I can see too well. The light is very, very bright, but anyway, welcome to Direct Connect. Thank you for joining us today, and it is great to see so many good friends, partners, and customers, and thank you for all your support. We have a very busy agenda. It's not just speakers and many demos for you to engage outside, as well as technology showcase with more than 40 different partners, and please stop by, and I will be also stopping by to see some of them. It is a great opportunity to share our priorities and also our progress. More important today is to listen to our customer. I would love your feedback so that we can better serve you. You know, I've been the CEO for about five weeks now.
Then some of the people asking me, since I joined Intel last year, are you committed to the foundry? The answer is yes. I'm so excited today to share with you. I'm committed to make the Intel Foundry successful. And I know there are areas we need to improve. I'm determined to strengthen our roadmap and our partnership and our execution going forward. As we drive this improvement, we see a lot of opportunity ahead of us. Semiconductors are foundational to our modern society. And they power our companies. And they enrich our experience of daily lives. And this will only increase in the world of AI. And clearly, you can see from the chart, this is going to be becoming a $1 trillion industry in the coming year, 2030. And a lot is driven by the AI. And so this is a very exciting time.
This is an amazing opportunity for all of us to pursue. And Intel has a very important role to play. We are the only company that does advanced leading-edge semiconductor R&D and manufacturing in the United States. Our R&D investments are driving new process technology and advanced packaging solutions. We'll continue to boost our manufacturing capabilities. And we're working to strengthen our position in each of these areas. Naga and Kevin, my coworker, will go into detail shortly. But let me set the stage. And clearly, you can see the slide: R&D, process technology, advanced packaging, and manufacturing. These are all the important roles that Intel wants to play and to serve you. We're investing in key technologies. And this includes Intel's differentiated advanced packaging capabilities. Our EMIB and Foveros solutions are being used to drive better power efficiency, bandwidth, and cost for our customers.
And a combination of RibbonFET gate- all-around with PowerVia backside power is a key differentiator for our next-generation process nodes. As you can tell, the four areas of technology that were doubled, tripled down, and then so that we can really show the differentiating technology. First and foremost is the Intel 18A. It is supporting the launch of our Panther Lake products. Our first Panther Lake SKU will be launched at the end of the year. And we'll have more SKUs coming in the first half of 2026. We'll continue to advance our 14A, which will enable us to serve a broader set of the market. We're working closely with key, I call it the node drivers, customers, to help us define the features and KPIs that will make 14A the best-in-class node. We also continue advancing in manufacturing.
As you can see from the slide, we build a diverse and resilient global supply chain. As you can see, it includes robust footprint across the Western Hemisphere, and we have the ability to expand the activity at several of these sites based on customer needs. I believe this will be the key strength of Intel moving forward, particularly here in the United States. I'm very pleased that the Trump administration has made American technology and manufacturing leadership a key priority. We are committed to working with them to achieve the shared goal, but to do that, we know we need to up our game. We are up against some strong competitors or partners, and we are focused to create the engineering-first culture of continuous improvement. We have many strengths on this to build. Intel technology powers the majority of PC, world PCs, and data centers.
This is a result of R&D, process technology, advanced packaging, and manufacturing investment I just spoke about. The past achievements do not guarantee future success. Our job now is to apply these strengths to the new way to meet the needs of the broader market. Every customer has their unique way of building products. We need to do a better job, make it easier for all of you to use our technology. We will be more rapidly embracing industrial standards, EDA tools, and best design practices. There are four pillars of design enablement that I'm most focused on: intellectual property, we call it IP, design, digital design flows, design for manufacturability, and design for yield. Let's talk about them each, beginning with IP. IP is a key to accelerate development, ensuring the right performance and compatibility, and enabling the first-time-pass culture through validated designs.
One of the industry IP leaders is Synopsys. They are a strong partner to Intel. And right now, at this moment, I'd like to welcome Sassine Ghazi to the stage now.
This is amazing.
Nice to see you. Glad to do it. First of all, I'd like to welcome the CEO of Synopsys, Sassine, to come here. Thank you for being here.
Great to be here with you.
There's a lot of great collaboration between Intel and Synopsys to get the IP, EDA, and ready for production design start, and can you share with us the progress that Synopsys and Intel put together?
Synopsys and Intel. Yes. We, as you know, have had a long history with Intel, many, many years of collaboration. With Intel Foundry in particular, as Synopsys has the privilege with our IP portfolio to be the first on-ramp between our customers and foundry. We start typically with Intel very, very early on at the 0.1 PDK with our DTCO TCAD technology to do calibration of process device modeling, and of course, with IP, that's what our joint customers really look at to see the entitlement around PPA for each process node. With 18A, we started very early, and we have pretty much our foundation IP and our entire I/O Interconnect IP portfolio enabled on 18A.
Wonderful. Thank you so much for this collaboration together. And clearly, on the 18A and down the road, 14A on the horizon, we're partnering with you and the whole team to give customer success. So maybe you can elaborate a little bit how together we can make customers more successful in their design.
Yes. The key when you start working on the next process technology is the speed of iteration to improve both the technology itself and make it easy to use for our customer to ramp up on. Actually, my team handed me, before I walked on stage, a test chip that we were working on with you. And this is on 18A, on 0.3 PDK, about a year ago, so right now, the work we're doing on 14A is no different. We're in that early stage. We call it day zero. We work with you with TCAD and with foundation IP, which is the essential standard cells in order to start our IP development. And the couple of observations, actually, that we're observing with Intel Foundry is not only from a process technology. You have packaging, which is very differentiated. And the last part is Foundry is all about services.
How well can you enable the customer and have an ease of ramp to your technology? And I believe Intel has made dramatic progress in terms of the ability to ramp up on your technology.
Wonderful. And you've been working with us together to serve the customer. And what are you hearing from the customers about Intel Foundry? And where are the areas that we can do better?
Customers are, I want to say, desperate for optionality and for technology that they can bring in. There's huge silicon demand, as you know. And having an option like Intel Foundry, and you mentioned it earlier, in terms of a North America technology foundry, that's something that's very important for our customers. What they're saying initially, what they used to wonder, is the difficulty to ramp up on an Intel technology versus what they're used to with other foundries. And I believe that the gap of design difficulty has absolutely shrunk.
To give you just a perspective, we started designing our IP on Intel technology around 10 nanometers, where the design effort used to be about 2 and 1/2 to 3x more to just design on Intel technology, compared to 18A, which is equivalent to the industry standard, I want to call it, which is a huge progress, which is very important for our customer. You cannot only win with technology. You need to have the whole process of enablement ready in order for the customer to see it as viable.
Wonderful. Sassine, thank you so much for taking the time. And this partnership is very important moving forward. And we count on you to continue the partnership and support with us.
You have our commitment.
Thank you so much.
Thank you. Thank you.
Another key area of focus is the digital design flow and enablement. This is critical to optimize the power, performance, and area, and scaling to new applications and driving the leading-edge nodes of the future. Cadence Design is a leader in this space. They are obviously a company that I know very well. And they have been very strong Intel partners. And I'm looking forward. And joining me to welcome Anirudh to the stage.
Hello, how are you?
