Great. We'd like to get started. Thank you so much for coming. I'm Toshiya Hari. I cover the semiconductor space at Goldman Sachs. Very excited, very honored to have Gianluca Romano, Executive Vice President and CFO from Seagate, with us this morning. I will kick off with questions, and I failed the last fireside chat, but I will try to leave room at the end for questions from the audience as well. Gianluca, first of all, thank you so much for coming.
Thank you, Toshiya.
I wanted to spend most of our time on longer-term strategic topics, but I did want to kick off with a near-term question. I think you guided September quarter revenue down 3% sequentially, after posting, you know, June quarter that came in below the midpoint of your guidance. What were some of the puts and takes you considered when constructing your calendar Q3 outlook, and how has the quarter progressed so far?
Thank you. Before we start, as usual, I will be making forward-looking statements today, and you can learn more about the risks associated with those statements on our website. Yes, so when we guided for the calendar Q3, now, what we said is, we expect mass capacity to start to improve, and that will offset the decline in legacy that was expected for the calendar quarter. So we were expecting hard disk drive revenue more or less flat, and we were also expecting a decline in the system business, so outside our hard disk drive revenue, and we guided basically $50 million sequentially lower revenue. At this point in the quarter, I would say mass capacity is slightly improving. Legacy is doing a bit worse than what we were expecting, mainly because of the situation in China.
China, I'll say we have basically all the segments in China, from China Cloud, China Enterprise, Video and Image Application, Mission-Critical, consumer, and client. I would say Video and Image Application is doing fairly good, so is sequentially improving as we were expecting, and this is a major driver of, you know, mass capacity, now slightly improving, sequentially. But all the other segments are actually very slow, and, and the expected revenue at this point is less than what we were expecting at the time of the guidance. So I would say, we are probably going toward the, low end, low end of our guidance range, for revenue and probably also for EPS. At the same time, we are very encouraged with discussion we are having with the, cloud customers.
It looks like the situation is improving in the cloud space, and we are also encouraged by the actions that are going to be taken by the Chinese government to support the local economy. So hopefully, this will also result in a faster economy in the next quarter or two. So between the cloud and hopefully a better recovery in China, we think this will be positive to our December and March quarters. But as I said, in the short term, probably no calendar Q3 towards the low end of the guided range.
Okay. And just to clarify, Gianluca, so mass capacity improving sequentially, largely in line with what you were contemplating, but the legacy-
Yes
portion of your business is tracking below?
Yeah, I would say, mass capacity except for the China Enterprise OEM, that is also a little bit impacted by the slow economy in China. The other segments are actually evolving as we were expecting, including Video and Image Application.
All right.
But it is a bit of an exception for the business we have in Asia and in China in particular.
Okay. And then, again, just to clarify, revenue and EPS tracking toward the low end is kind of the message?
Yeah, this is what we expect now. Of course, we're still a month-
Sure
to go through the end of the quarter, so could be maybe a little bit higher, but right now this is what we expect.
Got it. Very clear. Thank you. In terms of how to think about nearline demand going forward, obviously, it's going through a pretty big correction as we speak. I think it was down more than 50% year-over-year in the June quarter. To what extent do you feel like you're undershipping vis-a-vis end demand? And do you think the current correction is purely a cyclical dynamic, where customers are digesting inventory, maybe a little bit of budget sort of shifts, with AI becoming a big theme? Or should we be worried about this being more of a permanent shift in how customers think and spend on storage?
Yeah. Yeah, I say, right now it looks like we are undershipping. We see inventory consumption in the cloud space, in the enterprise OEM space. I would say enterprise started earlier, so there is less inventory to really consume. Cloud started probably in the June quarter, the big inventory correction. So, we did already a good step in the June quarter. I think we are doing another good step in the September quarter, and as I said, we think starting December, situation will start to improve. In terms of, you know, overall dynamics in the business, this cycle is different than what we have seen before. The main reason is, there is not only one factor that is driving the down cycle. It's not just an inventory correction. We had many different factors. We had.
