The distribution is no longer declining from the prior quarters. The inventory at the direct customers is a little bit harder to pin down, but overall combined, we are starting to see that this long-drawn inventory correction is finally sort of coming to completion. We have seen two meaningfully good trends. One is total bookings. I showed a graph, which you can really see on the website, that our January and February bookings were meaningfully higher than any other month in the prior year. The second trend was a total backlog on the company. Total backlog, which, you know, three years ago was, you know, multi-year backlog, just, you know, at the peak of the cycle.
But this backlog has been coming down now for a couple of years, and the backlog on the curve shows it bottomed out and now has a slight slope up, but the backlog is starting to increase. No hockey stick increase, but I think in the coming weeks and months we'll see that backlog continue to strengthen. And with the combination of those two, we think, we think we're coming to nearly the end of this correction. So the rest of my presentation was really on the nine-point recovery plan, and I'll pass it back to Joe in any order he wants to cover it.
Sure. Well, maybe starting on the business update, you know, this time last year, Ganesh was here and talked about bookings improving and the stock kind of got really strong, and that ended up being somewhat of a false start. You know, what is it that's different when you sort of look at the distribution trends and end market trends that you're seeing? What gives you the confidence in stabilization that you're talking about now?
So, so I'm looking at total distribution inventory, and I'm looking at, you know, what distribution is selling out to their customers and what distribution is buying from us. Now, what they're selling to their customers is still higher than what they're buying from us. But their inventory is not too idle where it should be. It's still higher than that, but it's becoming increasingly out of whack. So stocks are showing up, and they're starting to buy product to make up for what they need to have to serve their customers. So when you look at the overall distribution buying patterns, their purchasing is no longer decreasing from prior quarter. And I think that's a significant change.
Yeah. The thing that you can see on that bookings graph that we included this morning is, it shows last year and the data that we were seeing. You know, February was higher than January, March was higher, and April was higher, and then it just flatlined from there for about nine months. Now we've seen a meaningfully higher increase in bookings the last two months.
Great, and if I could just get one quick question out of the way, 'cause it's coming up a lot. Do you, is any of this, you think, related to tariffs, people building stuff to get ahead of potential tariffs that kick in?
We don't think so. We have met with all the distributors. We have met with many, many customers. And when the tariff question comes, people kind of just all, everybody thinks that it is above their pay grade. And most people think that tariffs are just a threat and in reality may not happen. And if they do happen, then everybody would be in the same boat and we'll deal with it, you know, like we dealt when tariffs were put on China a few years ago. In that case, we largely moved our assembly business out of China. Before that, we had about 8%-10% of our product that was assembled in China, which was the definition of made in China. And we moved it to Taiwan, Thailand, Philippines, Malaysia, Indonesia, wherever.
And today, only about 3%-4% of our business is assembled in China, and most of that goes to outside of the U.S., goes to Europe, goes to Asia. So therefore, when the tariffs are increased on China, it will almost really have no impact on us. Now, if you, today, as we speak, there are no tariffs on Taiwan, Malaysia, Indonesia, Philippines, Thailand. If tariffs are put on everybody, then everybody's in the same boat, and there's no place to hide, and, we'll be in the same boat as our competitors would be and our customers would be, and we'll deal with it.
Great. That's helpful. And then before we go through the nine-point plan, you know, there's a through line of questions. I know I've talked to a few investors this morning after your session, and it all kind of comes back to where's the actual revenue level once you get back to shipping at demand. So you were doing over $8 billion in revenue at the peak. You're, you know, close to $4 billion, you know, a little more than $4 billion run right now. You know, obviously the consumption is somewhere in between those two things, but all of this planning of manufacturing, manufacturing footprint, OpEx profile, long-term growth rate, like what's, you know, and I know you don't know exactly, but just goes through your thought process of how do you approach all those decisions, not knowing exactly where the starting point is.
Well, I mean, you know, if your question is what is the baseline of the revenue, I don't know it either. When you know it, let me know. But I think we have, you know, we have done significant analysis in peeling through where the excesses were and how the distribution inventory got built, how the direct inventory got built, and the, you know, that baseline of the revenue is well, well below the peak revenue.