How are you, my friend? First of all, I'm so glad to see you. And thank you for taking the time after the earnings call. That's a very busy schedule to even come here. And so delighted to have you here.
Absolutely, Lip-Bu. Great to be here.
Thank you. And we know there are several mega trends happening in our industry, one being the explosion of AI. And I always hear you talking about the general AI to agentic AI and physical AI, and that you really drive that initiative. And clearly, I still remember that you're driving a lot of AI initiatives to all the EDA tools across the board to make it more productive for the customer. And so with that, I think it would be good to kind of share about what are the mutual customers that we can address the demand of our customers.
Absolutely. Well, first of all, thank you for inviting me. Great to be here. Great to work with you again, even more closely, and as you all know, AI is a big mega trend, and there's a lot of talk of AI infrastructure, like data center, but also, I believe that that's a few hundred billion dollars in addition to the silicon industry, but also, the physical AI would be another few hundred billion, so I think the one trillion number may be pessimistic by 2030, driven by AI. And I think for us and Intel, it's great to see the collaboration across the board in digital, in EDA, in IP, also in system tools, packaging. I know Intel has a lot of investment in packaging. And so we're working with all the AI companies, helping them drive their infrastructure, whether it's the hyperscalers or the big leaders like NVIDIA.
Also, we are applying AI to our own tools, like you mentioned, our products, because this is a big opportunity to transform how chip design and system design has been done and to automate a lot of the mundane work. We see a lot of improvements. PPA can improve almost 10%-20%, which sometimes is equivalent to like a node jump. There's a big improvement and also productivity. I think together, working with Intel Foundry, there's a big opportunity to apply AI to the design process.
Wonderful. Anirudh, I'm new to this Intel Foundry. And what do you think the Intel Foundry and how we can your perception and listen to customers, what are the changes and improvements?
First of all, I'm so delighted that Lip-Bu is there at Intel because it's a national asset. And I think Lip-Bu is going to completely transform. And one of the main things, of course, I had the privilege of working with Lip-Bu for so long, is we always talk about team technology and customers. And Intel, and especially Intel Foundry, as you all know, is like a service business. So customer comes first. And I know Lip-Bu has very good instincts to understand what the customer wants. I know you already met a lot of the top customers already. And I think this is something that Intel, through its rich history, now needs to do more, right, listen to customers. I think that's one big change.
Now, in terms of technology, I mean, it's good to see this collaboration on, of course, packaging is going to be a big thing, 3D IC. And I think Intel has very good advanced packaging capabilities. And we are also collaborating more on that. And also, Intel is the first one to go to backside metal. And we see good improvements with backside metal. So in terms of technology, apart from the core transistors, I think packaging and backside metal can be a good thing. And I know you always focus on best product first, right, product wins. And then in terms of team, I know you're already making changes, but also working with the ecosystem. So as one team with Cadence and Intel.
So, I think Intel is a great company but can't do it by themselves, need the partnership of the ecosystem, which I think is even more important for the foundry. So, I'm glad you're doing that too.
Thank you. And in terms of what excites you most about working with Intel Foundry and Intel as a whole company?
So I'm really looking forward to it. I mean, like I said, I think this whole industry is going to be $1 trillion plus. The electronics market is $3 trillion plus. I saw all these reports that by 2035, even the foundry market could be like $500 billion plus. So it's a huge opportunity. So I look forward to working with you and Intel and really making a difference for our customers.
Wonderful. And then first of all, congratulations on the great earnings and continue the great work. And then looking forward to your partnership going forward.
Thank you. Thank you, Lip-Bu.
See you.
Take care.
See you.
OK, bye.
The third building block of successful foundry is design for manufacturability. DFM is driving higher yield and reliability and cost-effective throughout the fab. Siemens EDA is a leader in this space and a great partner to Intel Foundry. Siemens has this digital industrial software. That's really driving the design, engineering, and manufacturing and the life cycle for the industry and embrace the industry. I'm just delighted to welcome the CEO of Siemens EDA, Mike Ellow.
Nice to see you.
Nice to see you. Thank you so much for coming.
Yeah. I was a little jealous that you gave Anirudh a hug before me. So I thought I was a special one, but.
We worked together for how many years together?
Very many years.
Yes.
Through some of the rough times, as I remember.
That's right.
Yeah, the beginning.
Yeah, very much so. Yeah. Thank you. So taking your busy time to come over here.
I appreciate it.
And it's a very important partners for us. And then clearly, from the Siemens, I always admire and Roland and all my good friends. And then you guys building that whole system approach to the whole industry. And one thing that I'm really excited is how together we can empower the customer. And clearly, on the manufacturing side, Calibre is the industry standard that we all know. And then can you share with us our collaboration progress we make together?
Sure. So Lip-Bu Tan and the audience, as you know, Intel Foundry and Siemens EDA have been working together on certification of our technologies and physical verification and simulation for years. We're currently certified on Intel Foundry 18A. We're doing performance tuning on Intel Foundry 18A-P. And we started the process on Intel Foundry 14A. So lots of work going on in that space. But what also is interesting is that we've had an ongoing partnership for almost 15 years now with the Advanced Packaging team. So there's a lot of innovation going on with technology and methodology associated with where this packaging going to the future. And then lastly, we're proud to be one of the founding partners as part of the Intel Foundry Accelerator Chiplet Alliance. Yes, I knew I'd get that right. Yes, that's a mouthful.
But what I find interesting about this thing is, as you take a look at expectations of fabless customers, right, they expect great process technology. But on top of that, what they're expecting is predictability of cost, yield, and schedule through the manufacturing process, right? Power, performance, area, well, that's on the design teams themselves. But the fabs really got to give that predictability. And that's where Siemens EDA comes in with our industry-leading Calibre platform, because it's a trusted sign-off between the fabless companies and the foundry in order to make sure that the translation of the manufacturing process requirements goes into design.
And then on top of it, when you take a look at the connectivity that Calibre provides with all the fabless companies and the silicon IP providers, it really ties together an ecosystem so that everyone can stay in their natural design environments using methodologies, but still come into the foundry with predictability. So for me, when I take a look at where Intel Foundry and Siemens EDA are, we're really putting together a trusted, customer-centric solution. It's really good stuff.
Wonderful. Just dip a little bit, dive a little bit closer, deeper to that. On your perspective, we clearly mentioned about the advanced packaging, our EMIB and all the technology we have. How do we work together to really help our customers?
Yeah. So to me, EMIB is a fascinating technology platform with the promise that it offers with lower latency, lower power, and lower cost. I mean, it's got a lot of the lowers associated with it, and it's especially well-suited for where the world is heading in AI, and when you really contemplate what's going on with our customers and our customers' customers, they're undergoing a digital transformation where software now is the primary differentiation, right, so the idea of something that is software-defined and silicon-enabled is really taking hold within our industries, and with that, the days of a hardware team handing off a silicon platform to software groups and software's making compromises to make the system work are over. The hardware platform has to be customized to the software workload it is driving.
