Seagate Technology Holdings PLC

XBER:847

Earnings Call Transcript 2024-Q1

[Operator]

Welcome to the Seagate Technology Fiscal First Quarter 2024 Conference Call. [Operator Instructions] Please note this event is being recorded.

I would now like to turn the conference over to Shanye Hudson, Senior Vice President of Investor Relations. Please go ahead.

[Shanye Hudson, executive]

Thank you. Good morning, everyone, and welcome to today's call. Joining me are Dave Mosley, Seagate's Chief Executive Officer; and Gianluca Romano, our Chief Financial Officer. We've posted our earnings press release and detailed supplemental information for our September quarter results on the Investors section of our website.

During today's call, we'll refer to GAAP and non-GAAP measures. Non-GAAP figures are reconciled to GAAP figures in the earnings press release posted on our website and included in our Form 8-K. We've not reconciled certain non-GAAP outlook measures because material items that may impact these measures are out of our control and/or cannot be reasonably predicted. Therefore, a reconciliation to the corresponding GAAP measures is not available without unreasonable efforts.

Before we begin, I'd like to remind you that today's call contains forward-looking statements that reflect management's current views and assumptions based on information available to us as of today and should not be relied upon as of any subsequent date. Actual results may differ materially from those contained in or implied by these forward-looking statements as they are subject to risks and uncertainties associated with our business. To learn more about the risks, uncertainties and other factors that may affect our future business results, please refer to the press release issued today and our SEC filings, including our most recent annual report on Form 10-K and quarterly report on Form 10-Q, as well as the supplemental information, all of which may be found on the Investors section of our website.

As always, following our prepared remarks, we'll open the call up for questions. I'll now hand the call over to Dave for opening remarks.

[William Mosley, executive]

Thank you, Shanye, and welcome, everyone. Before I discuss our financial results, I want to acknowledge the situation taking place in the Middle East. Our thoughts are with all of the people in the region, including our Seagate team members, their families and loved ones.

Moving on to our September quarter results. Revenue came in at $1.45 billion with non-GAAP loss per share of $0.22. Consistent with our recent public commentary, we experienced softer-than-anticipated demand in the legacy markets, while the ongoing cloud inventory correction and weak economic trends in China continued to restrain near-term demand for hard drives. Looking ahead, we expect the pace of economic recovery in China to be uneven. However, we are encouraged by the positive progress of U.S. cloud inventory consumption.

Importantly, we continue to demonstrate financial discipline and strong execution on the priorities we outlined at the onset of this down cycle, namely to drive cash generation, strengthen our balance sheet and position the company for enhanced profitability as the markets recover.

We also continued to hit all key HAMR product development milestones, demonstrating our ability to drive significant areal density gains with this technology. These gains translate into lower storage costs on a per bit basis, enabling Seagate to offer a compelling TCO proposition for our customers while enhancing our future profitability.

Qualification and revenue ramp plans for our 30-plus terabyte products remain fully on track with high volume ramp starting early [indiscernible] is a competitive differentiator and increasingly important in light of the green shoots that we're starting to see with respect to cloud demand trends.

Within the mass capacity markets, we saw a modest uptick in demand for our high-capacity nearline products among U.S. cloud customers. We project incremental revenue growth from U.S. cloud customers again in the December quarter and are encouraged by constructive customer dialogue regarding our transition to a build-to-order model, making us more confident on demand fundamentals entering calendar 2024.

Additionally, industry analysis of cloud customer behavior suggests that their cost-optimization efforts are nearing a conclusion, while enterprises continue migrating new workloads to the cloud. These include both core IT workloads as well as AI-specific workloads. In addition to cost optimization efforts, spending priorities for CSPs have temporarily shifted towards AI-related infrastructure, which have further slowed the pace of demand recovery for mass capacity storage.

While AI-related spending remains a near-term priority, several cloud customers have indicated that investments in traditional servers and other IT hardware will resume in the coming quarters. All of these trends bode well for HDD demand recovery in both the cloud and enterprise OEM markets.

These same markets in China are lagging these early positive signals due to the regional economic conditions that I mentioned earlier. However, video and image applications were a notable exception, reflecting demand both within China and globally. Public and private investments in smart city and smart security projects have been key demand drivers for the VIA market. While we believe these underlying demand trends remain intact over the long term, the severe slowdown in China's property sector and broader global macro uncertainties are likely to temper demand over the next couple of quarters.

Near-term conditions aside, we are optimistic about the VIA market given the increasing use of AI and deep data analytics that enhance the effectiveness of VIA's systems. These systems are evolving from basic monitoring tools to more fulsome solutions incorporating advances, like high-definition AI cameras that offer more valuable insights and lead to longer data retention rates. These data-intensive solutions are well suited for hard disk storage in terms of cost, capacity and performance.