Mm-hmm.
But we don't know exactly where that number is. We made some conservative assumptions in our analysis in which we talked about how the inventory goes down, the point where we're sizing the factories down too, and we're leaving the room for, you know, upsides where we can rapidly respond because the inventory is high. And in our, you know, most advanced fab, which is the Oregon fab, we decided not to solve the problem purely through sizing down through layoffs. We're doing some layoffs, but we're doing a smaller layoff and leaving a two-week rotating time off in place in Fab 4. The reason being the rotating time off can be taken away next week.
Mm-hmm.
And the capacity immediately goes up. So, if we had gone too conservative anywhere, we're leaving room to rapidly bounce back up through that rotating time off. And then we have substantial equipment, which is currently underutilized, and we have substantial more equipment than we bought at the peak of the cycle. Some of it is not even hooked up, that without buying any more CapEx, we should have very low CapEx for the next several years and still have the ability to substantially, you know, increase the output from our factories.
Great. So maybe we could walk through the nine-point plan, and I really, you know, admire what you did here. You came in a very short time ago, made a very candid assessment of some of the challenges that, you know, the company was undergoing, and then you've made a lot of progress in a short period of time, and you gave us the update this morning. So maybe if we just kind of go in the sequence of the nine points, if we start with manufacturing, the consolidation of the manufacturing footprint, can you give us an update there? And again, what's the ultimate goal of that?
So I think on manufacturing, the assumption the company was running against is that, you know, keep building inventory. It'll be eventually sold. We'll have a major snapback, and you're gonna need all this product. And I think I look at that as a, you know, hope is a strategy and definitely a strategy. We're doing, you know, tangible things to improve business, design wins and revenue. And if we get some help from snapback, it'll be in addition to what we're trying to grow business without that. So I find the size of the factory was way too large for the size of our business. Now, remember that we were building inventory at the peak of the cycle. So the quarter in which we did $2.284 billion, we built inventory even on that, even at that revenue, which is usually crazy.
You know, at the peak of the cycle, it should be running on fumes.
Mm-hmm.
And then the inventory is very low. And when the low cycle comes, that's when the inventory builds. And then you start the upcycle with a good amount of inventory. I mean, that would be normal. But if you're building inventory at, you know, at the peak of the cycle, when the revenue is 2x higher than today, then we had just way too much capacity. So we have, you know, if you'd done it a couple of years ago, there were some better choices available, but the choices I had was too much inventory. I couldn't build any more. We had to sacrifice one of our Fab. And carefully looking at it, that became the Fab 2, you know, because 70% of the Fab 2 products have run into other fabs and we could rapidly transfer.
And as we started to bring those technologies in the other fabs, in some of the very first runs, we have had near perfect yield. So, you know, that proves because we have run them before. So we pulled the schedule in. Originally, we told you we'll close it in September, and we pulled it into May. That will even save money earlier, faster. And the other two fabs, Fab 4 and Fab 5, we're correcting them through a layoff. Fab 5, we're correcting a little deeper because the inventory from Fab 4 was even higher. Fab 4, as I said, we're correcting a little bit less and leaving a two-week rotating time off for a quick snapback if it happens.
And maybe if you could talk a little bit to the idea of headcount reduction, you know, in your book, you kind of talked about the importance of sharing pain. In 2001, 2009, you were able to mostly avoid that. Maybe it's a testament to, you know, how challenging the situation is. But can you just talk a little bit to that and, you know, how does that impact the corporate culture? How do you think about that?
So in our corporate culture, you know, we make great attempts to avoid layoffs and manage our business in a way that the need for layoffs do not arise. In 2008- 2009 global financial crisis, you know, our business was down by a third, but down about 35%, and everybody was closing their factories and shutting down and doing layoffs. When we met as a team, we said, how can we do differently, and that's when we designed this RTO program where, you know, in Fab, from every shift, two-thirds of the people are working and one-third are not for a week. And they have the week off unpaid, and they get the government benefits for that week. And then the next week, those people return and a different one-third are off, so we took one-third of the capacity out and related costs out, and everybody shared the pain.