And to me, that's where advanced packaging and 3D IC came in, because now they allowed a discrete combination of the hardware assets to map the workload that they're driving. And so that's where Intel Foundry comes in with a great technology platform, a world-class advanced packaging team. As I said earlier, we've been working with them for more than 15 years as far as partnership doing the delivery of technology and methodologies. And so as I take a look at an enablement point for this digital transformation, Intel Foundry has a wonderful opportunity for how they can change the world.
So thank you. Siemens, I always admire, is a system company and then enable the design from the EDA to manufacturing and then for the industrial applications. And Intel, we try to be the system foundry. And so there's a lot of area that we can collaborate and learn from each other, especially in the system know-how that you guys have in terms of industrial automation, power management. And so tell me a little bit, how can we change the world together and then driving this whole system approach to the design?
Yeah. So it's an interesting question, because what I find exciting about our opportunity together is the world is becoming more and more dependent on silicon-enabled devices and systems and system of systems. And as you take a look at the access to silicon becoming a national priority for many nations across the world, they're worried about the access to silicon because it could be a society disruption or a disruption to the very way of life. And when we start imagining the almost limitless possibilities with AI and how it can impact our lives together going into the future, it does create a challenge with the number of new designs that have to get started in order to address all of this. And then we end up with what's going on with access to silicon. Universities aren't graduating enough engineers in the engineering that we have. Well, they're retiring.
So we're faced with a challenge and an opportunity and dilemma all at the same time, is how do we take earlier in career engineers and allow them to address the multitude of new designs that they have to deliver and have silicon for it? And so when you contemplate what's going on in the world, the world is really dependent upon a resilient, robust, distributed, advanced node silicon supply chain. On top of that, the world really needs a set of AI-infused technologies to connect together the broader ecosystem in order to allow all of the design content to be created. So this is where we come in together, right? So the world, in my opinion, needs Intel Foundry to be successful with the delivery of advanced node technology and be able to manufacture it.
The world needs Siemens EDA and, as a matter of fact, all of EDA to deliver these AI-infused technologies to be workforce multipliers. So when you bring it all together, it's really one of those pivotal events when you take a look at the world depends on success from Intel Foundry and from Siemens EDA together. So I mean, I think it's a wonderful, wonderful opportunity for all of us.
Fantastic. I'm looking forward to a more deeper collaboration together.
I am too.
Thank you so much.
Oh, thank you.
Thank you.
Nice to see you.
A final area that I want to highlight is the importance of design for yield. Data, machine learning continue to unlock the new opportunity to improve yield and acceleration time to market. PDF Solutions is a clear leader in this space. And I really want to welcome the CEO of PDF Solutions, John Kibarian, to the stage. Thank you really so much for joining here.
Thank you.
And besides, we do hiking. And we also have CMU Technology Advisory Board together. And your wife actually is on the board with me at the CMU. So we have a lot of connection there.
Absolutely.
Then clearly, there's Intel and PDF Solutions working together to ensure the design for yield and manufacturability to build the process from the beginning. And that's very critical. So can you elaborate how we work together and then to help each other here?
Sure, and first of all, thank you very much for having us here and for the partnership that we've had with Intel and also in the partnership you and I have had over the years. Intel has been the leading IDM for decades, and when it decided to make advanced process nodes available to its outside customers, to its system foundry customers, it recognized that in order to be able to yield ramp fast, you need to really characterize the way processes and designs come together differently, differently for the foundry market. So about four years ago, it began using PDF Solutions in collaboration with its internal systems to build PDKs, to build data sets that allow the engineers and R&D to interact more closely with the design teams and the eventual production teams to achieve best performance and variability.
Below seven nanometers to maximize performance, profitability, and yields, you really got to co-optimize. And that co-optimization can't begin after the fact. The R&D teams need to be working with the product portfolios in mind from the beginning. And to do that, they've really got to start working much more early with the design information in manufacturing. We felt. We look at Siemens. It's bringing manufacturing data into design. We have to bring design information into manufacturing so manufacturers know how to take into account the products that they may not see yet, particularly as Intel moves to having more and more external customers. So we initiated a way of working with design information brought into the manufacturing flow, enable e-beam inspection, enable new ways of looking at characterization data so the R&D teams can respond to the future problems and address them early on.
So the technology is more ready, not just for Intel's internal products, but for the world's products.
Yeah. Thank you so much for sharing that. Clearly, yield is very important and how early we should engage with customers to helping them. And then the other part is from the Intel Foundry. I always consult you about how do we improve our yield for the different process nodes. And as you mentioned earlier, it's very important when you design, you want to make sure that the temperature, power are being addressed and overprovisioned in some ways. So especially in automotive and also in data center, this is critical. And then I think PDF Solutions is a very unique position to help customers to do that. Can you elaborate on that? How can we help from Intel Foundry and you, PDF Solutions, helping the customer to address some of this design to manufacturing opportunity?
Sure. First, let's talk about collaboration. Our industry delivers the most integrated products in the world, yet we operate in a disaggregated fashion. We all have to collaborate, and engineers in our industry, as you always say, have to be humble because the problems we take on are unbelievably daunting, and so to do that, they need to work hard together with their customers and their suppliers. The suppliers play a very important role. In any company, it's always critical, so collaboration, teamwork, and integrated analytics are required. We at PDF Solutions try to be the catalyst to enable Intel to work more efficiently with its suppliers through our secure communication networks, our data exchange networks, through our analytics with your customers.
The problems that need to get taken on is really to understand that trade-off as early as possible between the technology choices, the cost that those technology choices will make because we always have to sell products at a good price, and the performance that the customers want to get in order to be able to even go to the next nodes. So that analytics really is a shared team sport. And we try to enable Intel to work more collaboratively with its customers, with its suppliers. And use analytics. I always say electrical data is the lingua franca in our industry. R&D engineers think about cross sections and top views and TEMs, and designers think about their timing closure. But the lingua franca is the electrical characterization that brings together the different engineering teams. And that's our role to help you.
Wonderful. Thank you. In terms of the 14A and beyond, I mean, clearly, geometry is much smaller. The complexity increased a lot. And then the power is going to be an issue in the area. And then these are the areas that you're expert in that. And then can you share with us some of the customers clearly when they design? Really, these are very big challenges to figure out what is the next step to address and how early should they engage? Any advice from that?
Sure. I mean, I think always in these cases, innovation drives benefit. But innovation needs to, you need to know how to use the innovation. And when new architectures come out, it usually takes you a little while just to make your code work well on it. Well, the same thing is when you introduce things like backside power. The innovation that Intel has really brought forward with 18A and extending with 14A and in advanced packaging as well, it's a lot of capabilities that designers need to know how to take advantage of. To do that, they need to engage early. That's super critical. And for that to be possible, Intel has made a lot of investments early on. So the technologies, as the previous speaker said, the PDKs are easy to work with.
The variability in characterization data is there so engineers understand how to design to that process capability, and at advanced nodes, when you look at things like 18A and 14A, they are designed with advanced packaging in mind. It's built into the PDK. It's assuming you're going to put this into an advanced package. It's not a standalone die, well, that means that you have to start thinking about your test flows, multiple test insertions. Most of the engineering teams we see are using AI to manage that complexity. That means more data feed forward and more data analytics throughout that test flow in an area where PDF Solutions is contributing and where we're partnering with Intel because of the data sets that Intel are building on our systems that make that all AI is based on data.
