Hello, everyone, and welcome to the technology update from Intel and Micron. We have some exciting news to share with you today. I'm Lindsay Sutch with Intel's Global Communications Group. Before we get started, I would like to point your attention to a few legal disclaimers and also read through our Safe Harbor statement. During the course of this presentation, Micron and Intel may make projections or other forward looking statements.
Forward looking statements are projections and other statements about future events that are based on current expectations and as a result are subject to risks and uncertainties. Following today's keynote, we will have a question and answer session. If you're joining us from the webcast, please submit your question through the Ask a Question button. Now I'd like to welcome to the stage Rob Crook, Senior Vice President and General Manager of Intel's Non Volatile Memory Solutions Group and Mark Durkin, CEO of Micron.
Thank you, Lindsay, and I'd like to extend a warm welcome from Mark and myself here for joining us this morning, and thank you for coming. What we're here to talk about today is a development in technology delivered by Intel and Micron that is fundamentally a new class of storage and memory technology we call 3 d Crosspoint. And we'll have a better understanding of what 3 d Crosspoint is. Here we go. Okay.
Let me start over. So first, let me extend a warm welcome from Mark and myself for joining us here this morning, and thank you for coming. We're here today to talk about a new class of memory and storage technology that Intel and Micron have been working on for some time. As you may know, Intel and Micron have had a collaboration in nonvolatile memory for the last decade. And we've worked on a number of different technologies as part of that collaboration and you've seen the NAND technologies brought to market.
But this is actually different from that. This is a fundamentally different technology that we call 3 d Crosspoint. And you'll have a much better understanding of why we call it 3 d Crosspoint in the next few minutes. But it is fundamentally a different technology from the technologies that we're used to in storage, NAND technology or in traditional memory technologies, both from the physics in the way it operates as well as the attributes that it brings to the computing system, the speed and the performance and the capacity density of the memory technology. And it's the first time a new class of memory technology has been brought to market in quite some time and we thought we might take a moment and put that in perspective.
Yes. Thanks, Rob, and thank you all for coming. It's really amazing to think that in the 50, 65 plus years since the invention of the transistor back in 1947, we really only had 7 different classes or fundamental different types of memory in the industry. And each of those memory classes has really bought different attributes to the compute hierarchy or to the way we process data and the way we retrieve store and retrieve data. If you go way back in the beginning, you had early read only memories that were relatively dense, but could be written once and really didn't have the ability to feed data at a high speed to a processing unit.
Over time, we had SRAMs that were much faster but volatile, so the data came and went. You had to restore the data over time. We had one of my favorite memories DRAM which is system main memory today in many, many applications and has a nice combination of speed and density, but not non volatility. And so over time we had different kinds of non volatile memory. When I joined the industry over 30 years ago in 1984, we had the 1st NORFLASH product into the market.
And that was interesting because it brought a reprogrammable non volatile memory in a way that could interface directly with the processing unit. And then most recently NAND flash way back 2.5 decades ago. So Rob's right, when we talk about what we're doing today introducing a new class of memory, it truly is revolutionary and very, very exciting because it's bringing not only a new set of characteristics to the marketplace, but it's going to enable a whole new universe of applications and memory architectures and compute architectures. I think it really going to change the world and the way we think about what's possible in electronics. So I talked about some of the previous memories we've seen and how those different attributes allow those memories to play in certain pieces of the compute architecture or certain tiers of memory to service processing.
What's really exciting about the memory we're introducing today is that it's high performance, it can be very dense and it has non volatile characteristics, data retention, endurance characteristics that Rob will illuminate in more detail here in a minute that really allow us to do things that we haven't been able to do before. When you think about all the various mega trends that we see in the world around us today, whether it's mobility and all the things we like to do while we're on the go with our smartphones or the connectivity we want to have or whether it's the cloud and big data and bringing together large amounts of data to solve problems that previously we couldn't solve or networking and the Internet of Things and all the data that gets generated machine to machine and passed around the network in order to bring more data and more knowledge to the world. We really have an explosion of data. In fact, fun fact for you, by 2020, we're going to generate another 44 zettabytes of data. Does anybody know what a zettabyte is?
