Good afternoon, everyone, and welcome to our webcast today. Before I turn the stage over to Doctor. Siva Seavram, I do want to point out that at the bottom of the page that you're on right now, there is a Q and A box. So as you have questions during Siva's presentation today, please feel free to type in your questions and we'll spend a few minutes after Siva's presentation answering those questions. So with that, let me now turn the stage over to Doctor.
Siva Seavram. Thank you.
Thank you, Peter, and thank you all for spending your afternoon with me. I'm going to be talking about this remarkable journey that we have had over the last 5 years, trying to build a full spectrum storage company, a data storage company that is the largest in the world, trying to put the technologies and the companies together for the maximum synergy. What did it take? What was sum synergy. What did it take?
What was the original thesis? And what has played out? So the agenda as you would expect, I'll take you through where we came from, why this happened. The time line just before and just after and in the time between 2016 when the companies came together to now in 2021, what's going on? What are those big synergies between flash and HDD that we are talking about?
Together, where are we headed? If we are trying to get the future of the data center, what does it take, what's on store, why do these 2 companies that came together make the best music together. We'll go through where this is headed for the future. So this building where I am in now, in 2015, the building that is on display used to have a big red sand disc sign on top of it. Inside the company, we were creating flash technologies and our motto used to be have you killed HDD today.
That's where all of the focus of Sandisk was. At the same time, unbeknownst to us, there was another company Western Digital that was looking at us. That was looking at the technologies together, looking at to see how a full spectrum storage company that was doing both HDD and flash, together what can they do. And that bet was a very large bet that Western Digital made. They made the plate to acquire a company that was 40% smaller in revenue in revenue, and they paid almost 50% more than their own market cap to acquire that company.
A $16, 000, 000, 000 acquisition in 2016 created the current Western Digital. The thesis was 2 storage solutions together to satisfy a very wide range of customers all the way from retail to mobile to client to enterprise to new evolving architectures. How do we make sure we create solutions for all of those customers? And how do we continue to delight our customers? But if you were there in 2015, the HDD story was not very obvious.
In 2015, HDD was in decline in the total number of spindles sold, total number of drives. Western Digital had just consolidated with AGST to form the company, and the word that was used was rationalization of the footprint. We had just closed down our Odawara Drive plant. We were consolidating media factories. We were consolidating the footprint.
We had shut down the Kuala Lumpur facility. There was a worry what's happening to the HDD business. As mobile was going up, the next big battlefront was client SSDs or client computers where HDD was also in decline. There was a genuine worry. And at that time, NAND was booming.
Year over year, 35% growth, cost reduction in the 20%, 25%. So we were wondering if HDD is going to be relegated to the dustbin of history just like the 8 track tape or a floppy disk or a CD ROM. Where is this headed? How the times have changed? In the ensuing 5 years, after that acquisition, even as mobile was maturing to be a stable, smaller, regular growth business, the migration to the cloud started.
Big amount of data flow. Instagram happened. Netflix happened. Amazon Prime happened. IoT happened.
The number of newer newer applications that were producing data started expanding. The amount of data that was flowing to these data centers started multifold expanding. And that resulted in where we are today, a 35% growth year over year, either it is in NAND or in hard drive. Magically, that seems to be the number, about a 35% growth, meaning every 2.5 years or so, you are doubling the amount of bits needed. And towards the end of the decade, these 2 have again become full growth industries.
HDD was no longer in decline. The number of spindles was going up. The number of medias, number of head, the number of drives, whichever way you started to look at it, by the end of the decade, hard drive had become a growth engine again. We were starting to look at opportunities for carefully investing additional capital, carefully investing additional R and D dollars so that this growth engine could be sustained. If you looked at what happened in that same time period, in flash, the big driver was going from 2 d NAND to 3 d NAND.
That small 128 Gigabit Die that was the sweet spot in 2 d NAND, suddenly we are now making terabit dies, 8x growth in that small the basic unit on which drives are being built. But lo and behold, on the other side in hard drives, Helium had revolutionized growth at the same time. We were going from 4 to 8 terabyte at that time. 8 became 12, 12 became 14, 14 became 16, 18, 20. That 14 terabyte drive was a seminal drive.