And those data sets are going to enable Intel's customers to not only look at how to get the part taped in and initial yield, but how to run that production flow efficiently because it's a very complex test flow they're all going to take on.
Wonderful. Well, I'm looking forward to more deeper collaboration and partnership with you. And thank you so much for your busy schedule to come over. Really appreciate it.
Thank you.
Thank you. Thank you. As you can tell, this is important. I really embrace the industry standard, IP, and EDAs so that we can really help you and serve you better. And this is truly a service business. And that is built on the foundational principle of trust. And you have to be patient to earn your trust. And then we have to really drive some of the technology to make sure that the yield and performance and reliability meet your requirement. So one of my top priorities is to make it easier for the ecosystem to do business with Intel. This event is a great opportunity for us to discuss the best way to do that. And then as the name of the event say that this is a direct connecting. So highly recommend to connecting, networking with all of us. We'd like to hear from you, the feedback.
We have a lot of different demos and then some of the speeches. So feel free to reach out to us. We'd like to hear from you. If you have any good ideas for us to be even better to serve you, I think that would be great. We will be guided by the voice of the customer, all of you, and will act on your feedback. We are humble in that approach, as you can tell, because we know your success matters the most than anything else. We work every day to earn your trust. I'm grateful for all of your participation and partnership with us. I'm excited about the work ahead. Our team will be here to engage with you. I think at the end of the day, the best testimony is you.
And then how we can be serving you that you can praise us or can give more business to us, I think that would be great. And I would thank you, all of you, for being here. And I hope you have a great time today and then get something out of it. Most important, network together with you. And thank you so much. Have a wonderful day.
Please welcome Executive Vice President and GM with Intel Foundry, Dr. Naga Chandrasekaran.
Good morning. Thank you to all for joining us here today. It's been eight months since I joined Intel, and prior to that, for 23 years, I was at Micron Technology heading their technology development team. This is a proud moment for me to be here as part of this iconic company at a time of significance, not only for Intel, but also for the entire semiconductor industry and for our nation. We are in the midst of a significant transformation, as you all know, and that transformation is built on the foundation that was laid by several leaders ahead of us. I want to thank one such leader, Dr. Ann Kelleher, who has served Intel for over three decades and recently announced her retirement. Thank you, Ann. I also want to thank several leaders and technical experts at Intel who have brought us to this point in our technology.
It is by standing on their shoulders that we are able to see further. Today, I want to share some of our vision towards the future as to how we will become a successful foundry. As Lip-Bu highlighted, in order for us to be a successful foundry, we have to earn the trust of our customers. And foundation to that are a few elements. The first piece is technology. We have to have innovative technology that can deliver value to our customers. And along with technology, we need to provide a global resilient supply chain. The second aspect, as Lip-Bu greatly highlighted in his opening talk, is our partnership with our ecosystem providers as well as the broader supplier base in the industry. We have to collaborate, and that's the only way we can win. And the third piece is to be predictable. Predictability is key to our customer's success.
To earn their trust, we have to be predictable. Today, I would like to share a few thoughts on how we are making this transformation journey possible across Intel to earn our customers' trust with the customer-first mindset. First, let me start with the first element of that Venn diagram, which is on technology and global capacity. In the last four years, Intel has invested almost $90 billion in capital. Broadly speaking, this investment was in two areas. As the leading R&D expert and provider in the U.S., almost 20% of this $90 billion was invested to shore up our competitiveness in R&D, both in front-end as well as back-end technology. The remaining 80% was spent primarily in expanding our global footprint.
We have continued to invest in space as well as technology transitions that can be provided to our customers and make Intel competitive as a foundry provider. On the technology front, this chart shows what we have done in the last four years. Everyone knows that Intel fell behind with our 10-nanometer technology. Our first effort in the last four years was to shore up our technology competitiveness with Intel 10. This technology is now in high-volume manufacturing. Not only for Intel 10, but we have continued to explore progressive ways to monetize our 14-nm as well as Intel 16. We are working with foundry customers to enable success with Intel 16. Later, my colleague Kevin will talk more about it. We have partnered with UMC to bring their 12-nm technology to Intel.
This joint partnership is helping Intel learn about foundry and how to provide this 12-nanometer technology to our foundry customers. Our next focus was to transition to EUV. We primarily focused on getting Intel 4 developed and transferred into our Ireland factory. Today, we are continuing to develop this node as a foundry-capable node. We have enabled TSV technology on Intel 3, which would be a base die for many of the AI solutions. We are continuing to make improvements in performance with Intel 3. We are offering more voltage options as well as thick gate oxide options so that the performance can continue to improve on this node. In the last two years, our R&D team has shifted focus to 18A. 18A is facing ups and downs. With any new technology, we can have these puts and takes. But the team has continued to make progress.
I'll share more details on 18A. I do want to share that today, we have approved risk start manufacturing with 18A. We are looking forward to serve this node to our customers. We have also made significant investments in our advanced packaging roadmap. I'll share more on that later. It was not just on the roadmap, but it was also with the global supply chain. Lip-Bu highlighted this earlier, where we have invested in our footprint for front-end across North America and also in Europe, in Ireland, as well as in Israel. In these front-end facilities, we have built new factory space and also enabled the transition to EUV technology as our roadmap has progressed. We have a strong presence with our assembly test manufacturing footprint in Asia.
And more importantly, in the recent years, we have invested heavily in advanced packaging technology as well as capacity. We now have advanced packaging capabilities in New Mexico, in Oregon, and also in our Asia factories. I want to take this opportunity to take you all on a journey and show some of our sites to you through a few videos. First, let us go to Israel. Intel 7 node is fully ramped in our Israel factory. We are 100% utilized in our non-EUV factory space, and we have achieved mature yield on this technology node. But we are not stopping there. I'm very proud of the Israel team for what they are doing. They are continuing to increase the performance bins and also bringing a continuous improvement mindset in delivering cost improvement activities as well as yield improvement. A big part of Israel's deliverable is predictability.
That's one of the key cultural changes that we are driving. And I'm proud to say that so far, even with all the challenges facing them, the Israel team has never missed a single milestone. And that's something to be very proud of. We are also utilizing this team to contribute to all of our other nodes across the factories. In addition, we are enabling the Israel team to support us as an AI center of excellence to drive industrial automation across Intel factories. So thank you to the Israel team. Now, from Israel, let's go to Ireland, where we are ramping our Intel 4 technology node.
Our foundry journey has been really exciting for Intel. And Intel Ireland is proud to have run the very first foundry customer wafers.
We're working hard to transform our systems, our job roles, and our skills to ensure we can perform an outstanding job for you, our foundry customers.
I want to thank Intel's customers for helping us grow into a service organization and look forward to welcoming many more customers as we grow Intel Foundry.
We have learned a huge amount about customer expectations and how we need to serve you. And based on this, we're making major transformations in our business, all of our systems, and our culture. Intel 4 is Intel's first EUV node. And I would also note here that it's the leading-edge node that's ramped in Europe. And we have learned a lot as part of this node. I want to highlight a few cultural transformations that we have made. In the last couple of quarters, we pulled in transfer of this node to our manufacturing facility.