Looking back across our 45-year history, cost-effective, high-capacity storage has been vital to the enterprise's ability to harness the benefits of every generational technology mega trend that we have experienced. From personal computing to the Internet, mobile, to big data to the ongoing migration to the cloud. We anticipate the same will be true with the rise of AI and generative AI applications, which contributes to our long-term view for a return to healthy exabyte growth.

Seagate's mass capacity storage portfolio sets us up strongly with this growth backdrop. Last week, we announced our latest high-capacity nearline product boasting 2.4 terabytes per disc and leveraging our proven 10-disc platform to deliver capacity starting at 24 terabytes. We continue to offer customers the flexibility to deploy these drives as a conventional CMR drive or is a shingled SMR configuration based on their specific capacity and architectural needs. We are engaging with a number of cloud and enterprise customers on qualification and expect volume shipments to begin in the first half of calendar '24.

We also expect to begin aggressively ramping 3 terabyte per disk products based on HAMR technology in early calendar 2024. These drives deliver capacity starting at 30 terabytes and offer customers the same flexibility to adopt either CMR or SMR configurations to further boost areal density into the mid-30 terabyte range. Initial customer qualifications are progressing very well, and we continue to hit our reliability and yield metrics. We are getting extremely positive customer feedback, and we are broadening the number of customer qualifications as planned.

We've been very thoughtful in building our product road map to stage HAMR technology, leveraging existing product design and process commonality where possible. For example, virtually all of the capital invested for the 20-plus terabyte PMR drives is compatible with HAMR products. The 30-plus terabyte HAMR drives utilize many of the same components in electronics as our 20-plus terabyte products. They represent the fourth generation product using our 10-disk platform and the seventh generation that leverages glass substrates. These actions improve capital efficiency, reduce manufacturing complexity ensure reliability and hasten time to market.

While many aspects of our product design are evolutionary in nature, HAMR revolutionizes areal density advancements through years of persistent research and development investment, innumerable design iterations and optimization cycles across all elements of the drive from mechanical and electrical designs to wafer processing and firmware, we have now reached the appropriate balance between areal density gains cost optimization and reliability to launch HAMR in volume.

Our execution and cycles of learning have enabled us to continue strengthening our portfolio, and we expect to launch products yielding 4 terabytes per disc in less than 2 years' time, significantly differentiating Seagate and addressing the full spectrum of mass capacity demand.

Architecturally speaking, in today's data-driven business economy, mass capacity storage is a crucial tier. The HDD areal density advancements that we are delivering affirm and sustain the existing TCO advantages relative to NAND for mass capacity storage. Simply put, we offer customers mass data storage at less than 1/5 the cost of comparable NAND solutions on a per bit basis. We don't foresee that value gap closing over the next decade relative to data center architectures.

In addition to optimizing costs, customers are intensely focused on conserving data center power and floor space. Customers can realize benefits across each of these objectives by upgrading their existing installed base of HDDs to higher-capacity drives.

The 30-plus terabyte HAMR drives currently in qualification are more than 2x the capacity compared to the average installed base across large data centers. This HAMR-based upgrade would more than double their existing storage cost and saving or offer a 50% reduction in operating costs for the same storage capacity using about half the power and floor space. These are compelling savings for customers and offer valuable optionality to best monetize their storage assets or reallocate floor space and power budgets for other uses or even defer new data center build-outs to maximize their capital dollars.

As we deliver these benefits to our customers, we are also focused on capturing the value of our product portfolio. As noted on our last call, we are continuing efforts to adjust price commensurate with that value, which ensures both a healthy industry supply chain and offers customers the opportunity for improved TCO over the long term. We have already seen some benefit from this strategy, which we anticipate will take a few quarters to implement more broadly across the end markets we serve.

I'll now hand the call over to Gianluca for further details on the September quarter results and share our outlook.

[Gianluca Romano, executive]

Thank you, Dave. Seagate's September quarter financial results were consistent with our revised expectations. We generated revenue of $1.45 billion and a non-GAAP loss of $0.22 per share. Despite the sequential decline in revenue, we expanded total company non-GAAP gross margin by about 30 basis points and HDD non-GAAP gross margin by more than 130 basis points, reflecting our focus on enhancing profitability.

Within our hard disk drive business, revenue declined 6% sequentially to $1.3 billion, reflecting a modest improvement in mass capacity sales, offset by steeper decline in the legacy market than we had originally expected. The mix shift towards higher-capacity drives resulted in total HDD shipments of 90 exabytes, essentially flat with the prior quarter. Average capacity per drive increased 17% sequentially to roughly 7.5 terabytes per drive.

Mass capacity revenue increased 3% sequentially to just over $1 billion, driven mainly by the anticipated improvement in the VIA market. Mass capacity shipments totaled 79 exabytes compared with 75 exabytes in the June quarter. The mass capacity shipments as a percentage of total HDD exabytes were roughly 88%, up from 82% in the June quarter.