That 14 terabyte completely transformed the hard drive business and the data centers where that became the sweet spot. Both technologies, hard drive and flash, were showing consistent year over year cost reduction and consistent year over year bid growth. And 1 company would be in the forefront of all of that change. Along with our partner, we were driving the 3 d NAND transformation and on hard drives leading the areal density wars so that we could produce the number of products that were in that spectrum, whether it is in client, whether it is in gaming, whether it is in retail, whether it is in network attached storages, whether it is in very large enterprise drives, whether it is in performance NVMe drives, whichever way you go, Western Digital was now at the forefront of delivering products on both HDD and Flash. So look at what happened in that time period.
The big growth in that same 5, 6 year time period, Flash was growing at 10x. And HDD is starting from a very high bar, was continuing to grow at 3.5x, 4x. In that time period between 2016 2021, in 2016, the entire industry was about 100 exabytes for flash. This year, Western Digital alone will ship that much in flash. Similarly, in hard drive, we had gone from 600, 700 terabytes exabytes to now we are talking about 2, 3 zettabytes of growth.
Applications, use cases, they all had gotten integrated in the data center. When it comes to search, when it comes to database acceleration, Flash needed to be in the forefront. When it comes to video, when it comes to images, when it needed streaming, HDD was the place to go. Together, they were developing symbiotically. No longer was this a replacement business.
Whereas in the prior generations, as flash cost came down, it was replacing HDD. In the cloud, they were existing side by side growing at similar competing 30% to 35% growth rate. So what was the thesis? What was the thesis with respect to synergies? Given that these 2 are growing together, why do they need to be in 1 roof?
Why do they need to be together? What do they need to be gaining from each other for this transformation to continue? Let's talk about 4 major categories. Clearly, the financial synergies are always there. The fact that you put the companies together, HR, Legal, Finance, put the organizations together, get synergies, obvious procurement and those kind of synergies, we took care of all of that.
Those are fine. But the bigger synergies came from 4 major categories. I'll take you through them, where we went in the marketplace, how we ran internally, operationally, what new technologies got created and in the end, what does it mean to the customer. And you can see the synergies as we develop through these. Let me go through this 1 at a time.
When it came to the market, if you take nothing out of this whole seminar, the 1 word I want you to remember is channel. Channel, the ability for a mom and pop to order a 100 drives here, 200 drives there in very, very, intricate rich marketplaces around the world was carefully created by hard drive marketplace. The hard drive business as it grew and consolidate over a long time, Western Digital itself is a conglomeration of COMAG, Read Write, AGST, IBM, all of these came together. Over those consolidations, hard drive had created a set of distributors, value added resellers, various ways for people to access that products around the world. When that rich go to market synergy was introduced to flash, you could see the explosion in our sales motion, in our ability to reach customers far and beyond what we could do earlier.
Channel richness is a synergy that people do not realize how important it is when the 2 came together. On the flip side is retail. Sandisk had a very strong retail presence. WD had a retail presence. Together, today, they are a $4, 000, 000, 000 business a year.
Retail was considered a declining business, no longer. Because of the synergy between the 2, because of our ability to mold the business, because an external hard drive and an external SSD could be shaped between the 2. Retail is once more a growth business. We have 400, 000 retail outlets in 80 countries whether you want to buy an external hard drive or an external SSD, whether you want to buy a USB drive or whether you want to buy a micro SD card. It doesn't matter.
Those channels work together. What this does however is something that we least expected before but we now make full use of it, yet understanding. These are highly correlated marketplaces. These markets affect each other. We can see storm clouds gathering or wind gathering behind our back well ahead of anyone else.