I didn't, I have to look it up. A zettabyte is 1,000,000,000 terabytes and a terabyte for those you don't know is 1,000 gigabytes. So we've got 1,000 and 1,000,000,000 of orders of density of data coming beyond what we typically store in a single chip today. So really a phenomenal amount of data being created in the world. That's really only useful if we can get it close to the processor and do something with it, either add it or subtract it or compare it to each other or use it in a way to really create useful knowledge as opposed to random information.
And that's what this new technology is all about. It's about getting that data somewhere where we can get a lot of it close to processing and into a useful format. Might turn it over to Rob to talk a little bit more about it.
Sure. So this breakthrough in technology is something that we're very excited about, both the way it works and the results it delivers. And so it's something that we view as a new class of memory because it's not just a little bit faster. It's not a simple extension of one of the existing memory technologies. It's a lot faster, it's a lot higher endurance.
In fact, it's up to a 1000 times faster than today's existing NAND technology. It's up to a 1000 times higher endurance than today's NAND technologies, fundamentally different than NAND technology. It means you get to write it a lot without worrying about it. And then when we compare it to conventional memory, it's 10 times or more the density of today's conventional memory. And yet it's nonvolatile, so it can be used as storage as well.
And this is a fundamental breakthrough in delivering capability to the computing platform that enables us to scale computing even further than we might have expected. It brings new capabilities, so that when we scale the CPU and the performance, the memory and the data can scale with it, be close to the processor and we can turn that data into information to make companies run more smoothly, to make cities run more smoothly and deliver scientific breakthroughs. So we're pretty excited about this. This is something that many people thought was impossible and many people gave up trying to accomplish. And it takes the power of companies like Micron and Intel and the sustained investment to work at something that other people thought was impossible.
And it is pretty exciting, the results, but it's also pretty exciting for the technologists about how it works, because it does work fundamentally different than today's memory technologies. And I'd like to talk to you about that. 3 d cross point memory is a unique and innovative cell architecture that is a cross point memory that is able to stack in the 3 dimensions, hence 3 d cross point memory. And this picture that you're seeing here is an example of a small section, obviously, much higher scale than our memory technology of our first products. So our first product that will be delivered on this technology will be 128 gigabit product that is stacked in 2 dimensions there.
So it will be 2 layers high, total 120 gigabit, which means there's 128,000,000,000 memory cells on each chip, which is about the same capacity as high volume NAND based technologies have today. And so what's so interesting about this is its unique set of inventions and innovations that enable us to get to that cross point architecture. We are able to we have a unique switch and memory cell design that allow us to eliminate the transistor from this and scale to much finer dimensions, enable that cross point architecture. And a cross point is effectively you could think of it like a screen that you might have on your door, where 2 wires cross wherever there are 2 wires that cross, we have a sub microscopic pillar, if you will, at each one of those junctions of material that consists of a switch and a memory cell. And it enables us to select each memory cell individually.
And that enables us to read and write individual pieces of memory, not have to block arrays like traditional storage and then rewrite. It allows that to be very, very fast. So we have a fast memory cell in there, but it's interconnected by metal lines that allow us fast access to that memory cell. And then when the memory cell itself switches quickly and is activated quickly by the selector, we get a totally in total, we get an architecture that delivers fast memory. Now there's some breakthrough in the materials and how this works.
And so it's not just important to have a unique memory cell that works quickly, but also a compatible switch. And that's what enables the cross point architecture, the combination of those two innovations at the same time. And this enables us to get these thin columns of memory, which allows it to scale very well in the x, y dimension, if you will, get the pillars closer together over time, as well as expand our density by stacking more layers of memory over time as well. And that allows us to get a much higher performance. The materials themselves, the memory cell itself doesn't store its memory in terms of trapping electrons, either in a floating gate memory or like we would have a NAND or trapping electrons on a capacitor, put storing electrons on a capacitor.