Now, inflation, we have the war in Ukraine, we have the COVID-19, we have the fourth phase of recovery out of COVID-19, that is still strongly impacting China, and we have the inventory correction. So in general, now, technology business has always different phases, no? We have phase of expansions and phase of consumption. Of course, we are in a big phase of consumption, but there are also those other factors that are keeping this down cycle longer than usual. So as every cycle will end, and we think it's starting to end already, you know, in the December quarter and March quarter, as I said before, we see, we see improvement coming, but it lasted many, many quarters, probably double a normal cycle. And this has brought some different actions, now, that companies had to take during this prolonged down cycle.
But again, we are, we think we are coming out from that cycle fairly soon, and we will enjoy the up cycle, you know, in the next two or three quarters.
Okay. And Gianluca, can you maybe talk about the visibility you have into customer inventory? Is that something that you have a pretty good kind of pulse on, or is it extremely difficult to, to tell how utilized or not utilized your, your drives are, particularly at the cloud customers?
Yeah. I'll say, it's not difficult, you know, because we have a little bit of visibility on their inventory. The most difficult part is understanding the utilization, because, you know, a lot of what we call inventory digestion is actually not hard disk that have not been already implemented in various data centers. They're already there, but they're not fully utilized. So for us, it's very difficult to understand if, you know, a specific data center has hard disk utilization at 40% or 70% or 90%. So that is a difficult part for us to estimate. Now, what has been shipped and still not installed, that we have a little bit of visibility, of course, talking to our customers.
Okay. But to your point, you see the December quarter for your nearline business improving sequentially?
Yeah, no, and I said before, enterprise OEM started earlier, so it's more of a gradual improvement that already started. But cloud, you know, really declined the exabyte need in the June quarter. We are keeping that low in September quarter, but there are good conversation. I think December quarter will start to improve.
Okay, got it. You talked about VIA in China being pretty healthy, you know, within overall China that's still-
Yes
... relatively challenged. I guess, September quarter and beyond, what are your expectations for the overall VIA business, within mass capacity?
Yeah, we said in the past, and I think there are also no external company, you know, analyzing that business. I think there is a fairly good consensus on a CAGR, exabyte CAGR, in the mid-teens. So it's not growing as fast as cloud, but it's still a very healthy growth. So it will be one of the segments that, you know, Seagate will really focus on for the present and for the future. So Cloud, Enterprise, OEM, Video Image Application, those are the three major segments that we will continue to focus and serve, also with our new drives, now based on the HAMR technology.
Got it. We'll definitely get to, to HAMR.
Yeah.
So, so historically, you've said that exabytes in your mass capacity business should grow at a CAGR of roughly 30%, give or take, I, I believe. Is, is this still the relevant long-term through-cycle target, or given what you've gone through over the past, you know, six, eight quarters, is that number a little bit lower, a little bit higher? How do you think about the through-cycle growth rate of your mass capacity business-
Yeah
...a nd exabytes?
Yeah. I would say, you know, you, you always need to take a certain number of years to identify the right CAGR. So you need to go through the phase of expansion and the phase of digestion. And, you know, in the past, we had a year where we were growing, 50% or 60%. And of course, we had last fiscal year, where actually the growth was negative. So when you put all together, you know, the up cycle and the down cycle, now, what we expect is probably in the, you know, mid-20s%, and so a little bit lower than what we were saying before. Now, that is not maybe 30%, it's more, you know, 25%, but it's still a very, very healthy growth.
Right.
Again, don't expect a linear growth of 25% every year. You will have years, especially at the beginning of the up cycle, where we grow much more, and then we will have phases of no consumptions, where, you know, the growth will be much lower.
Yep, makes sense. AI is a big focus for investors. I think the compute segment tends to get a lot of the mind share. Based on, you know, what you're seeing, what you're hearing from customers, what are your thoughts on AI's impact on storage? Do you expect it to drive an acceleration in your revenue growth, and if so, around what timeframe?
