That’s the — but I think it will be challenged because I think that’s what a lot of enterprises are kind of assuming to make their models look okay right now. And that’s why the factory workers are getting strained. But we’re going to have to navigate our way through this and be as communicative as we can. I don’t think the industry capacity is what it was a year ago. And I think if you don’t tell us now, then it may actually be even more stressful for you a year from now.Operator The next question comes from Aaron Rakers of Wells Fargo.Dave Mosley Aaron, we can’t hear you.Aaron Rakers Yes. Sorry about that, guys. Can you hear me?Dave Mosley Yes, yes.Aaron Rakers Yes. Sorry about that. I guess I got two real quick. Dave, just on the heels of that last question, I’m just curious of if you’re asked kind of the longer-term growth of nearline capacity shift trends, where do you think that is today?
Like as we come out of this, do you think we — maybe I will stop there. What do you think that long-term capacity shift growth trend looks like now?Dave Mosley Yes. It’s an interesting question because I think if we go back a couple of years, you remember there was a stall and then there was an enormous growth pop back. I think it’s not going to happen like that again. I mean, from an exabyte perspective, if by that time, we’re on mid-3 terabytes per disc or even potentially 4 terabytes per disc that our exabyte output is going to be significantly higher at better economics for us as well. But will we have enough capacity for everyone all at the time? I think the answer to that is probably no.So — and I don’t think we’ll see the magnitude of the swing that we did before.
Maybe we can’t from theoretical demand, but I don’t think there’s going to be supply right now because of the stresses that are going on in the industry. And the fact that people just can’t start materials, they won’t have a whole lot of inventory, they won’t have a whole lot of materials, they won’t be leaning forward into some of those really great value propositions nearly as hard because they just can’t do that speculatively on long lead times. So it’s actually an interesting problem set versus what it was a few years ago.Aaron Rakers Yes. And then the follow-up question, just trying to think about gross margin kind of normalizing itself or, I guess, taken another way, if we kind of think about the progression of lifting out the underutilization charge, appreciate you’re not guiding beyond this next quarter.
But as I look back in time, when I look back and say, hey, if this model gets back to total capacity shipment levels of 150 exabytes a quarter, 130 exabytes a quarter. How do we think about that relative to lifting out the underutilization charges in the P&L?Gianluca Romano Well, of course, it depends on what will be the capacity installed at that point. As you know, we are reducing capacity right now. So gross margin, what we see today for the next several quarters is despite the short-term decline in revenue, we guided in a way that is implying an improvement in gross margin already in the June quarter. So I think that will continue for the next several quarters. And of course, as Dave was saying before, demand weakened it and capacity will be at a certain level, and hopefully, supply and demand will be well balanced.
And we have to manage this short-term demand decline. But again, we are ready to do it. We have done that in the past, and we think we know how to do it.Aaron Rakers So underutilization charges continue beyond the June quarter through the back half of the calendar year?Gianluca Romano We will have some, I think, also in September. But it is declining sequentially.Dave Mosley And some of that is a reset to fixed and variable cost as well, Aaron. So I think as we look forward, we have to project what we think the supply that we’ll need to meet that demand is, of course, and we’ll be adjusting and trimming as need be.Operator The next question comes from C.J. Muse of Evercore ISI.C.J. Muse Yes. I was hoping to maybe drill down a little bit deeper on the nearline outlook.
If you think about your vision three months ago versus today, obviously a change. And I guess I was hoping you could kind of isolate between cloud, enterprise and China and maybe rank order the change statements there in terms of the slowdown as well as from an end demand overall perspective versus just realizing that there’s just too much inventory downstream, and that’s what kind of caught you by surprise. So I would love to hear kind of maybe the moving parts there. And if you could prioritize what has really driven the change, that would be helpful.Dave Mosley Good. Thanks, C.J. I think it is good to break it down in a few different pieces. The most relevant, obviously, is the major cloud service providers around the world, and there are a few different behaviors depending on geography and so on.
If I think about the big applications in the cloud, they are growing a lot. I think right now, there are priorities being made on compute and their priorities being made on other investments. And then there’s people questioning business models and all of that just takes time to sort out, but the data keeps growing. So I think that’s largely what we’re seeing.We would have forecast six months ago that we would be coming out of it by now, just based on all these trends and run rates, and I think people can always hold on for another three months or six months. They can’t hold on much longer than that. So that’s the way I think about the major cloud service providers. The difference — but Mike, the subtle difference might be in China, where you’ve got cloud providers there that really have been kind of on a pause for quite a while and are starting to get a little bit more optimistic, again, like I said — I alluded to earlier, not necessarily purchase orders just yet, but get a little bit more optimistic about their investments.
And so that’s a watch item for when that might come back.On the enterprise side, I know there’s a lot of discussion about servers and other componentry. From our perspective, data does continue to grow on-prem. There are stressed budgets in the CIO’s world today. But I think some of the offerings are actually pretty efficient right now. So I do think for totally different reasons — well, unless you consider macro being the underlying driver of everything, I think the on-prem enterprise is going to take six months probably as well.I do think that there’s not too much inventory there. So we’ve got anecdotal evidence that, that inventory has been well managed. And so if there’s an opportunity to break north into some of those investments, we could go necessarily quicker.C.J. Muse Very helpful.