For nearline products, shipment of 56 exabytes were slightly up quarter-over-quarter. Average capacity per nearline drive increased 12% sequentially as demand trends among U.S. cloud customers began to modestly improve.

We believe that the industry continues to ship below end consumption and is making progress in reducing existing inventory at our cloud customers. As we mentioned last quarter, we anticipate that it will take at least through the end of the calendar year for inventory levels among CSP customers to rebalance and for demand to improve more broadly.

Specific to the VIA market, revenue was up sequentially as expected in the September quarter. However, as David noted earlier, the uncertain economic environment in China seems unlikely to change in the near term. As a result, we anticipate the VIA market will reflect an even pattern of recovery going forward.

Legacy product revenue was $278 million, down 31% sequentially, with lower demand in each of the 3 markets served, mission-critical, clients and consumer.

Finally, revenue for our non-HDD business decreased slightly more than anticipated to $159 million compared with $218 million last quarter. Reserved IT spending patterns in light of economic uncertainties remain a headwind to our enterprise system business, and we expect similar revenue levels in the December quarter.

Moving to our operational performance. Consistent with lower revenue levels in the September quarter, non-GAAP gross profit decreased by $25 million to $288 million. Non-GAAP gross margin of 19.8% expanded slightly compared to the prior quarter. Pricing adjustments enacted during the quarter and cost savings from earlier restructuring activities more than mitigated the 9% decrease in revenue and increase in underutilization cost, which were approximately $59 million. We expect to see further margin benefit in future quarters as we continue to execute price adjustment across the entire portfolio and achieve full realization of projected cost savings.

I note that, beginning with the September quarter, our results reflected change in the estimated useful life of certain capital equipment used in manufacturing. Our ability to increase the efficiency of our existing fixed asset base has enabled us to extend the useful lives from a range of 3 to 7 years to a range of 3 to 10 years. This change reduced depreciation expense in the September quarter by approximately $9 million within cost of goods sold and is expected to increase by about $20 million in the December quarter.

Reduced non-GAAP operating expenses to $248 million, down from $258 million in the June quarter. While we continue to actively manage all areas of spending, we do expect non-GAAP OpEx in the December quarter to be up slightly a certain minor spending reduction measures begin to conclude.

Moving to cash flow and the balance sheet. We have continued to take actions to improve our debt profile and manage working capital to support positive free cash flow generation. September quarter, we reduced inventory by 8% sequentially to just under $1.1 billion. Capital expenditures were $70 million compared with $50 million in the prior quarter. For the fiscal year, we are still planning a significant reduction in CapEx spend compared with fiscal '23, and expect spending will be more heavily weighted to the first half of the fiscal year.

Free cash flow generation was $57 million after giving effect to approximately $90 million of restructuring-related payments that we had highlighted on our last earning calls. We used $145 million for the quarterly dividend and exited the quarter with 208 million shares outstanding.

We closed the September quarter with $2.3 billion in available liquidity, including our undrawn revolving credit facility. During the quarter, we raised $1.5 billion in new capital through the issuance of convertible notes bearing a low interest rate of 3.5%. A portion of the proceeds were used to fund a capital call transaction that increased the effective convert price to nearly $108 per share, reducing potential future share dilution. The majority of the remaining proceeds were used to retire an outstanding balance on our term loans, which total approximately $1.3 billion.

As a result of this debt restructuring actions we expect to realize cash interest savings of about $15 million on an animal basis. Additionally, we renegotiated the terms of our sales agreement, and with support from our lender group, we significantly relaxed the debt covenants through fiscal 2025.

Accounting for all actions that I just described, our debt balance was $5.7 billion at the end of the September quarter, up $115 million quarter-over-quarter. Non-GAAP interest expense was sequentially flat at $84 million, and we expect similar expense level in the December quarter.

Turning to our outlook. We expect mass capacity sales to move slightly higher in the December quarter, supported by incremental demand for our nearline products from both cloud and enterprise customers, offsetting softer sequential VIA demand. Within the legacy business, we are projecting higher seasonal demand mainly from the consumer market, while non-HDD revenue is expected to be essentially flat.

With better context, we expect December quarter revenue to be in a range of $1.55 billion, plus or minus $150 million. At the midpoint of our revenue guidance, we expect non-GAAP operating margin to be in the mid-single-digit percentage range with underutilization costs expected to be relatively flat with the September quarter. We expect to narrow our non-GAAP loss per share to $0.10, plus or minus $0.20, based on a share count of approximately 210 million shares and a non-GAAP tax expense in the $15 million range.

I will now turn the call back to Dave for final comments.