It doesn't use that methodology. It stores its memory in a fundamentally different way. It uses the property change of the memory cell material itself. And it uses all of that material, and we call that a bulk material property change. And that allows us to do better scaling as well, because we're not using just a portion of the memory cell, we're using all of the memory cell itself.
And that allows us to get to smaller dimensions, because we're using all of the material. So it's pretty exciting, dramatic improvements in performance and scalability and a technology that operates fundamentally differently than the technology that exists today. It's not electron based, it's material based and it's a cross point architecture and that gives us the performance and the scalability we want. Now of course, you want to ask yourself, what does that mean to us? We're pretty excited about the technology as you can tell.
But at the end of the day is what can we do with that technology. And this is something that is a breakthrough. And one of the things we're most excited about, to be honest with you, and the reason we're talking with you here today is that is what people can do with this, what they can innovate on top of this that we can't predict today, because we believe that this is a fundamental game changer and will enable new things to be done with computing. But at a category level, what are the things that will benefit from this? Things that require much more memory than we can deliver today.
So I think a massive in memory database, things that require better service level agreements, faster system recovery time, because this memory is not volatile. When something happens, the power is removed, it does the data doesn't go away and things can recover much better, so better service level agreement for folks. And then storage that has very low latency, so really fast, low delay access to storage and things that require high endurance. And we do think we are excited about the things that haven't been thought of yet, but we did in fact think of and have been working with some folks about some of the applications that might be exciting. We'll talk to you a little bit about those.
Thanks,
Rob. Yes. So as Rob points out, the really exciting thing here is that when you can get this much memory close to a processor and you don't have to refresh it because it's non volatile and it's high performance and you can read it and write it many, many times, you really have the opportunity to do an amazing set of things that you couldn't previously do. And we don't know what all of them are even going to be yet, but for those of you that aren't high performance computing geeks in the audience, Even your kids know that gaming is limited by the amount of data flowing to the processor. So we've all got children that are out there playing their video games every day.
And what you'll notice is that the images are being rendered real time as data is fed to a processor. And then every so often, the character in the game might move to another scene and the game will stop and a little video will run as it reloads more data up close to the processor. That is the fundamental bottleneck in terms of the amount of data you can get close to the processor. You'll also notice sometimes in those games maybe windows in a house in the game or a TV and typically there's a frozen image in there because the memory just can't keep up with the images within the images within the images that you would really find in a real world. And so as we deploy technology like this into the gaming world, you'll see you won't have those little videos, people will walk seamlessly from one room to the other and the TVs and the games will actually have real content streaming through them or there will be activity outside the window, outside the house, etcetera.
So we'll have a much richer gaming experience. But it really goes beyond that because that technology is really very, very similar to what you do in any advanced simulation environment. I was joking with Rob in the back a little earlier about too bad we didn't have this great technology when we were doing all the materials research to make these switches and memory elements work because had we had it then, we probably would have got there a little bit faster. And so a really phenomenal high performance computing surge I think we're going to see with this new memory as we get more memory close to the processor. Think of all the ways that you use pattern recognition in your life today, whether it's speech recognition on your phone or in your car, whether it's voice recognition, face recognition, all the biometrics that you want maybe want to use in the future for security, the games you play and hand gestures, all of those types of pattern recognition activities require to get really high fidelity require large data sets to be matched with other large data sets.
And again, in this realm, a new memory like this can deliver a completely new world in terms of security methodologies and recognition activity. So I think we'll see a sea changes as this memory is deployed into applications like that. And finally, one of the great miracles going on around us every day is the rapid improvement in the understanding of the human body and medicine. And a lot of that being driven by genomics, but really underlying a lot of the advances we're seeing in medicine and health today is the ability to understand what goes on with DNA and to understand how different pieces of that large data set impact different medical outcomes and be able to manipulate those and operate on them in a way that creates different medical outcomes. So I think within the medical world as we think about having a memory like this enable us to do real time gene sequencing or quickly determine whether a cancer we've been diagnosed with has a genetic component to its origin or is something else will really change the way we think about our healthcare and about how we can access it on a go forward basis.