In the past, Intel would have waited longer for reaching maturity, and we have now started moving this end-to-end and getting the nodes earlier into our manufacturing facility, and by moving it earlier, what we have done now is enabled our manufacturing team to start looking at continuous improvement on yield, reliability, and focus on delivering it to the customers. One of the key aspects of delivering to customers is quality, and in the past, Intel's focus was on using test methodologies to bin the die and sell it to our customers, but now, as a foundry, we have to shift left. We have to have more quality focus in line, and moving Intel 3 earlier into our manufacturing facility highlighted how we have to invest more into inline characterization, into advanced capabilities that can detect faults earlier and have a velocity in terms of yield improvement.
The team is able to accelerate this learning and show that we can actually do development in manufacturing. One part of that is to change from the copy-exact philosophy to make sure we are innovating at every part of Intel. Now we want to make sure that the innovation engine can deliver solutions faster to our customers. We are also running Intel 16 in Ireland. This node has taught us a lot as well. We are partnering with our first foundry customer, MediaTek. With their support, we are able to make significant learning advances in how Intel behaves as a foundry.
We want to thank our customer for teaching us on variation control, on how are we going to improve in delivery to the customers, how do we respond to the customers fast, and more importantly, collaborating with our customers by putting them at the center of success. And these learnings would not have been possible if we did not have a customer working hand in hand in our factories with our team members. And I want to thank MediaTek for teaching us how to be a customer-centric mindset. The third video that I wanted to share now is bringing us close to home. Let's go and look at our Arizona site.
Here at Intel Arizona, we're excited and ready for Intel 18A, the most advanced process node developed and ramping in high volume here in the United States.
Our fab and our team are ready to serve our customers here on this innovative technology node. Intel 18A is the most advanced node that's developed and ramped here in the United States, and this node is industry-first on several fronts. It's the first gate-all-around technology that will be ramped in high-volume manufacturing. It's the first node to have backside power and offerings to our customers, and we are very delighted to have this node achieve the risk start manufacturing milestone. Currently, we are running this node in Oregon, and we have five products that are already running in production. We have delivered our engineering samples to the customers. We are in the process of running our qualification samples and continue to make improvements with our 18A technology node. We are partnering with our ecosystem partners to offer this 18A technology solution to all of our customers.
In Arizona, we also have the UMC joint-developed 12-nanometer technology that we are working on to port over into the Intel factories. This will also be a foundry-enabled node that can be offered to our customers. I want to share some progress on 18A that we have achieved. If I can go back one slide, please. This first slide shows our defect density tracking on 18A. We have achieved over 15% performance improvement, and we are on track to deliver to our performance commitments. There's a 30% chip density improvement on 18A, and our defect density is tracking to our plan by which we will be able to ramp it in production towards the end of 2025, towards the latter part of the second half. Our best wafers have already achieved the defect density targets.
We have forward-looking vintages that's now running in the factory that gives us confidence that we will hit this milestone. We have also heard feedback from our customers that 18A is not targeted towards the broader market. It was designed with high-performance compute in mind. We are now proud to have the next revision of this, which is the 18A-P that will be offered to our customers. 18A-P has design rule compatibility to 18A. We are working on improving the performance by another 8% on 18A-P offerings. It'll have additional ribbon size. It'll have corner tightening. It'll have additional VT ranges available for our customers. This would be targeted for a broader ecosystem for us to deliver. We are making good progress on both 18A and 18A-P.
This slide shows on the top right for me, it shows the performance of our ring oscillators, the frequency leakage curve, and today, for the wafers that we are running in risk production mode, we are achieving around 90%-95% of our performance target, and the bottom graph shows our current forward-looking vintages that are targeted for production start in the second half of 2025, and you can see that the ring oscillator performances are actually hitting the target, and we are on track to continue to improve the performance. In addition, all of our SRAM Vmin , both for the high-density and the high-current cells, are achieving our targets, and we are continuing to make progress in these areas so 18A, while having its ups and downs, we are at a point today where we feel confident with the progress we are making.
We are humbled that we can still have challenges with any new technology nodes. We are paranoid. But we are making good progress and confident that towards the second half of 2025, we will be able to serve our customers with volume production on this node. If I look beyond our 18A in terms of our future-looking roadmap, I will start with some of the mature nodes where we want to continue to partner with UMC and deliver additional offerings in the future. On Intel 3, there are some ease-of-design use challenges because of how the node was developed. But we want to partner with customers who have interest in Intel 3 to develop further offerings on this node. We are working closely with the memory customers to see how do we enable TSV on Intel 3 that can be used for high-bandwidth memory type solutions.
We are also interested in offering additional capabilities that can be enabled on this node as a foundry node in the future, and we are willing to work with our customers to enable this node. On 18A, I already mentioned about 18A-P. In addition, we'll have a TSV version on 18A as well, and the 18A-P TSV will enable it as a base die. This is going to provide this node a lot longer of longevity because in the AI applications where we can provide system solution with 14A and 18A-P being the base die and offer this to our customers, so we see the possibilities of 18A not just for what it can offer immediately, but from a longer term, 18A-P and 18A-PT will continue to provide advantages for our customers. We have now started focusing a lot on delivering our 14A technology node.
We are running this in Oregon today. Our progress is good. We are taking a lot of learnings from our 18A and moving it to 14A so that we can deliver this node predictably. 14A is expected to provide 15%-20% improvement in performance per watt. It will also offer scaling benefits of almost 30% improvement in chip density. We expect a 25% power reduction and maybe even more as we continue to optimize this node. One of the important things with 14A is that this is a second-generation RibbonFET and second-generation backside power. A lot of the challenges that we faced on 18A and the lessons learned on 18A will port over. That is what is giving us confidence with our roadmap with 14A. In addition, we are moving to a power-direct solution architecture for 14A from the PowerVia on 18A.
And that will provide additional performance benefits. We have started working with the ecosystem partners much earlier. Like the previous speakers mentioned, we have started engaging with the ecosystem partners on 14A much earlier. We are also making sure our silicon data that's supporting the PDKs is healthier and more industry standard based on the learning that we have gotten from Cadence and Synopsys and Siemens. And in partnership, we are making sure we have a better solution. We are working with PDF Solutions to have their test chips running in our factories to ensure we are collecting the right data that can provide the collaterals to our ecosystem partners. We are also partnering with PDF Solutions to see how we can use their solutions in our manufacturing facilities that can now enable yield ramp and manufacturability faster.
As we move away from copy-exact and as we democratize innovation across the Intel factories, we need Intel factories to work on variability reduction, cost improvement, yield improvement, and reliability improvement. And that requires us to partner very closely with our ecosystem partners. One of the differentiating factors on 14A is High-NA EUV tool. In Oregon, we have landed our second High-NA EUV tool. In fact, this tool came up much faster compared to our first one. And we are working in great collaboration with ASML. The tool is performing well to its expectations. I want to make two points for the audience today. The first one is Intel still has the option to have either a Low-NA or a High-NA solution on our 14A technology. And it's design rule compatible. There will be no impact to the customers depending on the path that we choose.