Yeah, absolutely. So AI has two main phases, you know, the training part, where AI just accesses data that is already available. So it's not creating new data, but is, you know, accessing data that is available in order to then create something new. Because AI needs information, needs data, data becomes more valuable. So data that maybe today could be deleted, in the future, you know, people and companies will probably keep it for longer because AI will have to access all that data to create something that is even more valuable. And then there is the second phase, that is the one that will impact the most storage, that is what AI will create from accessing all that data. So it access and create something new.
If that something new is something valuable for, for individuals, for companies, that will be stored, and that, that is a new data that needs to be stored, and of course, consume, data storage and create an additional need for data storage. Now, AI today is still in the first phase. You know, it's just assessing the data, it's not really creating a lot of new things. People are still learning how to use it. Companies are still learning how to use it. I would say in, you know, in one or two years, AI will be a normal tool for companies to improve their performance, and so they will have to store much more of what AI is creating. So I think it's very positive for, for storage.
Of course, in the short term, it's more positive for companies that are selling different components because, you know, the big clouds have to get ready to use AI, and so you need different components, right, to create the infrastructure. But once the infrastructure is ready and AI get up to speed, the result is more storage and more data to get stored in the cloud.
Got it. So when you're deploying models, when the inference is going on, that's when you start to see an impact-
Yes.
-on storage?
Absolutely.
Okay. Makes sense. The competitive sort of situation vis-à-vis flash is a topic-
Mm.
-that we get questions on. There, I guess, are a couple of storage OEMs who are very vocal, very adamant about this transition away from hard disk drives to enterprise SSDs. I know it's been ongoing for a very long time. And I guess it's encouraging to see hard disk drives. You know, you're going through a correction, but I think on a relative basis, you're actually holding up better than SSDs, despite NAND flash being at a price level that is, frankly, unsustainable. So I guess that's encouraging, but how do you internally at Seagate think about kind of the push.
I agree with you. I would say, first of all, now you have seen different trends between the two technologies, you know, in hard disk, despite this major down cycle. You know, company are still generating a decent gross margin. If you look at the NAND business, that margin is negative. So there's a huge difference on how the two businesses are managed and the results they can provide, not only in the upcycle, but even in the downcycle. Of course, this is a matter of consolidation and how the companies in the industry, you know, are driving those businesses. Second is, as you said, even if the NAND price is very, very low today and they generate negative margin, but there's been no change in the structure of the data center.
90% of the data was stored in hard disk before the down cycle, and it's still 90% today. So no change there, I think, is a confirmation of what we are saying before. In the data center, NAND and hard disk are used for different parts of the application. So they will both continue to grow, but they will not overlap. You know, data center, you know, is separating the storage, that is hard disk, and the support to analytics and compute, that is the NAND part. So two different parts of the big data center. Hopefully, they will both be very successful, but we don't see any overlaps there. There is, of course, overlap in low capacity.
So if you look at the consumer, the client, the mission-critical, so where you have hard disk that are, you know, 1, 2, maybe 4 TB, there is for sure overlap there. And no, we have been very clear since several years. We know that part of the business, that we call legacy, is going to decline year- after- year, and that is the part that can be replaced by the NAND longer term.
Got it. On your most recent earnings call, you talked a little bit about raising pricing in select markets. I was hoping you could expand on that. What catalyzed this, you know, shift in—I guess, not really a shift in strategy, but what led to this decision to raise pricing? You know, are you seeing good traction with your price changes? And how should we think about the risk around market share, to the extent your competitors don't follow suit?