Just three examples. There's many, many more. As you can see, we're really excited about this new memory. As I think back through my 30 years in the industry, it's always been I think a vision of those of us that make memory to have a true non volatile high density cross point memory that could deliver all this functionality and all this performance. And our relationship is 10 years old.
We've been working for it for less than that, but we've as a team we've been talking about it the whole time and we're really there. We have today product production in our joint manufacturing facility in Lehi, Utah. We will be rolling out products into the marketplace independently as we move through 2016. And we just couldn't be happier to be able to tell you guys about this phenomenal new technology and how it's going to change the world.
And we brought a special treat for you here today. We brought the first of the wafers from our new production facility, our existing production facility rolling these out. We thought we would take a moment and show you that new wafer. This is a wafer with 128 gigabit memory cells memory chips on it and in our production facility today. So thank you for coming.
And we have some more we're ready to do some Q and A for you, I guess.
Thank you, Rob and Mark. We're now going to hold a question and answer session. As a reminder, if you're joining us from the webcast, please submit your question through the Ask Question button. A few other quick notes before we get started. The announcement today is focused on the joint technology between Intel and Micron.
As such, we will not be discussing or answering questions on individual system products from either company, nor will we be taking any financial related questions. I would also like to point your attention to a couple of URLs on the slide, both the Intel newsroom and the Micron newsroom, where you can find the press materials and other information on the technology. And also if you're tweeting about the news today, and we hope you are, please take note of the 3 d Crosspoint hashtag that's also on the slide. With that, I'd like to open it up for questions. We'll start with 1 in the room.
Yes, Mark. You noticed something very interesting. You said that the product has started the manufacturer indefinitely. So can you talk about how the joint venture is going to lead to manufacturing? And was there any specific message that you want to relay by stating the product has been in production independently?
What I was really trying to communicate to you is this is a real technology. It's not a PowerPoint presentation. There is a real wafer here on the stage with us today. This is a manufacturable technology that Intel and Micron will be manufacturing first in our joint venture facility in Lehi, Utah. Beyond that, we'll have to see where we go together.
But it's been a great long term relationship and we'll do the early manufacturing together, take the products to market separately.
We'll take another question from the room.
Hi, good morning. You talked about the performance differences, endurance differences. Can you just help us understand cost per bit differences relative to DRAM? And then and is this technology let's say longer term more viable with EUV technology or can you get cost economics with conventional lithography technology?
Yes, sure. The technology is very scalable, right. And so we're at about 128 gigabits for this technology. And so hence, density is very, very high. It is nonvolatile and much higher density than DRAM, much faster than NAND technology.
And so you could put the cost somewhere between NAND and DRAM, of course. So from a cost per bit, it's likely to be in between those somewhere. And a but the actual costs will be resulting from the final products that we delivered to the marketplace in 2016. It is a very scalable technology, independent of EUV or what have you. And it has the same scalability challenges that many other long line lithographies have.
We'll now take a question from the webcast.
How does the 3 d cross point compare to 3 d NAND flash that you announced earlier this year? And how scalable and manufacturable relative to NAND and DRAM is this technology?
So maybe I'll take that one. 3 d NAND flash is an exciting new technology as well, but it's really a modification of the existing NAND technology. So what we're talking about today with 3 d cross point is a fundamentally new class of memory that has fundamentally new characteristics. The 3 d NAND is very, very exciting because it allows us to bring high density memory products to the marketplace and to continue the scaling that we've seen over the last number of years with NAND. This technology is exciting because it enables whole new ways of computing, whole new computer architectures and tiering of memory and some of the applications I talked about.
The characteristics that Rob alluded to are what differentiated the enhanced cyclability or the number of times we can rewrite the memory, the enhanced performance a 1000 times faster than NAND and the ability then to get that memory actually close to the processor and use it in a fundamentally new ways.