In the non-factory situations in the offices, where the work is not always dependent on the amount of revenue you have, rather than you know cutting R&D projects and dropping our customers from service and all that and R&D and all that, in all office areas, we did a 10% pay cut around the world. In general, their average bonuses are another 10%, and bonuses went to zero, so 10 plus 10 became 20. And then the rest of it is discretionary travel and computers and you know all the other stuff, so we took our OpEx down 24% in 2008- 2009, without laying off a single employee. When the business came back rapidly, there was a snapback in 2010. Our factory folks simply went from every third week off to every fourth week off to every sixth week off to working full-time to working overtime.
And we had a 57% revenue increase in the following year, was the best performance in the industry, where our competitors had closed the factories and people had moved away, and it was very difficult to reassemble the workforce and took them a year. So, you know, those are some of the lessons we learned in the past. But this time around, the cycle was so deep that we tried RTO. We tried pay cut for people who were on pay cut for 10 or 11 months, and our revenue was still dropping. So when I came, you know, back as CEO, I assessed that entire situation. And you cannot put those people through another pain for another year.
At some point in time, you know, the people have to work full-time, they have to get paid, they have to get a merit increase, and certain people need to go away. So at this point in time, that was the only option remaining.
Great. And then just one last manufacturing question on the business mix of, you know, 40% internal fab, 60% foundry. You said on the call you intend to maintain that. Is there any thought, I mean, you have this kind of fixed cost to factories that are put into place in sourcing some of the foundry wafers?
We don't own the technology to bring the other wafers in.
Got it. Okay.
Wherever we own the technologies. So in foundries, the majority of our foundry business is designed on their technologies. So, you know, on their proprietary technology, we can build the wafers there, but we don't have any IP rights to bring that inside. On a very small area, small number of technologies, the technology that runs in foundry is our technology, where we developed it, transferred it to them during good times for a higher capacity, and we have the right to bring it in. And most of that has been brought in. The ones that hasn't, we're working on it. But that's gonna make a very small difference because the majority of it is foundry's technology.
Right. A point to make there is that anything that's on a process technology that's 90 nm or below, we don't have capabilities of doing that in-house. Anything 12-inch, we have to outsource.
Okay.
Correct.
Great. Thank you. So the second point of the nine, inventory management, you know, the inventory's got very high. You're taking aggressive action. Maybe you could just tell us what's going on there.
So our inventory is 266 days as of last quarter. And, you know, our normal inventory would be 130-150 days. You know, so clearly, I mean, that's a very high inventory. We're not proud of it. But the inventory was building and building and building, you know, under the premise that we don't need to take any actions in the factory and just the snapback will solve all problems. So inventory got out of control, and once it gets out of control, then you have to take really tangible actions to bring it down, which is what we're taking by closing one fab down and significantly downsizing the other two fabs, also downsizing our backend facilities. So our current footprint, with the number of people we're keeping, will be a factory size that is below our demand rate, which will take the inventory down.
And we'll be watching it, and as the inventory bleeds down, you know, at the right point in time, we'll start, you know, growing the capacity again by hiring people. We have all the equipment needed.
Great. Okay. And then the third initiative is the growth drivers. Maybe you could talk about this a little. I mean, there's so many businesses that Microchip acquired that had really rich growth trajectories that are sort of, you know, I think of as separate from the three segments that you guys give. And I think you've elected to kind of not talk about a lot of those things over the last few years. But can you just, as you kind of reevaluate the growth portfolio of businesses you've acquired, you know, what are you most excited about, and how are you reorienting the investment around those things?
So, you know, in our growth initiatives, we talked about two this morning. One is our Total System Solutions, and other is our focus on Megatrends. In the Total System Solutions, we talked about Anchor Products, and then the other catalog products attached around our Anchor Products. In the Anchor Products, we showed a bar chart of three different sets of Anchor Products: our MPU products, our 16-bit microcontrollers, and our networking products. And a lot of those products are many years old, and therefore the attached products around them are saturated and going up slightly, but not much. And they're on the average of four to five products attached around our Anchor Products from zero years ago. So we've been making progress in about four to five products attached.