Our ability to understand because of what hard drive is telling us what it is going to mean or when you look at flash and say what is it going to mean for hard drive, that predictability together we're able to shape the marketplace. Market is a place of synergy that when we talk about HDD and Flash, we don't often think of. Between our ability to utilize the channel, ability to utilize the retail marketplace, ability to understand the highly correlated marketplace, we have a huge advantage of able to move products and bits across the different channels. When a WD Green or a WD Blue or a WD Red or a WD Purple, we are able to take products that you could go from hard drive to flash and flash to hard drive and address those specific markets through this very, very rich common sales force through that very, very integrated channel. So that's the first 1 in synergies.
Then let's look internally with respect to the way we run our own operations. Clearly, the fab part of the 2 businesses, the flash fab as opposed to a smaller head fab are very, very different. We use the JV for the front end of the fab in flash and we make our own front end heads and media and our fabs here. However, once these come and we move into the back end where we assemble and test and deliver the products, remarkably similar in the way the manufacturing is done. There, the learning between the 2 organization is substantive.
Our geographic footprint is similar. So if you go to Malaysia, we could be in Johor, you could be in Penang, you could be in Thailand, you could be in China, whether you are in Shenzhen or in Shanghai. When we need to look at labor forces, whether it is in manufacturing capabilities, whether it is with respect to supplier management and procurement, there is a lot of synergies. Flash had gotten a lot far ahead with respect to automation in its factories. By 2015, the back end factories had become lights out factories, highly automated lights off factories.
Now we are learning and implementing the same things in hard drive. The hard drive back end factories are becoming more and more automated. Given the volume that we run, in a flash factory we run about 5, 000, 000 single die equivalents per day, Per day, 5, 000, 000 single die equivalents. There, AI and ML become useful. The volume that you are starting to run, taking care of machine vision and artificial intelligence and getting this together as a manufacturing efficiency that's shared between hard drive and flash, knowledge transfer, information transfer.
So together, the synergies on our internal operations have been dramatic. Actually, they now work as 1 organization. Hard drive and flash manufacturing and operations within the company or 1 organization. Let's talk about technology, an area that is close to my heart. Both hard drive and flash are block storage devices, which means they are limited by the interface that with which they talk to the rest of the computing system.
Whether from the hard drive days with the SATA or the SAS or the newer NVMe in in in in in in in in in SSDs or the UFS and and and the eMMCs and the USBs and all of these standards, both companies have been dramatically good in developing, implementing and driving standards. Now we have a common organization that drives our best known methods with respect to standard development. Clear, easy win. Let's get a little bit more technical. Both of these technologies are signal processing driven technologies.
You need to be able to sense extremely small signals. Error correction and checking, ECC, was developed actually the modified Hemming code first came into HDD that then came into flash. The first place we look at it, okay, let's go look at it together. Where can we develop the redundancies that are needed so that we can minimize wastage in additional bits when it comes to both hard drive and flash. So now the ASICs, the SOCs, they are driven by 1 organization.
Commonality in our ASICs and technologies and towards creating those engines that manage these memory bits, that's common as you would expect. Now look at it little farther out. When the companies were separate, you make this brain dead ideas such as hybrid drives. A hybrid drive was supposed to be NAND and HDD together so you can get the best of both worlds And that's what an outsider's view of it when you put this together. Clearly what happened, you got the worst of both worlds.
So the hybrid drives died in the vine. They never took off. Now when we are 1 company together, you start to think about where can we take it, you come up with NAND enhanced hard drives, where you think of using the NAND to reduce the amount of DRAM that is needed, to reduce the amount of metadata that needs to be stored or can optimize it, to drive use the use the use the memory for optimizing the drive functions. These are products that are coming right in front of us. And even more exciting are new memory technologies.
Imagine taking the same materials that have been used in HDD media for the last 20 years, Use that to create a switching device, put it in a solid state architecture that I can make a 3 d device, that's MRAM. MRAM is that device that leverages the strength of both organizations. You are going to see products such as this coming in front of you all the time now, now that the 2 companies are working together. So in technologies, whether it is in standards, whether it is in developing ECC, whether it is in developing ASICs and SoCs or now creating new products as NAND enhanced HDD or materials that use magnetic materials to make solid state devices. Technology is there for the development.