Yes, I think if I were to try and make it really simple, is that they're fundamentally different principles of physics involved, right? So the cells are completely different in how they work. And this 3 d Crosspoint technology is way faster, right, not a little bit faster, it's a 1000 times faster than NAND technology. And it's writable in small amounts. So we can write words and so it can act more like memory as well as storage.
But it's like NAND and that is nonvolatile. And the data stays when the power goes off. So you can use it as storage. It's just wicked fast storage.
We'll take another question from the room.
Yes, you guys both mentioned that this is the solution is going to be up to 1000x faster than traditional NAND. Can you sort of elaborate on that from a real world perspective?
Well, I think at a high level the way you might think of it is, NAND flash memory storage, if you're thinking about it from a storage perspective, is about a 1000 times faster than hard disk. And this memory technology is up to a 1000 times faster than a solid state, than NAND that's in solid state drives. And so the level of innovation that you saw there when folks move from rotating hard disks to solid state drives in the enterprise and data centers and cloud data centers and in clients is has the potential to happen again with a new technology.
Another question from the room.
Chris Preinsberger from Eweek. You said earlier that many people in the business thought that this was going to be impossible to make successfully. Can you name the 2 or 3 major hurdles that you had to clear during the development process to get this
out? Maybe I'll take a couple and Bob can take a couple. Well, obviously we talked about the material set. This is a fundamentally new switch, which required some very complicated and different difficult materials research and development in order to provide a switch that have the right on off characteristics and could provide the right current through the switch in this stacked in film format. The memory element itself obviously highly optimized to give us this bulk switching characteristic that Rob talked about.
And then the architecture and how this whole thing is put together also very, very unique. Rob? Yes,
I think that is those are the biggest hurdles I think in the process technologies is the invention of those technologies that are compatible with each other and integrated into a technology stack that works well together and can be produced in high volume. And that is those are the key breakthroughs and some of its hard work and persistence and then a lot of innovation and ingenuity in the technology development.
I guess I could add, the processing is we think is not fundamentally outside the realm of what one would normally do in semiconductor processing, but really, really tough. These are new materials and I just I can't say enough about the technology development teams at Micron and Intel that work together to actually make this whole thing come together. It's one thing to think about something and how it might work. It's a completely different thing to actually deliver to the marketplace.
I think also just to sort of on to the technology teams, it's one thing to pursue a problem that you know is solvable. And whether that's the next generation of something, right? It's another thing to pursue a problem for an extended period of time that you don't know is solvable. It might be impossible. And that it requires commitment and investment and vision and capability from these technology development teams that is hard to find.
And I think that's one of the unique things that has enabled this technology to come to life.
Question in the room?
Hi, Ian King from Bloomberg here. Thanks. And both companies know very well just how brutally competitive the memory chip has been over the years. You want to say anything about where this new technology does in terms of positioning you relative to what else you see coming from competitors, what else the industry is looking at, please? Thank you.
We're not or at least I'm not aware of anything coming from the competitors that fills this niche today. Now there have been a lot of papers written and there have been a lot of people talk about how they would like to create a memory like this. This one is real. It's in our fab. We've got a wafer here on stage with us and we plan to ship it to customers.
So we're not the only companies that are thinking about new ways of bringing resistive elements to form memory, but I think the memory we have created is
unique. Yes, I agree. And as far as being in the memory business, it's good to have something other people don't have, that's for sure.
We'll take one from the webcast and then switch back to the room.
Thank you, Lindsay. Will both companies be making products based on this technology? Will that mean that you'll be competing with one another?
Yes, we're both going to be making technology to making products based on the technology in a similar way that we do with NAND flash memory today. And it's worked out quite well, I think, for both companies. And we expect it to work well here. There is always the opportunity for cooperation amongst the teams on jointly developing products if we want to. But today, we're really talking about the technology itself and the disruption of this particular technology
on its own.
We'll take a question from the room.