And then we introduced, before we introduced the new Anchor products, there are two other Anchor product sets. One are our data center products, and the second are our FPGA products. Those are newer products, acquired a few years ago. And when we acquired them, there was no Microchip attached around them. And as we have been attaching products around them, they're all on a very steep rise, where year over year, the number of products that are attaching to data center products and FPGA products is rapidly rising, and there is a lot more to go. And then we introduced two new Anchor products. One was a, you know, an Ethernet 10BASE-T1 product, which is rapidly becoming the standard of connectivity in automobiles, homes, industrial, defense, and other places. And we have a leadership product in that area.
Many, many automobiles, I think, you know, we are the lead design in with that product. I quoted the sources in my presentation. It's projected to be about a $4 billion TAM in 2030 from very small numbers today. So we think that will be a very sizable growth driver for Microchip, and we'll have a large amount of attached around it. The second Anchor product I talked about was PIC64, which we introduced last year. You know, this is about a $5 billion TAM, and I quoted the sources, and it's growing at about 10% compounded annual growth rate. It will have a very high amount of attached also around it as a new product. So that's one, which is TSS, initiative. The second initiative we have been on for a few years is the focus on Megatrends of the industry.
We have had six Megatrends historically. They were IoT, 5G, data centers, sustainability, electric vehicles, and ADAS, which is advanced driver assistance. After comprehensive review in the last 100 days, I have made two changes to those Megatrends. First, we removed 5G because it's kind of mature, and we replaced it by artificial intelligence and machine learning. The second one, ADAS, which is advanced driver assistance, we changed that into connectivity, which is a much broader trend because ADAS was only in the automobiles. The Ethernet connectivity is not only in automobiles, it's also in homes, it's also in industrial, it's also in defense. ADAS is now part of that trend, but the connectivity trend is a broader trend. Those are the two changes we made.
We shared some data with the investors where, since fiscal year 2021, our Megatrend revenue has grown at more than 2X the rate of our overall business. Our overall business grew at 11.95% compounded annual growth rate. Our Megatrend grew at 25.78% growth rate, and our non-Megatrend only grew at 3.5% growth rate. So I think with Megatrends growing significantly faster, that is a, you know, significant growth driver for Microchip.
Thank you. And how do you assess the investment levels around those businesses? Coming back to the company after some time away, you know, these growth initiatives, FPGAs is an area where you've grown your presence. Are you investing the correct amount into those businesses?
We have 22 product lines in the company or had 22 product lines in the company. I don't even know whether I can name all of them, but 8-bit, 16-bit, 32-bit microcontrollers, analog power group, mixed signal group, data center group, communication business unit, FPGA, silicon carbide group. I'm only up to nine.
USB and Ethernet.
USB, Ethernet, you know, on and on and on. I have a list of.
Memory, timing, security.
Memory, timing, licenses, security, computing group. So there are all these various product lines which make up our total business. And I wanted to look at, you know, every single business unit and take a deep dive on what the competitive factors are, which ones are working, which ones are not working, which still are synergistic with company's overall, you know, growth Megatrends and which are not. And after reviewing all that, we made a few changes. We made a few organizational changes, and we combined three other business units into other business units for higher efficiency, more synergy, and things like that.
One of the changes we made was we combined an 8 and 32-bit microcontroller into a single business unit and moved the development tool, which was in a separate business unit, into that business unit so that we can work on the most advanced AI tools and all that, and serve the customers who want to go from 8-bit business unit to 32-bit business units a little bit better. Before that, there were two segment groups, two separate groups, and, you know, they were cooperating from a pricing standpoint, so there's no competition at the price of conversion, but it wasn't as smooth as it can be when they're together. The other one was, our 16-bit dsPIC business unit worked, usually in the same board where our analog power was providing digital power to data centers and motor control applications and other.