[William Mosley, executive]

Thanks, Gianluca. We are operating in a longer-than-typical cycle, and I'm very proud of our team's shared determination and resilience. We've continued to drive our financial, operational and innovation priorities, which is evident by the actions we've discussed today. We are focusing our tactical business decisions on free cash flow generation. We are strengthening our balance sheet through debt restructuring actions. And we are executing on our mass capacity product road map to address future data growth.

Signs of recovery has started to emerge as we look past the end of calendar 2023. And as industry conditions improve, Seagate is ready to capitalize. We are a stronger, more efficient company with a technology road map that extends our areal density leadership, positioning Seagate to deliver enhanced value to our customers and shareholders.

Thanks to all of our stakeholders for your ongoing support of Seagate. Operator, let's open up the call for questions.

[Operator]

[Operator Instructions] The first question comes from Erik Woodring of Morgan Stanley.

[Erik Woodring, analyst]

So Dave, I kind of want to take a step back. And if we go back to your Analyst Day, the expectation was that mass capacity exabytes shipped could grow at a 35% annual rate. Now the world is clearly different today both from a macro outlook and the emergence of AI as kind of Board-level investment priority.

So curious, as we look forward, how are you thinking about kind of the long-term 3- to 5-year rate at which mass capacity exabytes could grow? And maybe help us think about the linearity of that. Could that strengthen over time? Is that relatively linear? Just any thoughts that you have kind of longer term on what mass capacity exabyte growth could be.

[William Mosley, executive]

Thanks for the question, Erik. Yes, we were coming off the back of a number that was almost 80% in 1 year. So to your point, back in the last Analyst Day, the mid-30s was feeling pretty good for us, given all the data growth that we know is happening. And then we've gone through these current events, I'll say, that we're all seeing.

Look, to your point about linearity, I don't think it's a very linear function. I think it can be very choppy. Up one year, down the next, it can be. As we look out 3 to 5 years now, I would temper that somewhat and say in the mid-20s is probably a good modeling range. But we will see probably more growth from time to time.

I do think that, given the move of -- that happened in the middle of the pandemic, of data into data centers for enterprise applications that were being run in data centers. I think that data is still going to grow in those data centers. So we're not at peak data center growth. And then you lather on top all these new applications that will be really sped up with features like AI and generative AI on top of them, I think that there's a big, healthy demand growth coming.

25% would be healthy for any kind of CAGR that's going to go out for 3 to 5 years or a decade or something like that. So we think there's healthy demand growth. But I do think it won't be linear. There will be periods of big growth and then there will be periods of digestion as well. We'll continue to see this as the economy flutters.

[Operator]

The next question comes from Wamsi Mohan of Bank of America.

[Wamsi Mohan, analyst]

Dave, can you talk about HAMR qualification? How many customers are you expecting qualifications in the near term? And maybe share color on what you expect in aggregate, given what you know about the ramp in terms of unit shipments, maybe 1, 2 years out?

And if you could maybe just comment on how this higher areal density shift might change the dollars per terabyte relative to current product. That would be super helpful.

[William Mosley, executive]

Yes. Interesting. I can take a crack at that. So relative to qualification, we are prioritizing customers, not necessarily the easiest customers first. Sometimes they can be more difficult, challenging customers. But we're staying very tight with customers, and we're trying to make sure that every of the initial drives that we build has a home. Make sure that we -- as we bring on more qualifications, like we talked about in our prepared remarks, that we will make sure that we have the supply that's adequate for those because I do think there will be a strong demand.

What we're showing customers right now, exactly to your point on the value proposition, is a projection for what their TCO benefits will be. That -- part of that is the acquisition cost of the drive itself. Part of it is the power and floor space improvements that they'll get as we model TCO with some of them. So I do think that there's going to be a big push for the higher capacities, just like there always has been in our industry. And I think this is going to be one of the biggest jumps in areal density that we've done in the last 5 years.

So the healthy growth is right ahead of us. We are filling the lines with -- the wafer lines with HAMR parts. And we expect to see volume shipments in the millions of drives next year, or next calendar year. It'll all time out based on when the qualifications are, but the parts are coming.

[Operator]

The next question comes from Sidney Ho of Deutsche Bank.

[Sidney Ho, analyst]

It's great to see some green shoots in the U.S. cloud market. Dave, last quarter, you talked about expecting cloud inventory to normalize in a couple of quarters. Can you give us an update about the timing there? But more importantly, are you getting indications that the rate of recovery is going to -- what kind of rate of recovery is going to look like beyond the December quarter, especially given your efforts to increase visibility through earlier collaboration or longer-term collaboration with your customers?

[William Mosley, executive]

Very good. Thanks. Yes, I think as we are giving predictability on what products that we have in our pipeline, the customers are likewise giving more predictability on what they will need out there in time.