Yeah, I don't know if that is a fair assumption. I think the industry has the same view and the same need-
Yeah
- to generate a certain level of profit, a minimum level of profit. So, you know, I think, it's not only our desire to, you know, generate that level of profitability. So the need of increasing pricing, I think, will be more an industry need. We will know, fairly soon, in the next quarter or two, but this is my expectation. Our customer needs to understand, you know, the customer base is fairly concentrated in storage, especially in the cloud space, but the supply base is even more concentrated. So if you look at nearline, there are basically two companies. Toshiba is very small in the nearline space, has less than 10% of market share, so it's very consolidated.
As I said before, we need to generate a certain level of profit because supply chain, our supply chain, is also in a need to be supported. So there are many companies, you know, that need to generate a certain level of cash flow, and of course, we need to help them with that cash flow, and at the same time, our customers need to help us.
You know, we think we arrive at a time where it was necessary to increase a little bit our pricing, especially right now, because, you know, we see the end of the cycle coming, and we want to be ready at the end of the down cycle with a pricing that is a little bit stronger, so we can generate better cash flow when we go into the upcycle and volume will be higher of what it is today.
Your message about supporting the supply chain certainly makes sense, and we talked to a bunch of companies that make up that ecosystem. Does that message resonate with your customers? When you go to your customers and say, "Hey, we need to support our, you know, small suppliers in Southeast Asia, give us a break." Do they say, "Understood, let's take up pricing a little bit," or is it tough in this environment?
I would say the reaction is not always the same.
Yeah.
But our customers that now have worked with Seagate for, you know, 10, 20, 30 years, and they probably understand that. Customers that are more new into the space, they probably need to understand better to be better. So again, in the short term, you, you always have a little bit of pushback when you raise price, but you need to keep the position, and, as I said before, you know, the industry is concentrated, is consolidated. I think it's the right thing to do for, for us, for the industry, and for our suppliers.
Makes sense. You also talked about this concept of build-to-order-
Yes.
-on the earnings call. I think from wafer to shipment, we're talking months, sometimes quarters, in terms of lead times for hard disk drives, right? We're not talking days, we're not talking weeks.
Yeah.
So when you say build to order, what's the starting point? And if I'm a customer, and place an order today, when would I get my set of drives under that sort of?
Yeah, well, if we really start from the wafer, the cycle time is, you know, at least six months.
Right.
And depending from what drive you want to buy, it can be even longer. So this is, this is another improvement we want to have. Now, we want to have a little bit more certitude on what we need to start. Of course, when you are in an upcycle and you see volumes going up, you can build more into your inventory and then, you know, define the right mix through the end of the quarter. Right now, it's more difficult because, you know, we don't want to create inventory ourselves, we want to preserve cash, so we need to have a good understanding of what we need to start. What is the product that the customer want in three, six months from now?
Now, in the past, it's been, it's been a little bit difficult, and, you know, the real order confirmation always arrive a little bit late. So we had to build based on a forecast, but it was not an order, and if the forecast change, we need to manage that inventory, and if at quarter end you have excess inventory, you need to move the inventory, that, of course, not helping the pricing situation. So we need to change, and many, as many other, you know, technology industry, we need to get orders, and when we get the order, we will deliver exactly what the customer wants at the time they want. But, it's not the time right now for us to speculate on what they will want in three or six months from now. We need to have a little bit more visibility.
Okay. So is it fair to say that you've sort of introduced this long-term order kind of concept, or am I interpreting-
Yes.
...i t the wrong way?
No, I think it's correct.
Okay. Okay. And again, the customers, is it resonating with the.
It's a new concept.
Even if it doesn't, seems to be, you know, very difficult to implement, you always get some pushback when you ask for, you know, longer term commitment, and it's a little bit like the pricing, there was a little bit of pushback at the beginning, but then it's just the right way of working.
Right.
Now, it's a good partnership, now they tell us what they need, we will produce what they need, and deliver what they need when they want. But, you know, always expecting Seagate, or in general, the hard disk industry to, you know, build wafers, and build medias, and wait to assemble until the last moment, I don't think is a good way to manage with the industry long term.