So what does this all mean for the customer in the end? 1st and foremost, the biggest realization is there is a common set of customers, A common set of customers with a common set of procurement teams, common set of processes, common set of things that we do business together. Just that customer intimacy. Imagine, I'm selling to the same customer $200, 000, 000 of HDD and $300, 000, 000 of Flash. Suddenly, a $500, 000, 000 engagement is a lot better than any of those $200, 000, 000 engagements.
Same with having a $1, 000, 000, 000 hyperscaler. When you are in a relationship where somebody is spending $1, 000, 000, 000 with you, your whole customer intimacy changes. The way you think about it, the way you deal with them, the way you interact with them, you think about the future, all are dramatically different because of that customer intimacy. 1 of the prime examples of this is the client SSD marketplace. The client SSD marketplace, Sandisk had a spectacular SSD set of products.
However, we could never cross that 12%, 13% market share. Now look at it. When the 2 companies are together, with the depth of relationship that the HDD guys had with a OEM customer, a PC OEM customer, where we know what their testing methodology is, where their labs are actually reproduced in our own laboratories within the company. Now you are able to get in there, get things done and qualify and ramp so fast that now we are talking about a 30% market share. The leadership in client SSD marketplace is a big result out of the fact that we work together with the same customer.
The same thing happens when a large hyperscaler wants to venture into consumer products, whether it's gaming products or AR, VR headset, wherever. A hyperscaler ventures out into it. Where do they go next? They go to ask that $1, 000, 000, 000 supplier of theirs to say, hey, you have a flash product to go with it. That's the reason today we lead in gaming.
That's the reason our augmented virtual reality glasses business is growing quarter over quarter in a dramatic pace because of that customer intimacy, the same set of customers when they make a broad set of products that we are they are with us. So you saw the 4 major things I talked about where the synergies are, whether it is in market, whether it is internal operations and manufacturing, technology, customer, these were not explicitly understood when the companies came together. And now together, we are now ensuring that in a data center, the future of flash is flash. The future of HDD is HDD. It is not no longer substitutional.
We know what is coming where. When you go to the data center, when the growth happens, we know the use cases. We know the data flow, we know where we need to introduce what products. This is a big revelation from where we were when HDD was going down in volume to now being a growth business. We can now see how HDD and flash in the data center both thrive, both grow rapidly.
So given Western Digital's knowledge of where the technologies are going, When the customer is developing new products, new systems, new architectures, whether they ask us to work with them, whether they come and ask for components, we can develop scalable, composable infrastructure products involving both flash and HDD. That's the big power of them together is that we have knowledge of where the technology is headed, where the products are headed, where the customer is headed, where the architecture is headed. We can deploy it in any way that the customer wants. 1 prime example is this idea of ZNS, Zoned Name Storage. In HDD, about a decade ago, we introduced a concept called shingled magnetic recording, SMR.
Today, the 1, 2, 4 terabyte devices are all SMR. They are managed by the by the system itself. Even in capacity enterprise drives, we are slowly headed towards more and more and more incorporation of SMR. That idea when translated into solid state becomes ZMS. Exactly the same idea of shaping incoming data so that you can sequentially record in a way that is compatible with the media.
So the media and the data are matched with each other. That's ZNS. It allows a dramatic reduction in the amount of DRAM needed in reducing the cost, improving the performance. ZNS has become a standard. Just like we talked about earlier, we are now delivering ZNS as an open standard.
Much as I don't like it, but it still warms my heart, a competitor is now announcing a ZMS product. We know what it is to establish a standard in the marketplace. We know where it is headed. This is a classic example of an SMR technology from HDD coming into solid state and becoming a standard. So you can see where this is headed across the spectrum.