Second, the High-NA EUV is performing to the expectations, and we will introduce it at the right time. We already have data on 18A as well as 14A that shows yield parity between our low-NA-based solution and a High-NA-based solution, so we are continuing to make progress on the technology front and ensuring that we have the right options available for us to make sure the solution we deliver to our customers has the lowest risk and the best reward in terms of the decisions we make. Our investments are not only in the front end but also in the back end area, which is a significant differentiating factor for Intel, and we expect to offer these solutions to our customers. I want to show a small video clip of our construction and space expansion in the back end area.
The video that you just saw showed some of the clips of construction work that's happening in Malaysia as well as in our New Mexico facility. We are happy to host you in our New Mexico facility where you will see the campus as a vibrant hub of the most advanced packaging solutions in the United States that we have to offer for our customers. It has end-to-end capability. We run our silicon interposer solutions in this facility. We have EMIB technology development that's happening in our New Mexico facility. We also have die sort, die prep capabilities, test capabilities, as well as the full advanced packaging solutions for Foveros. These solutions collectively give us the end-to-end technology development to manufacturing solution in our campus. In Malaysia, the facility that we have built is the first of its kind in Malaysia.
It's the first one that will have a full silicon wafer capability in a back-end manufacturing facility. It's not available in any of the other factories in Malaysia. And we have timed it so that in future, as our packaging strength grows and we have more customers, we will be able to ramp into this facility and again be able to provide our assembly test manufacturing solution as well as the advanced packaging solution and the backside metallization solutions in a single factory. Our roadmap today continues with our Flip Chip BGA-based solutions where we are making continued progress in terms of pitch scaling and providing better interconnect solutions and providing better package solutions to our customers. We have the silicon interposer-based Foveros package solution that we now offer for all of the AI PC-related applications. And we are engaged with many of our customers in this area.
Intel's unique EMIB technology solution continues to make progress. We have made several EMIB-based products, and we are continuing to work on reliability improvements and putting additional products into the field. This year, we want to announce some additions to our roadmap. On the EMIB front, we are introducing a TSV capability with our EMIB technology. This TSV capability and the EMIB solution will provide improved power delivery and signaling capabilities for HBM4 and UCIe 32-based solutions. We can also have better AC and DC noise with this EMIB-T-based solutions, and we can have MIM capabilities on the silicon as well. We have already shown that we can have EMIB on panels, so this EMIB-T-based solution will be able to scale fast for the bigger complex solutions that the customers want to build with more silicon area and an increasing number of HBM on the package-level solutions.
By having EMIB-T-based solutions, we can also test the SoCs earlier before we have to commit to putting more of the HBM-based solutions on the package. And this could be a unique advantage that Intel can offer. On the Foveros side, we heard all of your feedback. And in order to continue to improve on the cost, we are introducing two new solutions, the Foveros-B and the Foveros-R-based solutions. And these Foveros solutions will enable us to provide a substrate-based buildup capability that can be used by our customers. And beyond Foveros, Intel is introducing the 3D IC-based solution where we have hybrid bonding capabilities. And we are now working on products that will be available for end customers.
In addition to space, we are continuing to invest in our technology capability and ensuring that we can offer these innovative solutions to our customers both on the front end as well as the back end side. This slide shows our strategy on the optical interconnects and the co-packaged optics. Optical interconnect offers significant advantages over the traditional electrical IO. It offers benefits in terms of shoreline density improvements. It offers benefits with bandwidth, latency, and power efficiency. When we can take optical interconnect and bring it to a chip-to-chip level along with Intel's advanced packaging capabilities, this solution is going to provide significant benefits in how we can scale up and scale out AI-based solutions. It will provide more denser and advanced interconnect capabilities. Also, in the compute space, we can provide latency and higher throughput by having the co-packaged optics solution.
So Intel is very happy to work with our customers to provide our advanced packaging solutions to enable your co-packaged optics-based desires and needs. I want to bring us back on this whole technology and capacity aspect and indicate how we are doing towards our future. I want to say that the slide is more of a representation and not to an exact scale. And it might change depending on the technologies that we run in the factories. What this slide shows is on the solid bars, what's our front-end capacity that we are running today and already planned to deploy based on the investments that we have made in the last four years. It shows both our EUV as well as our non-EUV capacity plan. But by working closely with you, we can actually enable additional capacity. We have invested in space that's available at different stages.
If our customers come and ask from us, within six to eight quarters, we will be able to take these spaces, build them up, and actually provide the tooling in order to get extra capacity for our customers. This is across Arizona, Oregon, Israel, Ireland, as well as the future buildings that we are continuing to place in Ohio. A similar picture for the back end. We have traditional assembly test capability worldwide, predominantly in the Asian factories. We also have the advanced packaging capability to offer. Between Foveros, Foveros Direct, as well as the EMIB and EMIB-T capacity, Intel has one of the highest capacity offerings that we can provide to our customers in advanced packaging.
So like Lip-Bu highlighted, it's not just the technology, but also continue to provide a global supply chain that our customers can depend on and come to Intel for both technology and capacity. The next pillar that I wanted to mention briefly is on the ecosystem partnership. I want to thank all of our ecosystem partners for not only working with us, but also teaching us in how to become a better foundry and how to listen to the voice of the customer and make sure our customers can have the right solutions. On the front-end side, we have enabled five strong vendors for our PDKs. And we are continuing to work with them. On 14A, we have one module that's not ready as of today. And we expect it to be completed in the next quarter. From 18A to 18A-P, as I mentioned, we will have design rule flexibility.
As a result, most of the IP should port over seamlessly between the two technology nodes. Also, on the IP portfolio for 18A-P, we expect to release these between Q3 of 2025 to early 2026, and we have strong partnership with the EDA vendors to ensure this stays on track. Intel also offers package assembly design kit, and these design kits enable our customers to evaluate EMIB-based package designs and also validate them. The design kits will have design flows. They have guiding principles and rules, and they also offer all the requirements that our customers need in order to design their EMIB-based solutions and validate them. As you can see, we have several modules already available, and we are working with our customers in providing these solutions to them so that they can evaluate our EMIB-based technology for their use. I want to close with the last section here.
We talked about technology and our global capacity. We talked about ecosystem partnership. But the most important thing is we have to earn our customers' trust. And we have to change our culture to be customer-first mindset. And it's easier said than done. It is going to take time. But we are making several progress in this area. There are four pillars along which we are focused on where we are putting customer-first mindset. The first is predictability. We have to be predictable to our customers. And we are not there today. We have to earn customer trust by delivering and executing stronger and being predictable to our customers. The second pillar is quality. Intel has to make significant improvements in ensuring the customers have no impact, negative impact from our quality-based deviations. We have to change from a test-based quality detection back to an inline quality detection.
We are investing heavily in this area, not only through test chips, but by partnering with our equipment suppliers to install all the inline detection capabilities. The third aspect is velocity. We have to not only move fast, but we have to move in the right direction. This is where we need partnership with our customers and ecosystem partners. Velocity is going to be a key aspect of how Intel delivers to our customers so that they can feel they can get to market faster. Time to market is going to be a key criteria for our success. The last one is on affordability. We have to do all of this while we are prudent with our spending. I showed the CapEx spending that we are making, that we made in the past of $90 billion.