And now we put them under the same leader so they can have, you know, much higher efficiency of operation. We combined our silicon carbide group with a, you know, power module group, you know, because a lot of the silicon carbide products were being sold as modules, and there was a higher synergy doing that, and it gave it an economy of scale. We took our power of our Ethernet group, which was a very small $30 million group, but it had a lot of resources, and it wasn't making money. So we'll keep selling the products and harness the gross margin, but we took all the resources out of that and combined it into a Megatrend, which is the Ethernet connectivity group, which is gonna be a $4 billion TAM in 2030 with a substantial opportunity.
We combine all the resources because all these people have Ethernet expertise. You know, there are a few other smaller ones, but, you know, these are the changes we made so that our R&D resources are most organized to achieve the results that we want to achieve.
And then I guess for the next couple of points, you know, the same question with regards to distribution strategy and customers. You know, how have you found the situation with both distributors and customers, and what are you doing to improve that situation?
So on distribution, you know, unlike some of our competitors that have gone in different directions, we continue to count distribution as a very major portion of our go-to-market strategy. Today it's still 45% of our business. A number of years ago, it was 50%, and it has kind of slowly come down, but 45% of our business is still very substantial. And we don't really work from a single distributor like some of the others do. And we find that, you know, every distributor has strengths and weaknesses. No distributor is strong in every geography. No distributor is strong in every end market. And therefore, when you have narrow distribution, you're working with a lot of weaknesses. By having a broad distribution through global distributors, regional distributors, and catalog houses, you know, we're able to, you know, get the best distribution in each geography.
Over time, we have done very well. So, you know, so as a result, we have historically had, you know, much higher incentives to distributors than our competitors did. And as the competitors came down in incentives in a few different places, you know, there was an air gap between what we were offering to the distributors and where the right incentive could be. And we shaved them in two or three areas. You know, one was when we gave distributors a design-in socket registration, then distributors can make the higher margin on that socket for the life of the socket. And sometimes, you know, at Microchip, it can be, you know, a decade plus. And we felt that that was way too generous. Nobody else does that. So we decided to truncate that benefit after a few years, and then it remained proprietary.
After that, the flag became fulfillment, and the distribution margin came down. Because what we want to have distribution do is not sit on that juicy margin, but instead show our new products with better prices, with other features, and, you know, get the customer to adopt our new products. You know, otherwise, you expose that socket and that customer to the competitor. And in the other case, I think we lowered the fulfillment margin a little bit. Our fulfillment margin, again, was on the way high end of competitors, and we brought it down where it is still on the higher end, but there's not an air gap anymore. And I think these are, you know, smaller changes. I have met with all the major U.S. distributors, you know, Arrow, Avnet, and Future, their presidents. You know, it's fine with them. The changes are small.
They still see it as a very good program compared to what they're getting from our competitors, and tonight, I'm heading to Asia, and I'm getting all of our Asian distributors together in one geography. We're meeting all in Thailand. For the next two days, I'll meet with every one of them and, you know, talk about the programs and all that, and for what we have done for Asia, those are very good margins because Asian margins were always lower, so that's what we have done with distribution, and with direct customers, I think, you know, there is no competitor that dealt with their customers perfect during the supercycle. There was no way to do it. I mean, just the product wasn't available. Demand was too high.
So everybody made choices, and the customers that served well and the customers that didn't serve well, and there were price increases, due to inflation pressures. Everybody increased prices. So every customer we have met and said, "Well, you know, everybody did things to them that were not desirable." So it becomes a question of degree, and it becomes a question of, you know, who are they upset with and who are they not upset with. So what we found after analyzing and getting feedback on about 7,000 customers, which is a very large sample size, at 24% of the customers, our customer relationships actually improved. So we served them better than our competitors did. At 64% of the customers, the relationships were the same. And at 12% of the customers, our relationship declined somewhat.
