cheap? Is there any interest in levering up the balance sheet to buy more stock?
Well, that's a good question, Wamsi. I think that the board has considered that, and in fact, they authorized a $200 million share buyback. We initiated the first tranche of that in January, and it was $50 million. That's certainly something that the board has looked at closely.
For the level of investment, maybe, Charles, you said you're investing quite heavily into hiring software engineers, and that's been a big focus. Can you maybe expand a little bit on that? What specifically are you hiring these software engineers for from a solution perspective? What are your customers asking for? What is the capability that Super Micro is going to have in the future because of these software professionals that you're hiring?
Yeah, because it supports Tier 2, Tier 3 enterprise, Tier 2, still our main focus, right? Tier 1 data center, once we have capacity, we can grow that area. To support those enterprise Tier 2, Tier 1, and Telco company, they really need our total solution in terms of firmware, kind of like a BMC, supply chain security, for example, and also some application patch. Their whatever operating system or application, we work with them closely to make sure we provide them a plug-and-play solution. Software matters, who are invest their own firmware engineer, software engineer to study the interface between their environment and our machine. That's really a good value for our customers. As I say, we still have a lot of room to grow enterprise Tier 2, Tier 1, Tier 3, Tier 2, including Tier 1 as well.
Those are the software firmware team, which we invest in the last five years, is very helpful, and that grow our value a lot.
Okay, thank you for that. I have a question from an investor here who's tuned in, whose question is really: What steps are being taken to ensure that there is support and the right level of corporate infrastructure to achieve these growth targets? In terms of systems, customer service support, FCPA compliance, all these things, what steps are you taking to ensure that all of these are in place to support the level of growth that you're targeting?
Do you want to start?
Sure. Wamsi, we've invested over the last three years. You saw our revenues kind of level off, but actually, in the background, we were making a lot of investments. Two years ago, right about this time in May, I was over in Taiwan for groundbreaking on the new campus, which is just being completed now in the first part of July. In the same way, over the last three years, we've added about 100 people to general and administrative functions. We've added an internal audit department. We've added a compliance department. We've added more people to our finance department. Really, across the board, we've added additional software and systems and procedures in order to really get us ready for the next stages of growth. Remember, we grew between 2010 and 2018.
We grew at a 21% compound annual growth rate. We know that we have to keep making investments in infrastructure to facilitate our growth, and that's exactly what we're doing and what we've done.
Let me add a little bit. Indeed, in the last five years, especially the last two years, we invest a lot on SAP. SAP, our accounting, our operation software, especially kind of operation automation, including B2B, B2C auto configurator, and also customer order online. The system have been greatly improved and about ready to open to more and more customers, will be open to every customer very soon . The other one is service, a kind of global service network, global service talent, training, including on-site material, a kind of hub. All of those have been well established.
Okay. No, that's great. Appreciate that perspective. Maybe just to talk a little bit about when you think about your footprint out in Asia now that you're expanding. Are you worried about the last administration, we saw some tariffs being put in place, and other structures like that pose challenges. When you think about your footprint in Asia, how concerned are you that you might be subject to some of these various things like tariffs or other issues, particularly when the U.S. seems to be very focused on driving manufacturing and everything here in the U.S., and now you're talking about this administration now maybe, especially in semiconductor manufacturing anyway, continuing to want to invest in the future quite aggressively.
Do you see a path where this footprint is not just sort of Taiwan-centric, but if needed, you expand here in the U.S., or do you think that the Taiwan capacity is really the place to be to support this future demand?
Very big question. For sure, we have a very aggressive expansion plan. I shared with our senior management, hey, because of green passion, the passion for green, I like to grow the company bigger, quicker, and however, in the next few years, I believe it's time for Super Micro to grow our value. Grow our value for shareholders, for employees as well. We will continue to invest everywhere, but at the same time, we will focus more on how to grow company value for shareholders and employees as well.
Talking about other territories, Taiwan, we started to focus on Taiwan kind of seriously about five years ago. Now we have a very good market share in Taiwan, Japan, Korea, same same. We getting a bit more people hire and trained there, and Europe, and the East Coast, especially. Now, Patrick was assigned as Super Micro East Coast President.
We are ready to service our potential customer on the East Coast. I like the East Coast, people, company, and territory. We have a very aggressive plan. Patrick, you add a little bit.
Sure. Charles. Back to your original question on capacity. Not only are we adding a building in Taiwan right now, we have one coming online right here in San Jose. We're continuing to invest capacity resources right here in Silicon Valley. We have capacity out in the Netherlands as well, and that's an area that we can flex up if we need to. The way that we see it is, it's a global network of capacity. We are positioned to serve our global customers wherever they need to ship, wherever the data centres are, wherever the towers and 5G deployments they need. We've got an opportunity there. Just to add on to what Charles was saying earlier about East Coast, it's a tremendous opportunity for the company. It's an area that I just see blue skies there.
Really, what we want to do is replicate the success that we've seen in California over on the other side of the country. It'll be an exciting couple of years going forward.
Yeah. No, thank you, Patrick. We wish you luck with that. Unfortunately, we're just about out of time, so maybe just for closing, Charles, maybe you want to address the investors who are tuned in and talk to them about why they should be excited about an investment in Super Micro.
Thank you. Thank you once again, thank you all investors and potential investors. I'm very excited. After 10-K delay problem, finally 100% fixed, and then COVID-19 USA is kind of much better under control now. I'm very excited to grow our business. Now we have capacity, and we have infrastructure, including SAP operation and business automation and service network already. We are a USA company. We have lots of products made in USA. I got to say, we have the highest percentage of server storage products made in USA compared with any of our competitors.
We are very excited to grow in USA and ready to grow in Asia with big scale, Europe, East Coast. I feel very excited. Green computing not just helps customers save energy costs, but indeed kind of make our earth more beautiful, more green, simply. I'm very excited.
I sleep less hour every day now. Other than getting old, maybe don't need that much sleep, really very excited every day. Thank you for the opportunity. Thank you, everyone. You have a good day, everyone.
Thank you so much. Thank you, Charles. Thank you, Patrick.
Thank you.
Thank you so much. Appreciate it.
Thank you.