So I would say, if you look at 2 years ago, the demand was quite strong, and then we entered into this period about 6 quarters ago now where some customers were saying, "We really don't need very much and we're going to make allocations of what we've already got in our data centers last for another year." Or something like that. And the entire industry has been suffering through that.

It's very hard to run factories like that. And so that's why we've gone out and said, "Okay, let's get predictable." What we're showing -- the numbers we're showing for customers right now in these long-term forecasts for build-to-order are nothing compared to the volumes of where we were at 2 years ago. They're much smaller. But I think the customers are showing us predictability and then asking us for upside on top of that. So it gets into a really interesting discussion about what's the true demand.

And exactly to your point, I think it will help us get through the back of this period. But then I think, ultimately, we're not satisfying true demand when the -- because the growth of data is still large. 25% CAGR is still very, very large. And so we're trying to make sure that we start the parts that ultimately will get paid for and showing people exactly what we have in process for them. And they're trying desperately to show us a more predictable schedule that we can all manage better for better economics on both sides, of ourselves and our customers.

[Operator]

The next question comes from Karl Ackerman of BNP Paribas.

[Karl Ackerman, analyst]

On HAMR, could you discuss the reliability metrics and perhaps power efficiency metrics relative to your existing SMR drives? I ask because at OCP last week, it's clear that hyperscalers continue to prioritize those metrics as they introduce newer technology.

And I guess as you address that question, could you also discuss maybe the number of customer engagements you have with HAMR perhaps beyond the existing customer?

[William Mosley, executive]

Yes. We're -- as we said before, Karl, we're bringing on multiple customers right now. So we've already shipped qualification units out to multiple customers. I won't talk about them in particular.

The HAMR reliability metrics will be identical to the reliability metrics of traditional drives, so there's really no change there. And there's no change in the power, either. I mean people have pointed to the laser subsystem and things like that. Yes, there's very small changes according to that, but we're also working power down in normal course on these products.

I think one thing important to realize about the family is that the mechanical and electrical design points of the 2.4 terabit per platter drive that we just announced are very similar to the 3 terabyte per drive nodes or even beyond that, we should get into the mid-3 terabytes per disc drives with those same parts. So same electronics, same mechanics, so there's really no change in power, there's no change in reliability expectations either.

[Karl Ackerman, analyst]

Super helpful. If I may sneak another one. Just on your systems business. That has moderated roughly 40% the last 2 quarters. Just curious, I suppose, why and kind of the outlook for that. And within that, how much of that is maybe softness in flash prices versus perhaps the CoreVault hard drive offering?

[William Mosley, executive]

Yes. I don't really think Flash has much to do with it. It's more, what I'll call, on-prem traditional enterprise applications. And thankfully, that space has not been as bad as we thought it was at the back of the year. It's not nearly as cyclical as what the cloud has been in this recent cycle, but it was down year-over-year. And I think it will recover over time as well because I think on-prem enterprise should benefit from all the growth of data. In many cases, you can't move the data off-site or into the cloud. It's either too massive or you've got regulatory requirements or sovereignty requirements and you want to keep multiple copies anyway.

So I do think on-prem enterprise should have a good recovery at some point, and we're happy with the systems business, and our penetration into multiple accounts on that front to recover when the on-prem enterprise business does recover.

[Operator]

The next question comes from Aaron Rakers of Wells Fargo.

[Aaron Rakers, analyst]

I wanted to maybe just unpack the gross margin a little bit, appreciating that you guys don't give a defined kind of guide on a forward basis. Could you help us to understand kind of the impact of underutilization you expect going into the December quarter relative to the $59 million reported this last quarter?

And how do we think about, as maybe that starts to lift out of the model, kind of the glide path to back to that 30% level as HAMR starts ramping, et cetera. I'm just curious how we think about or how you guys are thinking about the gross margin trajectory, kind of the variables within that looking out over the next couple of quarters.

[Gianluca Romano, executive]

Thank you for the question. Yes, our guidance for December quarter is, of course, implying an improvement in gross margin. This is not coming from underutilization. We think underutilization will be fairly flat with the September quarter. But it's coming from, of course, there the pricing actions that we are taking, and we have already started in the prior quarter, and a better cost structure. Part of the improvement is for sure coming from the full impact of the restructuring plan that we started in the prior quarter. So now we start seeing the full impact in our cost structure. So those are the major drivers for the improvement in gross margin.

Then we will continue to improve our structure, and we think our revenue will continue also to increase sequentially. Now to go back to the, let's say, the 30% gross margin, we think we need to have a revenue that is lower than the prior peak, we think at least 20% lower. So we can achieve that level of profitability with a much lower revenue.

[Operator]

The next question comes from Timothy Arcuri of UBS.

[Timothy Arcuri, analyst]

I also wanted to ask on gross margin. Gianluca, as HAMR ramps, I think there's some controversy in terms of whether it's initially negative for gross margin? Or like what the crossover point will be for when it becomes positive to gross margin. I think it depends on your yields, obviously. But can you just talk about sort of how that plays into the answer to your question, the trajectory of margins off the bottom here?