And out of the 12% of the customers, half of them, where we were still preferred customers or we were still approved and customers were buying from them, and we were getting new designs, we were on the approved list. And that left the remaining 6% of the customers where we were not preferred customers where the, you know, there was a challenge was a little larger. So we, after looking at it, we believe that that was a very small number. And I'm sure there's nobody who has higher than 94% customer satisfaction rating. That would be a very high rating. But in any case, we collected those 6,000 customers and made a big program that over, you know, 100-120 days, we're going to have a senior-level meeting with all these customers. And I quoted data on how many are those customers we have already visited.
The output from those meetings have been very well. I think we're improving customer relationships. Many times, customers just wanted somebody to say, "We're sorry," and, you know, listen to their pain and what they went through and talk about how it happened, why it happened, what steps are being taken so next cycle it would be better, you know. But otherwise, you know, we build proprietary products. These customers have our development tools. They have years of legacy designing with us. So it's not like we're being, you know, dropped, locked, stuck, and barreled, but I think the result at the end of the day, when we get done through it, we would have actually improved customer relationships even in those places where they were already very good.
Great. I do wanna ask about the financial model. Maybe starting with growth. You talked about growing, industry growth plus. I, I assume that's kind of a consumption comment. So you're not saying revenue from today. You're saying when we sort of get back to consumption, that's the level of growth.
Yeah. Long term, yes.
Okay. And then with regards to the margin structure, 65% gross, 40% operating, you know, very healthy targets. You know, I guess how do you make those targets without committing to a revenue level and, you know, what adjustments might you have to make to get to that depending on where revenue ends up?
So we didn't put a revenue level at which we, you know, achieved those targets. But what we showed is a path that it's not just, you know, we pulled a number out of the air. There's some thought that has gone in. So, you know, let me reconstruct the 65% gross margin. Last quarter, our gross margin was 55.4%. And last quarter, we took an $82 million inventory write-off. You know, when we have very high inventory, you build inventory and write much of it off, right? Because you're building excess inventory. So we wrote off $82 million of inventory, and we took a $43 million underutilization charge in our factories. So that adds up to $125 million.
If longer term, I capture even $100 million out of the $125 million at the current revenue run rate, you know, that's 1,000 basis points of excess gross margin, which gets us to 65. So it's not really pie in the sky. I think, you know, as we, as we lower the inventory, very rapidly, we'll no longer be writing off inventory. Because once the inventory starts going down, then inventory isn't popping up at the level where it's higher than, you know, 18 months of demand, and you're writing it off. So once that goes away, the gross margin rapidly rises. And the other part rises with the utilization. As the inventory bleeds out, you start reramping the factories. When you reramp the factories, then the underutilization charge, 43, starts going down.
And you only have to capture, you know, you don't even have to capture all of it to get to 65. And in that 65, we've also built a, you know, possibility of China, possibility of competitive prices, you know, readjusting some of the old peak performance prices going away. You know, all that got us to that kind of number.
Okay. And then similarly on the OpEx, to get to 25%, the decision-making there?
So at the peak of the cycle, our OpEx was just a little over 20%, I think. And that wasn't sustainable either. It was you were dividing by a very large revenue that was growing from quarter over quarter and not able to hire expenses, you know, on that additional revenue. And our long-term model was 22.5%. So we changed that to 25%. We believe, you know, we need to make investments, especially in R&D, you know, across these broad different product lines, to grow the revenue, you know, higher than the revenue of the industry. And if you constrain the OpEx to 22.5%, I believe it will not be the right decision for our investors. I think our investors will benefit from 25% OpEx if it can deliver long-term higher revenue growth at higher gross margin.
40% operating profit would still be in the echelon group of this industry. So we think that's the right decision.
So just to wrap up, I mean, you've come back. You were away, you know, from the exec chair for the CEO role, for the last couple of years. You came back. You feel like the growth drivers are intact, the margin's achievable, you know, and after 100 days, it's obviously still some work to get there, but you still feel pretty confident in the direction of the company?
Yes. Very much. And looking forward to sharing the next 100 days.
Great. Appreciate that. Well, we'll wrap it up there. Thank you very much.
Thank you.