Look at NVMe Ore Fabrics. For the ability to network a large number of solid state drives, our computational storage, where if there are limited primitive set of commands that need to be done, let's say encryption, decryption, those kinds of things can be done faster in storage than having to move all the data up to the compute and back. Whether that computational storage resides in a controller, in a drive itself or on a fabric sitting outside being able to handle multiple storage devices. Storage accelerators. With the evolution of CXL, we now are talking about multiple accelerators hanging off of a single bus.
We are able to access coherently memory. Now we are able to grade that memory. I now need to be able to say, okay, there is a fast DRAM, there is a set of memory above it, there is a set of storage above it, etcetera. We are able to work with people on it. Or the new novel and smart network interface cards that are coming up.
Every 1 of these places, as the customers develop new architectures and technologies, we are there with them with our knowledge of what it is that we can deliver. So this knowledge of where flash and HDD are headed together. It's only a matter of time before we will have gone from 24 layer to 48 layer to 96 layers to 110 layers to 100 and 62 layers to 200 plus layers. Flash is going on its trajectory. Hard drive, you saw the same thing from 8 to 14 to now 20 to 24 to soon on a 2 inch form factor 50 plus terabytes.
We know when we are going to be introducing these. Together, we know where the flash technology is headed what are its characteristics where is the HDD is going where are its characteristics and that brings us together to be at the leading edge of storage. This is the reason the 2 companies came together. This is the synergy that we were looking for, taking technologies so that we can help customers going forward, taking technologies and their road maps and be able to predict where the technologies are headed, Use the market synergy. Use the customer intimacy.
Make sure that we are manufacturing together and develop technologies together. This is the new Western Digital. Thank you.
Okay. Thank you very much. Siva, we do have a few questions that have come in. Just give me a quick second here to organize them. So first, Siva, thank you for the presentation.
You talked about the bet that Western Digital made combining these 2 storage technologies into 1 company, all of the benefits you got from that. But can you please explain why a few months ago, Western Digital decided to separate the businesses into 2 separate business units.
When the new management team came here together, 1 of the first things we needed was the thesis that we developed is still true, but we did have a few execution fumbles in the past. What was needed was focus on products and portfolios. So the 2 new organizations that have come up, the Flash under Rob Sodaburi and HTD under Ashley Gorakhpurwala, the intent is to justify this thesis even with the focus on the 2 product divisions. So, what we do need to get to deliver is to deliver on the promise of this thesis, which is still very, very true given the growth in the marketplace, given the growth in the product portfolio, given the growth in the systematization of delivery mechanisms, the 2 product organizations still need to have considerable focus so that we can execute on each 1 of these markets, whether it is in retail, whether it is in emerging data center architectures or in mobile or in client or in the new IoT spaces that are coming up. That's the reason the the 2 the 2 new business units have been formed.
Okay.
Thank you. Here's a question that just came in. It's focused more on a specific end market for us, which is the Cloud. How should we think about demand cycles for SSDs and HDDs in the cloud? Are SSDs more closely tied to compute cycles and HDD more storage cycles?
Or just at a high level, how should we think about that?
The way you think about it is, it is in the same marketplace. They grow symbiotically. HDD has a funny sort of ingestion digestion cycle that is in each hyperscaler. Each hyperscaler goes on a release, they go ingest and then they digest the product. SSDs given the volumes are not quite as high as the HDDs and bits are a little bit more linear in their adoption, but the qualification cycles in both are very, very long.
So they are less, about cyclicality than more about the hyperscalers' own growth and their ability to qualify the next generation technologies. They are not as cyclical. Okay.
I have another end market question. This time, it's more on the retail side of the house. Can you provide more details on how WD is able to mold its $4, 000, 000, 000 retail business into a long term growth business?
Yes. The retail business has continued to be a jewel in our business universe. We have always worried that this may be a over time diminishing business place, and that has not been the case. We have been introducing new products on both sides, both in hard drives and in flash. We have seen external SSDs grow from really no volume to now being we have about close to 40% market share on it and continue to grow dramatically.