That is giving us a pretty good baseline to start with both from a technology and a space capability. As we look at the future, our CapEx investments immediately will be to ensure our technology competitiveness stays on track. We are able to transition our factories to meet the customer demand and then decide how we are going to continue to expand and invest into capital. We are bringing a lot more efficiency-based activities into our factories. How do we utilize our capital better? I've used the word no CapEx left behind, ensuring we get more wafer starts from the installed capital, ensuring we are getting more wafers from our space that we have already constructed. We are driving a lot of these improvements across our factory, but with the customer-first mindset in terms of how we are going to serve our customers.
We are going to do this across four vectors: people, process, systems, and structure. We are ensuring that we get outside-in views across the company, not just at the top level, but at every level of the organization. How do we get outside-in view? A big part of it is our customer feedback as well as our ecosystem partner feedback. We are ensuring that our systems and structure are changing to the new reality of serving as a foundry. We are making sure unnecessary business processes are eliminated. We streamline our activities with velocity in mind. The change from copy exact to continuous improvement mindset. The change from telling others to listening from our customers and our suppliers. The change from measuring our success by what Intel delivers to measuring our success as what the customers see as Intel delivering to them.
These are changes that we are making to ensure we have a customer-first mindset in what we deliver. On the affordability side, we are driving all of the AI and automation solutions into our factories. I want to show a last video here on how we are using automation-based solutions within our factories. Our automation team is driving these improvements in our factories. To share more about it, I want to invite Joe and Jessica on stage, please. Joe and Jessica. Joe, what do you have in store for us today? I want you to meet one of our newest team members. Please welcome to the stage Chip, our robotic inspector. Come on, Chip. Come on up the stage. Why don't you go ahead and give us a bow? What you saw on today's video just a moment ago is that we're empowering our technicians.
Chip's facilitating this by allowing us to collect more consistent data. In just a moment, we're going to demonstrate Chip's thermal imaging capability. If you look across the stage, we have an actual motor from one of our facilities. We heated it up backstage with a heat lamp because we wanted to simulate what a bad motor winding would look like. In just a moment, Jessica will dial in the camera. We'll be seeing the heat signature of this motor on stage. Oh, wow. There it is. It actually held the heat pretty good. It's at about 189 degrees. Definitely something wrong. That's hot. Definitely something wrong. Keep in mind, all of this data back at our facility would be streaming live to our advanced processing control system. It's an AI analytic platform that monitors this data for us.
In this scenario, it would have auto-generated a maintenance ticket. We would have had a technician go out and address this issue before it became a real problem for our factory. That's amazing. This is an example of what I mean by democratizing innovation. We always think of innovation as the new next Chip. Walking here is very interesting. Yeah, I'm more focused on Chip than what I need to say. We always think of innovation as the next generation transistor. What are we doing with CMOS or new materials? Innovation happens in every part of our factory and in every part of our business.
This is one such example where we partnered with Boston Dynamics to ensure that we have the right scalability, the software stack has the right applications, and bringing it and working together to ensure the right solutions are put in place inside our factory. So Joe, in addition to this, what can Chip do for improving efficiency in our factories?
So the way I see it today, data is like gold. And Chip has allowed us to expand the amount of data we collect from our assets, allowing us to understand the health of our assets and fix them before they become failure.
Very nice. Very nice, Joe. Jessica, in addition to thermal imaging, what else can we expect Chip to do that can provide efficiency improvements for us?
Yeah. So Chip enables our workforce to be more productive.
We're able to scale our productivity by dedicating our skilled resources exactly where they're needed the most. So we have a full robotic fleet of robots just like Chip around the globe in all of our facilities. And they are helping to make work in our facilities smarter, safer, and more efficient.
Excellent. Thank you very much, Joe and Jessica. Please give a round of applause for this automation team.
I was just standing here to ensure he doesn't ram into me. I do have one last thing that I wanted to share and show our commitment. I'm a fab person, so it always feels strange to hold a wafer by hand, always cringe a little bit. But this is our leading-edge 14A node technology wafer. And this is running in our Oregon factory. And we continue to make progress with this technology.
What I want to make sure we say today is we are committed. We are committed to foundry. We are committed to our customers. And we are committed to working closely with our ecosystem partners to make Intel Foundry successful for the customers. It's not just important for Intel. It's important for all the employees that work in Intel, all the communities where Intel operates, all the hundreds of thousands of individuals who support Intel. And more importantly, it's important for the semiconductor industry to ensure that our innovation progresses and our solutions for the customers continue to progress. And we commit to listen. We commit to be humble. And we commit to listen from you and deliver solutions to you. So thank you very much for joining today. Thanks for all your support.
Please welcome Senior Vice President and GM with Intel Foundry, Kevin O'Buckley.
Hey, Thank you all very much for being here. That was really cool. Naga gets all the toys. It's amazing what $90 billion will buy you. Anybody else see Black Mirror? I did too. Same reaction to the robot. My goal today is to connect some dots for all of us as a team. You're going to hear a lot. You heard a lot this morning, and you'll hear a lot this afternoon that I would characterize as the what. You'll hear about whether it's angstroms or millivolts or silicon bridges. My goal is to talk a little bit more together about the why so that everything you hear today, I want you to hear intentionally done through the lens of what our customers are asking for. So that's our goal together.
You heard from Lip-Bu this morning about the importance of strengthening our roadmap, collaborating with a broader ecosystem, and making our customers successful. You heard directly from a number of our most important ecosystem partners about how they're investing to enable our capability for our customers. And you heard directly from Naga about how we're deploying our resources for development, capacity, and our capital plans. I'll focus this morning on how we're leveraging our rapidly growing relationship with customers to expand what we're doing with three key priorities. The first of those priorities is better matching our technology to their diverse needs, attributes of the technology, not just in the details of the technical details, but also the schedule of what we're delivering for our technology and the predictability.
Second, we're going to talk about how we're transforming our business and technical processes to improve our execution, our responsiveness to our customers, and again, the predictability of what we're doing for our customers, and third, we've talked a lot today about ecosystem. I'm going to talk about how we're making some new investments specifically to expand our ecosystem even more based on the feedback we're getting from our customers. All three of these priorities for us are about how we are intentionally transforming Intel Foundry into a service-first company founded on trust. The rate of pace and change that we see in what we call together the AI era really is unprecedented. Every company, every company in our industry has somehow magically, over the last couple of years, become an AI company. We at Intel Foundry are focused on becoming an AI services company.
Our goal is to enable our customers' vision, and that vision is reasonably well represented here, vision of the next generation products. This is what our customers are telling us they need us to deliver for them. It's a 3D construction with multiple fully utilized base reticle die with many compute tiles stacked on top. It's surrounded by an extraordinary massive memory capacity, both the highest performing HBM and also, in this case, LPDDR die. And these are all coherently connected by both electrical and optical interfaces, tens of terabits of bandwidth, and what you're looking at here is in excess of a 12x reticle size solution. At Intel Foundry, we believe, based on what our customers are telling us, that we are uniquely positioned across the entire industry to deliver these massive heterogeneous systems on a package.