[Gianluca Romano, executive]

Yes. As Dave said before, no, HAMR will strongly increase our capacity per drive, and that will, for sure, improve our gross margin. It will be accretive to our gross margin since the beginning. But of course, when we go to 3, 4, 5 terabytes per drive, you will see even a bigger improvement.

Now we think we will start our HAMR revenue fairly strongly in the first 6 months of the calendar '24. We think we have about a million units as opportunity to be sold. That will help, of course, as every time you go into a new technology and new products, we could have a little bit lower yield, and that could limit in the first quarter or 2, the improvement to the gross margin. But we see gross margin improving sequentially.

[William Mosley, executive]

And Tim, just I would add, remember that it's not just about the highest capacity point, although that does drive a lot of heads and media in our factories. It's also about mid-range, if you will, capacity points, like 20 terabytes or 24 terabytes again. Or those -- the areal density enables us to go address those capacity points with improved cost structure.

[Timothy Arcuri, analyst]

Great, yes. So I also jumped on just to smidge late, so maybe you talked about this. But can you talk about how the change to build-to-order is impacting bookings? So I know that revenue is being guided pretty flat, but it seems like bookings are improving, and I wonder how much of that is due to the shift to build-to-order and how much of that is due to just the customers having worked through inventory? And what do the bookings tell you about the trajectory of where revenue is going to go into the first half of next year?

[William Mosley, executive]

Yes. It's actually an interesting question. So we do have some customers that are embracing the predictability. And there are reasons for that. Maybe some of it is because they have burned through their inventory completely, and so they know that they're going to be buying. I think there are customers who are not leaning into multiple years just yet for various reasons. They make procurement decisions all the time, and so do we, right? Everybody has to make these tough decisions. .

But generally speaking, I think it's giving us better visibility, at least at the lower rate that the industry now runs, because remember, the industry just doesn't have the money to speculatively start a bunch of products right now. We have to make sure that what we do start, the suppliers are going to get paid for and so on.

I think the model is generally working out pretty well. And as we show higher and higher value like the 3-plus terabyte per disk capacity points and then the 4-plus terabyte capacity point, I think people will want to make sure that they can take advantage of that TCO proposition we put out in front of them. So I anticipate it will pick up steam.

We are still doing things that are 2 quarters or 4 quarters or something like that. And so there's negotiations in everything, and that can be frustrated by that. But as those negotiations continue on, I think we will continue to make sure that we right the industry and ourselves, our suppliers, so that we can at least get back to a point where returning value to everyone so we can keep investing in the industry. I think that's an important point. That's one of the reasons we've done this.

[Operator]

The next question comes from Krish Sankar of TD Cowen.

[Sreekrishnan Sankarnarayanan, analyst]

I had a 2-part question. First is you folks shipped about 56 exabytes to nearline customers in the quarter. Given the view that nearline is going to improve in calendar '24, do you think you will hit the 100 exabyte run rate sometime in calendar '24?

And then a follow-up is, Dave and Gianluca, it seems to give most confident about HAMR and gross margin this quarter versus all the prior quarters. Kind of curious, what is the reason for that? Was there any improvement in the quarter that you had some milestones? Or was it more increased customer demand or better visibility into the purchasing of HAMR next year? Any color on that would be helpful.

[William Mosley, executive]

Yes. I would say all of the above, as time marches on, our teams make progress against the yield targets, the reliability targets. And the qualifications progress, and we can see -- maybe we have more and more certainty as we get towards the end of the qualification. So all of those things are factoring into our confidence, and I think we'll continue to update everyone going forward exactly how this is going. .

Remember that we're also starting into qualification with this 2.4 terabyte per platter, which again, I made the point before, it's almost the same box. And so we have a PMR outlet for the same parts, and we're driving the vendors to that commonality, of most of the vendors, the lion's share of the vendors have common parts through these 2 platforms. So we can drive that much more volume in and predictably get people paid and things like that.

[Sreekrishnan Sankarnarayanan, analyst]

And then on the exabytes for nearline. Can you hit 100 exabytes next year? Or is it too aggressive?

[Gianluca Romano, executive]

Yes. I think for the current fiscal year and probably the calendar '24, is not very probable that we can double the exabyte, but we think we will grow sequentially in all those quarters.

[Operator]

The next question comes from Blayne Curtis of Barclays.

[Thomas O'Malley, analyst]

Sorry about the name mixup here. This is Tom O'Malley on for Barclays. I just wanted to understand the timing of the recovery a bit better here. you previously have talked about Q4, and then you talked about at the end of December. And this is a near line in particular. And it's kind of pushed to Q1. You're seeing this U.S. cloud uptick you said, again, so kind of in September and in your expectations for December. But when do you expect the market step up in the market? Is that still expected for March? Or have things kind of elongated just given the inventory situation? Any update on that recovery would be helpful.