Can you talk about where do you see bigger opportunity for Crosspoint in the whether it's in the NAND market or in the DRAM market? And also if the cost reduction trajectory for Crosspoint, is it any different from what you are expecting on the NAND over or DRAM over the next 5 years? And last point just from a historical perspective, it kind of reminds me of Sigma RAM like which didn't quite do very well, sounded promising long time back. And also like another new type of memory that was introduced few years back was the phase change memory and like that wasn't as very successful. So what gives you the confidence that this is going to be really successful this time around?
Let me take the first piece and Rob you can jump in. First of all, you shouldn't think of this as NAND or DRAM. You should think of it as a whole new class of memory. It really does fill its own unique spot. Now it can be used in more of a storage type of application or it can be used more as a system memory and we think it will be used as both for different applications and for different reasons.
But it really kind of fits in that unique spot. Now relative I'm not familiar with Sigma Rand, I'm sorry, maybe Rob is. But relative to phase change which has been in the marketplace before and which Micron itself has some experience within the past. Again, this is a very different architecture in terms of the place it fills in the memory hierarchy because it has these dramatic improvements in speed and facility and performance?
Yes, I think the key is that we're working on a problem that we have a deep understanding of the computing gap, if you will. And it is not some technology chasing a problem, it's actually a technology that we're developing together to solve a problem that exists, right? We didn't develop this technology without a vision in mind of where it's going to be. There is a challenge with storage and memory technology, where in the gaming example, the storage is too slow. So the gamers have to orient their game around the systems capabilities.
And if we can relieve that bottleneck by either making bigger memory or faster storage, then we can fundamentally change and free the gaming designers to be creative about how they have caused the game flow to happen as opposed to having, okay, I can do this much and then I need to load a new level, right? Because I'm constrained in this amount of memory or the speed of storage, for example. And I think as far as some of the other technologies go, I think many of them did 2 out of the 3 things we're talking about here. They were either non volatile and dense or they were fast, right, and non volatile, like which one of the technologies, but you need it to be dense, fast and non volatile in order to fill that hole in the computing gap. And most have got 2 out of 3.
And relative to the scalability question, this is early days, right? This is a fundamentally new technology. But when we look at the characteristics of the technology, we believe it is inherently scalable on a go forward basis. Straight line layouts, material sets that are bulk mechanisms that allow us where things actually in some ways get better as we scale and a bulk mechanism that really allows us to scale very aggressively. We fundamentally believe that this technology can evolve, it can be multi bit, it can do all sorts of things to continue to scale.
Yes, we can scale in the X, Y dimension as well as into 3 d.
We'll take another question from the room.
Hi, Jim Harrison from Electronic Products Magazine. You talked a lot about the memory associated with the processor. You see this having a big effect on solid state drives as well? It will be really fast storage capability. And so we could attach either from a storage or memory attach point.
So for the highest performance storage, it will be a great alternative.
Let's take one from the webcast right now.
Thanks, Lindsay. There's a handful of questions around the intellectual property, who shares it, who owns it, what is the agreement around VIP for this technology?
Yes. Today we're really talking about the technology itself and what its capabilities are. This was a jointly developed technology that we both have rights to and beyond that probably doesn't make sense.
We'll take one from the room.
Hi there. Pat Moorhead with More Insights and Strategy. Thank you for the disclosure here. One question I had for you is you did outline a few of the workloads or use cases, which makes sense. One thing that I think if I understand the technology correctly is that it really could enable the thought of edge computing or edge analytics.
And is there are you nesting under big data or am I just overthinking the potential of this?
Actually, I think those kinds of things, Pat, are very possible. It allows bigger data at a lower cost, if you will, and non volatility and things like edge analytics are potentially a great area for this to be used.
Another question from the room?
Hi, it's Steven Lawson of IDG News. How much did it cost to develop this? And some people are concerned about the future of fundamental breakthroughs like this. Can you talk about the role of pure research in this and the continuation of that kind of pure research?