All right. Makes sense. Before we take questions from the audience, I just wanted to hit on HAMR. Obviously another topic that comes up a lot in conversations. Your engagements with customers seem to be broadening. I think a quarter ago or two quarters ago, you had talked about one specific U.S.-based cloud titan being qualified. Can you give us an update as to how things are going with HAMR?
Yeah.
How should we think about the ramp into 2024?
So qualification is progressing well. So, so far, we have not encountered any issue with the core. So it's progressing well. We will qualify other customers after this first one. And as we said, at the earnings release, we will start ramp volume in at the beginning of calendar 2024. So it's just four months from now, so we are very close to that time. And, no, but a little bit of time needed to ramp it up, but we expect, you know, a fairly good revenue already in the calendar Q2, I'll say 2024. So we will have revenue in calendar Q1, but more volume in calendar Q2.
Got it. Okay, great. I'll pause here and see if we have any questions from the audience. If not, I can keep going. Okay, so in terms of the supply side response, I think both you and your closest peer have been pretty proactive in taking supply offline. Can you remind us exactly how much manufacturing capacity you've taken offline? And I assume this is permanent, but talk about, you know, whether it be transitory or permanent as well.
Yeah, we took out about 25% of capacity. That means, you know, compared to the volume we are producing and selling today, there is still capacity available for the first part of the upcycle. But at a certain point, of course, depending from how strong is the demand, we will have to go back and spend a little bit of CapEx to create more demand if it's necessary. I would say there are two ways to reduce capacity. One is taking out equipment from the manufacturing, and one is reducing labor, and we did both. So unfortunately, we had to do many restructuring plans in the last few quarters, so our headcount reduced by about 10,000 people. This is about 25% of our headcount.
We also took out, of course, the equipment from the manufacturing floor. So at a certain point, of course, you can always hire back those people and reincrease part of the capacity, but no, it will take a lot of time to rehire thousands of people. We don't think we will need to rehire, you know, exactly the same number of people to go back to the same level of production. We can do it with a lower number of people. Now, we have identified opportunities, of course, for improvement during the down cycle and to be more efficient, so we think the cost structure will be better even at the same level of volume we had before. And of course, we also did a lot of restructuring in the OpEx.
Now, if you look at just a few years ago, our OpEx was about $400 million a quarter, and now it's almost $250 million. So a huge reduction. I don't think we will add OpEx when we go into the upcycle, but there will be, you know, one change, which is variable compensation, that today is basically zero. So hopefully, that variable compensation will come back. But that should be the only, you know, major item that will increase in our OpEx, not much more.
Okay, got it. And then in terms of gross margins, for the current quarter, I think you guided profitability to increase modestly from 19.5% in the June quarter. Perhaps that's a little off now, given some of your earlier comments. But what are some of the near-term puts and takes as you think about gross margins, and when do you expect underutilization charges to subside?
Yes. So underutilization charges this quarter will be higher than last quarter.
Okay.
Of course, with latest adjustment in revenue will not particularly help with the gross margin situation. But as I said, you know, mass capacity is starting to slightly improve. In the legacy, we also have a decline in mission-critical, that is generally good gross margin, so that will now be positive overall. But again, now I would say the most important is what is our feeling for the next few quarters, and we see that improving, and of course, gross margin will improve sequentially. We will have the impact of our restructuring plan, some impact already in the current quarter.
More in the December quarter is probably where we peak at the restructuring level in terms of cost decline, and that will stay with us for a few quarters until we really have to ramp more. So I would say there are positive things, both on the pricing side and on the cost side, so we should see an improvement in gross margin in the next few quarters.
Got it. Then specific to some of the pricing actions, is it possible to sort of quantify how much of a tailwind that could bring along over the next couple of quarters, or is that difficult?