[William Mosley, executive]

Yes, Tom, thanks. I think we're still on the same plan that we talked about last quarter. It's a gradual uptick. So we're watching the inventory being depleted. We're seeing new orders come in. I think the one variable would be maybe some of the global cloud customers, not the U.S. cloud customers, but maybe some of the global cloud customers, given some of the economic issues that we have in various parts of the world. But generally speaking, we're still on the same plan, and we'll see a gradual uptick rather than a hockey stick.

[Thomas O'Malley, analyst]

And then just on the gross margin side as well, just taking the midpoint of guidance and you're talking about operating expenses actually up slightly in the December quarter. It implies 200-plus basis points of sequential improvement in gross margin.

I understand that you pointed to some pricing increases, but that will be pretty quick in terms to get the full benefit there. Is there any other levers that are contributing to December? Obviously, you have some mix benefit with VIA gone -- or going down and nearline up. But just any other levers that you could point to, other than the pricing, that are impacting in December.

[Gianluca Romano, executive]

Yes. I will not say that pricing is the only thing driving the gross margin up. Actually, the cost side is very important. We have the full impact of all the restructuring plan that we executed in the prior quarters that are now going into our COGS, of course, also in our OpEx. OpEx could be a little bit higher, but not much higher. So we are talking about a few million dollars higher. So I would say both pricing and cost should go in the right direction to improve our gross margin sequentially.

[William Mosley, executive]

And sorry, just to be clear. We're also going through product transitions, right? So as we may raise price on one of the older products, and then there's a -- the customer can offset some of those increases by a better TCO proposition in the next drive, and we can go work with the cost on that. So the mix plays a role, and the customers can see the TCO benefit, and they're incentivized for that.

[Operator]

The next question comes from Steven Fox of Fox Advisors.

[Steven Fox, analyst]

Dave, understanding everything that you said about sort of a cyclical recovery, there's still a lot of macro headwinds out there. And obviously, your decline in sales over the last 12, 18 months has outdone what the macro is doing.

So I'm just curious, how can we get comfortable with the idea that, as you see more macro pressures, that business, your business keeps recovering? Is there anything you would point to in particular that may not limit cloud spending as much more than you think of? Or just like other cycles where you outgrew in tough environments.

[William Mosley, executive]

Right. Thanks. You're right. It is a tough environment. I will say that data continues to grow, and people want to improve their economics all the time. So the data centers that exist in the world have an enormous number of hard drives. So we're going to see some refresh of those for various reasons, power, some of them are just aging off. The upgrades are -- could be actually fairly large. If you think about buying a 32-terabyte drive and replacing 4 -terabyte drives that may still be in your system. I mean, those are market economical benefits that will ripple through the data centers.

And so I do think that, in some cases, since data is growing so big, it bucks the trend of what's going on in the macro. I mean, obviously, that everyone is paying attention to the macro. But I do think that there will still have to be some investments to make, and there will also be opportunities to go save costs with some of the new products that we have coming. And we've also been through the cycle, on the early side already of already the inventory is being depleted around the world. So there's not that massive inventory bubble out there anymore.

[Steven Fox, analyst]

That's interesting color. And then, Gianluca, can you just -- I mean, I assume you don't want to give like specific underutilization charges likely for the first half of next year. But can you sort of directionally give us a sense for how long you think we should keep modeling even if there are smaller charges in calendar Q1 and Q2 to keep that in our gross margin calculations?

[Gianluca Romano, executive]

Yes, we expect to have underutilization cost also in calendar Q1 and calendar Q2, probably a little bit -- could be a bit lower than what we have in December, of course. But of course, the volume that we are producing is still a little bit below what we had installed at the top of the cycle before.

[Operator]

The next question comes from Ananda Baruah of Loop Capital.

[Ananda Baruah, analyst]

Yes, I guess, Dave, just given that you're starting to see saying start to firm up and talking about sort of, I think, the Q-over-Q increases, in nearline as you go through '24. What's the opportunity for things to get tight as you go through '24 just given the capacity that you've taken offline?

[William Mosley, executive]

Well, I think that tightness is actually kind of interesting because, like we said before, the data is growing in the data centers, we all know that. The amount of data being generated is growing quite quickly. And I think relative to the hard drive industry quickly reacting to anything right now because of some of the damage that's been done in the supply chain and just some of the -- frankly, the lead times that exist on current parts, especially at the highest capacity points, there's not as much flex as there used to be.

And so I think we may see -- we may enter into an environment where people say, "Okay, I see the economics of upgrading part of my fleet here now. Let's go ahead and do it. It will save me power, it will save me space. It will allow me to answer the call for the data that's growing." And then once they get their orders in, we'll say, "Well, the lead time is x, and that might be the challenge."