And I guess as my follow-up to Gianluca, thinking about liquidity. You talked about intention to pay down the debt that’s coming due in June. Curious how you’re thinking about gross cash? And is there a plan today to draw on the revolver? Or that’s TBD depending on kind of the free cash flow generation into June and September?Gianluca Romano No. I think our cash balance will be fairly stable at the end of the quarter compared to the end of March. So I don’t think we need to use the — we don’t need to use the revolver. Now sometimes, we use a revolver during the quarter to cover for a few weeks. But in terms of cash balance, I think, it will be fairly similar to what we had at the end of March.Operator The next question comes from Krish Sankar of TD Cowen.Krish Sankar My first one is for Gianluca.
Just wanted to follow up on HAMR. How much of a drag is it on margins today? And when do you think it will improve towards the corporate average? Is it just purely a function of volume in eve that helps you improve HAMR margins? And then is the long-term gross margin target still 30% to 33% baking in HAMR? Then I have a follow-up for Dave.Gianluca Romano Yes. Well, I would say right now, we are very happy with how we are progressing with HAMR. As we said in the prepared remarks, we are shipping products for Qual, so right now, we don’t have revenue coming from HAMR. So it’s not really impacting our gross margin. But we think, of course, in the future, 3 terabyte per disk or 3.5 terabyte per disk and 4 terabytes per disk, those are very interesting propositions for our customers, and we think will be actually a very profitable gross margin improvement for Seagate, which is why this product is so important to us.
And in general, in the longer term for the entire industry.Krish Sankar Got it. Got it. Very helpful. And then a follow-up for Dave. Sorry, just to come back to this SSD versus HDD dynamic. I agree with you that the NAND prices are unsustainably low. But if you look at like the last down cycle, the SSD prices are almost 8x higher than HDD on a per gigabyte basis. Now it’s more like 2x to 4x. So that gap is definitely closing. So I’m kind of curious, when you talk to your cloud customers, is it a bigger — how does that discussion progress? And is that a certain part of your weak pricing per terabyte? Or is it part of your weak demand from cloud customers? Or I’m just kind of curious about it.Dave Mosley No, I would say when you get into mass capacity, the highest capacity points, I don’t subscribe to the notion it’s 2x to 4x.
I think it’s significantly higher than that. And we’ll have an ability to continue to drive to better and better value propositions as we add more capacity per drive through the transitions that we’re talking about. So what I would say is that right now, demand is low for everyone. And so what you see, the dynamics you see is just manufacturers trying to take some of their cash that they’ve tied up in inventory and turn it back into cash as quickly as possible. And if you start to look and say that’s below cost or something like that, hopefully, people stop doing that and try to look for other options. And that’s exactly how we’re thinking about ourselves as well, right, to make sure you don’t build.So I think in that kind of environment, it’s really hard to extrapolate on trends, exactly to your point, Krish.
I think some of these other things that we’re comparing against are unsustainable. They’re going to have to re-equilibrate at some point. I don’t know exactly when that is. The more we push into the value chains, the more we push inventory in there, the more people will build buffers and then continue that. So I think that’s one of the reasons why from our not hole in defense, we’ve continued — we’ve decided to really pull back on our builds and reduce our footprint.Operator The next question comes from Toshiya Hari of Goldman Sachs.Toshiya Hari I had one for Dave and then one quick one for Gianluca as well. So Dave, I know it’s a bit of a black box, but I was hoping you could give kind of your assessment of nearline inventory — or customer inventory in the nearline market relative to six months ago, 12 months ago, what’s your view there?
And when you compare and contrast what you’re selling into the nearline space versus what’s truly being consumed, what’s the percentage delta at the moment?Dave Mosley Yes. It’s actually an interesting thing because if I normally talk about inventory like distribution channel inventory, for example, I’ll look at weeks of supply or something like that. But then that’s — is that based on the last four weeks or 13 weeks? I mean, if we base things on a couple of years ago, the demand is low and there’s not too much inventory out there based on the run rates of number of drives that were a few years ago.Even on exabytes, the demand is fairly low — or sorry, the inventory is fairly low. But I will say that most cloud service providers need to have some inventory around based on the data center populations they’re — the new data center buildouts they’re doing and the refresh the spares that they need and things like that.
So they normally need to have a few months of inventory anyway.It’s more than a few months right now, but it’s not significantly more. And the gross number of drives is not huge, I think. So that’s the way I think about it. The exabyte growth will be substantial coming some time and what we’ve got to do is just make sure we’re not pushing any more into those chains. But I don’t know if a good way to think about it is really three months or six months like we might want to say because if you go back two years historically on the number of drives that were being pulled by people, that’s — it’s not really six months. It’s probably more like 3.Toshiya Hari Got it. And then any view on sell-in versus end consumption? Or is that too hard to really see.Dave Mosley Yes, here’s what I’d say is that we see a lot of the new drives, the higher capacity drives going into replacing lower capacity drives, obviously, but the lower capacity drives get put into other purpose applications as well.
So I think drives are living longer, but there’s still going to be demand for higher capacity drives, especially just the efficiency that you get out of that in the data centers. And then there’s new data center buildout as well. So I actually think that this is not a problem of not needing any product. This is a problem of reshuffling priorities inside of the data centers and getting ready for the next level of data center expansion.Toshiya Hari That makes sense. And then Gianluca, on gross margins, longer term, obviously, you’ve had this target of 30% to 33%, if I’m not mistaken. You’re obviously significantly below that today, but as you think about the path towards your long-term model over the next four-eight quarters, given the reduction to manufacturing footprint that you’re executing to, what sort of quarterly revenue run rate would you need to be at to be at that, call it, 30% range?