Well, no, we are, we have decided to increase pricing in all the segments. Of course, it takes time, and, you know, every segment take a different time. So we started, of course, with the channel and the distribution, and then we discussed with some of our OEM and cloud customers, and of course, now the Video and Image Application customers. So it takes a little bit of time, but we have been very clear. No, it's not a one-quarter action. This action will continue because we need to bring that, the, you know, profitability that we need for us and for our supply chain. And so until we are there, we need to increase pricing, and we need to continue to reduce our costs, because that is, of course, part of how we manage the business.
Okay, so you were at 30% ±, in gross margins, not long ago, right?
Yep.
It was a year and a half ago.
Yes.
T wo years ago. You're currently below 20%. With the price actions, with mass capacity demand improving over time, and all the restructuring, you know, programs that you implemented, is there a good line of sight back to that sort of 30-ish% level sometime in 2024? Or is it a little bit too early to think about-
Oh, I think so.
from the outside?
I think we will be able to, you know, achieve that level of gross margin well before the level of revenue we had six quarters ago. So, you know, six quarters ago, we were at about, you know, $3 billion and a little bit higher than 30% gross margin. I think we can achieve those levels with a much lower revenue, because of all the work we have done on the cost, that will stay with us, and because of the, you know, pricing actions we are taking right now, and we need to continue to take in the next few quarters.
Got it. Maybe one last question on, on the balance sheet and, and the dividend. I believe you reduced debt on your balance sheet by roughly $500 million in the June quarter. You now have less than 12% of your total debt due in the next two years, so you made a lot of progress. Given the outlook that we've just sort of discussed in terms of the business, how, how are you thinking about the balance sheet?
Yeah.
How are you thinking about the dividend? If you can kinda share your thoughts, that would be helpful.
Yeah. In the last, probably three quarters, we have reduced about $800 million in debt, and, of course, has been a focus of the company. During, you know, going into this down cycle, debt was a little bit elevated, so we had to take actions and reduce the debt. Every quarter, we look at opportunity to reduce the debt or improve the structure of the debt, and now pushing out the debt in time. As you said, right now, we only have, you know, 10%-12% of debt outstanding in the, you know, in the next, you know, couple of years. But no, we always do something.
Every quarter, we look at opportunity and improve the structure of the debt, and you know, repay what is becoming short term and try to take care of of the balance sheet and keep it strong balance sheet. In terms of dividends, you know, even during the down cycles, you know, our fiscal year 2023, we generated about $600 million of free cash flow, and so this is above the level of dividends that we pay. So I would say the company has been very, very focused on generating positive free cash flow, even in this very tough environment. And right now that we expect, you know, gross margin to start to improve in the next few quarters, I would say, shareholder return has always been a focus of the company.
Of course, we had to stop doing share buyback at the beginning of the down cycle, but, you know, we are very focused on keeping the dividends, and this is a priority for the company.
Got it. In the last 60 seconds that we have, I know you converse with a lot of investors and analysts. Anything about the Seagate story or anything about the hard disk drive industry more broadly, that we collectively overlook or underappreciate in your view?
Well, the only thing I can add is that data is growing, so the need for data storage will continue to grow. There are cycles in this industry, and of course, right now we are through a fairly big down cycle, but hopefully through the end of this down cycle. As we discussed before, we're seeing the CAGR for the exabyte growth is very healthy and, you know, we will need to have a lot of focus on how we manage our own business in terms of cost, how we support our supply chain, how are we, you know, partner with our customers.
But there is no reason why the data growth and the need for data storage will be different, you know, in the next year or two or five or 10 years, of what has been in the last four or five years. Actually, probably will be much higher because of AI, because of all the new application that people and company will start to use. So we are very positive on the long term, I would say in the medium and long term, and we start to see sign of improvement already, so we feel more confident. And you know, as usual, we focus on free cash flow, profitability, and much less on market share. We focus on our roadmap and push our roadmap with our customers, because that is what is important.
HAMR is a way for us to generate more revenue at a bigger level of profitability and continues the evolution of the industry, and I think this is what we have to do and what we are focusing on.
That's great. Thank you so much, Gianluca. I really enjoyed the conversation.