So that's how I think when things may get tight and it will manifest itself. I don't know exactly when, but I certainly think that we could get into a situation like that just simply because the hard drive industry does not have the immediate capacity gains that it used to.

[Ananda Baruah, analyst]

And then what the -- so what are the downstream impacts to that? I mean, in the past, it's been pricing goes up, long-term agreement, that type of deal. Is that still some of the stuff that could occur if you find...

[William Mosley, executive]

That's right. But that's exactly why we're addressing the customer base with these build-to-order models because we -- I think we're -- in some sense, we're helping the customers get a predictable financial outcome if they can give us at least some predictable visibility.

[Operator]

The next question comes from Mark Miller of the Benchmark Company. .

[Mark Miller, analyst]

You mentioned AI opportunities in VIA cameras and other areas. I'm just wondering, when do you think these opportunities will really start to ramp? And any idea on the magnitude of these opportunities in terms of sales for next year?

[William Mosley, executive]

Yes, Mark, I think it's really hard to quantify just yet, but there's a lot of stuff going on in, I'll say, say big data around AI. You've got people appointing Chief Data Officers now to be able to track where is the data, what's its value, might we want to retain that just in case that we end up with some tools that are allowing us to monetize it or understand more about our process? So on and so forth, with customer base, factories.

So my personal opinion is we're in the early innings of a move from throwing data away to keeping some of it longer term for the benefits of the corporations. And I do think that a lot of that will be edge. I don't think all of it will end up in the cloud. I think a lot of it will actually be on the edge.

And so this is something that we're tracking very carefully. I don't think it's really manifests itself just yet. I think people are using some of the new application capabilities that are being branded AI applications to get to know them and understand them. And I -- certainly, like things like generative AI, which I believe is kind of a new user interface, if you will, that will allow the application to be used much more efficiently and maybe queries to be made of these applications much more efficiently.

But I think we're still in the early innings. And I think once it does latch, we're going to know that data and the longevity of that data and the integrity of that data is all critical, and so I think that's good for us.

[Mark Miller, analyst]

You mentioned lower CapEx for fiscal '24. Can you give us a range or any idea would it would be?

[William Mosley, executive]

I think we'll probably stay inside of our existing range, 4% to 6% of revenue. But we are -- since the revenue is so far down, we are very mindful of the spending. However, we can see the tools that we need to bring HAMR up according to a certain pace. And as soon as -- if the pace quickens, we will get there as quickly as we possibly can. So we understand what the recipe is, and we understand what's needed to make the recipe really well and we'll spend accordingly. .

[Operator]

The last question will come from Vijay Rakesh of Mizuho.

[Vijay Rakesh, analyst]

Just quickly on the HAMR side. I was just wondering, when you look at exiting calendar '24, what your mix of HAMR would be either by exabyte or units?

[William Mosley, executive]

Calendar '24, you said?

[Vijay Rakesh, analyst]

Yes.

[Gianluca Romano, executive]

Yes, I don't think we guide so far in time precisely on the mix.

[William Mosley, executive]

Yes. We're not going to guide that. But I will say that we'll be very aggressive. When we talked about rates per disc and cost optimization of these platforms and things like that. I mean this is something that a hard drive industry has been doing for years and years and years, decades, right?

So we know how to do these transitions. We we're very confident in the technology, and we'll look to be very aggressive there. The wafer lead times are also quite long. So we're already starting on this journey because we're already in wafer. And so we're populating the wafer fabs with the parts that will support it.

[Vijay Rakesh, analyst]

Got it. And then on the nearline side, obviously, the shipments easements have come down quite a bit. As you look at the green shoots with some data center coming back, do you feel pretty comfortable, given what the inventory levels are at the OEMs and what you see in terms of a return on the spend? How would you characterize that if you look at those 2?

[William Mosley, executive]

Yes. I do think that the inventory has been depleted now to levels that if you think about the complexity of all the data centers of the world and how much material needs to be parked out of them in front of them for replacements and then what the data growth is in the data center. I think the inventory levels have come down to a point where we feel comfortable now that people are going to get back to more predictable buying patterns. .

[Operator]

This concludes our question-and-answer session. I would like to turn the conference back over to management for any closing remarks.

[William Mosley, executive]

Thanks, Andrea. As you heard today, Seagate remains focused on our key priorities, including executing our leading technology road map, which we believe positions us well to enhance profitability over the near term and to capture long-term opportunities for mass capacity storage.

I'll close by once again thanking all of our shareholders for their ongoing support of Seagate. Thanks for joining us today, and we look forward to speaking with you during the quarter.

[Toshiya Hari, analyst]

The conference has now concluded. Thank you for attending today's presentation, and you may now disconnect.