And it's coming very close in density to the hard drives, and we are able to product switch between those 2 depending on the flash cycles and the costs and the prices in the marketplace. And the markets are also the delivery mechanisms are very similar, whether it is e tail or retail or on our 1 of our big, retail outlets that we work with, we are able to get shelf spaces that are complementary. So with the brand identification, the uniqueness of our brand, we're able to continue to grow the retail business.
Okay. Here's another question that kind of leverages off your presentation a little bit. But you talked about the bet that Western Digital made to combine these 2 technologies under 1 company. What, in your opinion, of the 4 boxes that you showed, gives Western Digital the greatest leverage or the greatest opportunity for success?
It's like picking among my 4 children, but customer, It's all starts with the customer. The customer intimacy, the customer the minute we have a large customer working with you in a codependent fashion, they open up and tell you where things are headed, and that makes the biggest difference in the synergy between the 2,
okay? Here's a technology question. MRAM has been talked about it as an emerging technology for a number of years. What is it in your view on timing or what's needed for this technology to
have mass market adoption?
In addition, what benefits Who
would Whoever asked the question, they're being kind to me in saying some number of years. It's some number of decades that MRAM has been talked about. What has changed is the idea in a solid state device, when you develop an architecture, you need 3 things. You need a cell, you need a selector, and you need an architecture to go together with. We did not have a cell till the HDD folks came and showed us what can be done.
HDD folks had only a cell. The flash folks did not have had an architecture but no cell. So the 2 get together. Now the 3rd leg of this stool that has changed is the marketplace. In the hyperscale, data is flowing so much that each workload has become large enough that we can now start looking at specific workloads for whom we can introduce this technology Without having to change the entire software stack for everyone, we are now able to deliver a product that is custom developed for some large vertical workload.
These 3 together, the presence of the HDD and the flash and the customer makes MRAM an ideal candidate to grow.
Okay. Let's stay on this technology theme for a second here. Western Digital is investing in bringing NVMe over fabric as well as or is Western Digital investing in bringing NVMe or computational storage to market? Are you really investing in the processors? Are you going to more supply the memory that goes in the behind those processors?
I want to be careful the way we answer it, because, it is not eitheror. In all cases, we work with customers. In many places, customers would only ask for certain components from us, and we work with them to integrate. But internally, on our own research efforts, we will do the build the entire system to make sure we are building a fully compatible system. So with respect to a computational storage or a new storage accelerator, etcetera, Our job is to produce compatible components that make sure that, that system works.
But we will work on the entire system to make sure we do have a solution for the customer.
Okay. Let's bounce back here to end markets. 1 of Western Digital's recent growth opportunities has been in the enterprise SSD space. How are you thinking about the progression and leveraging your common customers to really grow your exposure within the enterprise SSD market?
Yes. This is a very, very important vector for us. Clearly, we are shipping millions of hard drives to these customers every quarter. Clearly, these customers trust us and have worked with us for a very, very long time in our hard drive business. We came through a very difficult period with respect to our enterprise SSD products.
These customers stayed with us during that time. I do not think it would have been possible for us to grow as rapidly as we have grown in our enterprise SSD space without the support and trust of our customers who are there with us during all this time. Now we are leveraging the 2 that we are able to ramp very rapidly in all of our hyperscale partners with solid state drives. Okay.
And I see 1 last question here. New innovations create new market opportunities. Cryptocurrency is a great example. Can you go into a little bit more detail on how Western Digital can leverage both its HDD and its flash business to better position itself to capture, more opportunities in this market as it grows?
So cryptocurrencies for us is another vertical. We treat it as another emerging vertical business. Any vertical business, we look at it the same way. How do we best serve the customer with my entire suite of products? Because we are the only company in the world that has got that combination of both hard drives and SSDs, we are able to come back and say at any application, what's the right product to ship, what's the right combination of products to ship and what is the right future for this so that we make the right investments for that in the future.
Whether we customize a product, we develop a product, that can be done because we have both of these technologies. And that's exactly how we have been treating the cryptocurrency market.
Okay. And with that, that's all the questions we have for right now.
Thank you, Peter. Thank you all. Thank you for spending your afternoon with me.