Well, I don't think we want to talk about the details of how much we spent, but I can tell you that the partners here have been actively engaged for a number of years at a very significant level to make this technology reality. It didn't come easily or cheaply, but it's the kind of investment that I think both Micron and Intel have proven really strong capability of bringing fundamental discovery to the marketplace in the past and this is just another one of those.
Yes, I agree. It's a significant multiyear investment. We've been working together for a decade and investing on technologies, how to tease apart exactly where we spent what. And certainly, this technology has been under development as part of our own fundamental research labs in both companies, but it had been an active development inside the 2 companies since 2012. And hundreds of engineers involved in it.
Another question from the room.
Hi, Karl Ackerman from Cowen and Company. I believe you referenced earlier that this could potentially hit the market in 2016. But I guess how should we think about this new architecture potentially catalyzing your existing NAND business and does it curtail your 3 d NAND roadmap in any way?
So it doesn't impact our 3 d NAND roadmap anytime in the foreseeable future. 3 d NAND is still a very cost effective solution for storage of mass amounts of data. Again, this kind of fits in a separate category between NAND and DRAM and we think provides unique value both on the storage side and on the system and memory side.
Another question from the room.
Chris Leonardo from Heart OCP. You've already addressed the scalability and the density benefits, but what sort of improvements might we see in power consumption and heat?
Yes, it has it's a nonvolatile memory, right. So from the traditional power management elements of it and it's very fast. And so getting in and out of power states with more of your memory being nonvolatile is obviously a benefit there. And getting just getting in and out of the power states very, very quickly would be a power benefit. And then of course, it doesn't need to be refreshed.
And so you can for a very large memory sets, it enables you to avoid the refresh power as well.
Okay. We'll take one here from the room.
Great. Thank you.
The room and webcast, sorry.
Thank you. Just follow-up, would you have to upgrade any of the facilities in Lehigh and for Intel would you consider collaborating with other manufacturers if there is a capacity constraint?
So companies are going to make capital investment in Utah to support the manufacturer of this product.
So the Intel question? Yes. It's part of our joint venture in Lehi, Utah. That is a shared facility and we would make the investment there as we have been quite frankly with the NAND based technologies that we've been evolving there. And as far as sort of a technology, can we leverage some of the equipment authority there?
The answer is yes, of course. It is similar in nature to us moving to a new node of technology, like we're moving from 2 d to 3 d or something like that. And so there is some shared and existing equipment there, but there will also be some new equipment as well. Right now, our supply that's available from we have 2 companies, of course. And then the Lehi, Utah facility is a large facility that should be able to support this.
And then we can move to new factories as it needs. So we have no real need to collaborate with other companies. We have the wherewithal to invest in capital between the two companies. And we've got a lot of very large factory networks.
We'll take a question from the webcast.
Thank you. So you say 3 d Crestpoint is 1000x faster. However interfaces like SATA III and PCIe are already saturated. What are you doing to open up the bottleneck?
Yes. I think your first example is great. Why have we been so focused on migrating and driving the definition of NVMe and PCIe for storage is because it is a re optimization for the platform for NAND and unleashes NAND, but it's incredibly important for technology like this. And so moving to NVMe over PCIe for storage is really important and it eliminates a lot of that bottleneck that you would see in SaaS or SATA, for example, and it would be critically important. And then we will look at other attach points in the platform in the future, and we'll talk more about those products in 'sixteen.
We have time for one more question either from the room or the webcast.
Jim Handy here, objective analysis. I came a little bit late and you may have already answered this, but this 1,000 times faster, I am guessing is only write speed, it's not read speed. And I was wondering if you could embellish a little bit on how it compares on read speed, which constitutes almost exclusively what NAND flash does?
Yes. The way you might look at it is, well, first, you'll see it in the final products that come out in 2016, the actual. But in fact, no, it is a very fast read mechanism as well. And you would see that kind of performance benefit in reads in the memory technology.
Thank you everybody for joining us both here in the room and on the webcast. Appreciate you attending.
Thank you.