Keep this image in the back of your head as we talk today because it is this vision, your vision, that informs the roadmap that we're developing together. We all know today's systems aren't just silicon. But a lot of the vision and the drivers of our systems do start with silicon. So that's where I'm going to start. Naga and team reviewed this. Over the past year, I've engaged with many, many customers. And all of our announcements that we've made today, this morning, and this afternoon are incorporating their feedback. The top of this chart is what we all euphemistically call the Angstrom Era.
For us at Intel, it is comprehensively enabled with both gate- all- around and backside power delivery, our evolving 18A platform, and our 14A platforms on a predictable two-year cycle because that's what our customers are telling us they need to bring their products to market. We'll share some highlights of our roadmap. But I do want to start with 18A. I want to start with 18A, not just because it's an incredibly important first technology for us. But when you hear about 18A as a technology today, team, I want you to also think of it as the foundation of all of the nodes we're building in the future for the coming few years.
In the relentless pursuit of next generation computing, Intel Foundry introduces the Intel 18A process node. Standing for 18 angstroms, 18A is the most advanced node fully designed and produced in North America.
It features an industry-first combination of technologies supported by a broad ecosystem of IP, EDA, and design services partners. It delivers better performance per watt and improved area efficiency, enabling new computing architectures for applications demanding the highest level of performance. Intel Foundry's PowerVia introduces a first-of-its-kind backside power delivery technology. Traditionally, power is routed through the front side of the die, creating congestion. PowerVia places power wires beneath the transistor layer, separating them from signal wires. By eliminating the need for power routing on the front side of the die, more resources become available to optimize signal routing. The result is a 10% density boost in cell utilization and a 4% improvement in iso- power performance. RibbonFET, a revolutionary leap in transistor technology, is featured in the 18A process. It uses ribbon-shaped channels surrounded by a gate, offering better control and higher drive current at all voltages.
The incredible combination of RibbonFET and PowerVia paves the way for innovations that will shape the future of high-performance computing, artificial intelligence, mobile, and networking. That's the power of Intel Foundry. As Naga said, 18A is in risk production today and is ramping for our first customers in the second half of this year. Our engagement with customers on 18A, because of the improving and maturity of the silicon and all the investments that our EDA ecosystem partners is resulting in substantial uplift in our 18A platform. We've had over 100 customer and ecosystems tapeouts since we've started to develop the technology. It's a great technology that we really are very proud of.
However, if you reflect back a few years ago when this technology was defined, there was really primarily one voice at the table, a voice describing the primary needs of a company that was developing general-purpose compute processors. We've discovered in our discussion with customers that work is needed to address their diverse design styles and support more Foundry customers in more markets, which is the why behind 18A-P, team. Using the voice of our customers, we're offering more VT flavors, which allows us to drive improvements in performance per watt, improve the characteristics of the technology at lower voltage, not necessarily the target market for general-purpose compute. 18A-P is design rule compatible with 18A, meaning companies that started to develop in 18A can reuse almost all of their work in 18A-P.
But 18A designs can be retimed to take advantage with our new libraries to maximize the benefits of our power performance in new libraries. And all of our IP partners are already underway re-characterizing their IP to take advantage of 18A-P. And we have two products already taped out in the fab today in 18A-P. Think back to that first animation I showed, the one with all the chips stacked on it, the future state of the art. That is a 3D base die packed with not just SRAM, but also tons of logic supporting network traffic functions and other fabric. This requires TSVs. I think we all know the state of the art today is enabled by Intel and our competitors. It's about a three-nanometer process. Intel will be leveraging our learning with Intel 3's TSV to develop 18A-PT. We're extending our TSV enablement to that 18A-P technology.
This will be the best next-generation base die for HPC, AI, and edge applications. And because we're using the 18A-P technology as the base platform for those TSVs, it allows full IP reuse. Even full design blocks can be reused for a fast transition. And our EDA partners are working to enable that effort seamlessly for all of our customers. Let's transition to 14A. Again, the why. You'll hear a lot more from my colleague Ben later today, who's going to talk about not only that we're progressing to the plan that we first laid out here last year, but if you look very carefully at some of his numbers, you'll also see we're setting our sails even higher for the PPA characteristics for our 14A technology than we originally disclosed. And that's based, not aspirational marketing.
That's based on the learning that this team is doing using and bringing up our 18A technology. Let's go back to the customer lens. We've already delivered the 14A PDK to multiple customers in multiple markets. Yes, AI, but also edge applications. The plans are already underway with multiple customers for test chips that were being developed with them. And if you reflect back on the lessons that Naga described for 18A, where we as a team, an Intel team, were in catch-up mode, we just were. We're developing two fundamental new technology elements in parallel with a variety of other priorities as we were bringing up other technologies. Our commitment based on that learning in 14A is for that technology, yes, to be second to none, but by building on all of our experience in 18A, this is our second-generation gate-all-around technology.
This will be our second-generation backside power delivery technology. We're committed to our customers to make our execution more predictable. And you also heard from an even larger expanding ecosystem of partners. We're moving our ecosystem efforts away from solely design enablement and IP partnerships into areas like process R&D, physical design, and ramping yield and quality. We're making major, major investments to build trust with our customers. We all know the journey to serve our industry comprehensively is not just about the Angstrom Era. Mature nodes don't just help us meet customer needs that are diverse. Frankly, enabling us to leverage existing capital assets is a critical part of building a healthy Foundry business. And we know we need to do that as part of Intel Foundry. Our first node that we enabled with customers is Intel 16. It's now been described based on our existing facility in Ireland.
We are very fortunate to have teaching customers and partners working with us as we developed our go-to-market capability. I'm pleased today to be part of a MediaTek announcement where they're announcing that their first product in Intel 16 not only taped out and is through the fab. Let's hear from them directly. Hello. I'm Vince Hu, and I'm the Corporate Vice President for Strategic Marketing and the Compute Business at MediaTek. I'm happy to share our experience working with Intel Foundry. Today, Intel Foundry's expertise has been instrumental to enhance chip performance while minimizing power consumption and area. The Intel Foundry team has been a reliable partner, consistently supporting us in achieving our customers' goals and pushing the boundaries of innovation. Their responsiveness to challenges has been remarkable. Whenever issues arose, they provided swift and effective solutions, helping us mitigate risks efficiently.
Given the global reliance on semiconductors, having a robust and diversified Foundry supply chain is essential. MediaTek values this diversification for business resiliency and continues to evaluate expansion opportunities. Intel Foundry's capabilities in fab interactions and operations have been instrumental in processing our test chips. This collaboration has provided valuable insights, allowing us to refine our designs and achieve A0 first silicon success with our first product before moving to full-scale production. Moreover, Intel Foundry's commitment to maintaining high standards in manufacturing and reliability has been affirmed through our fab audits. Their comprehensive quality defense systems and operation disciplines ensure that MediaTek products are ready for the market. In summary, our partnership with Intel Foundry has been positive. Their experience, reliability, and commitment to excellence have supported us in achieving several strategic goals. We look forward to continuing this collaboration and driving innovation together. Thank you.
Thank you, Vince, and to the whole MediaTek team. It's our privilege to work with you and serve you, and I think you all know A0 to production is the finish line that matters most in our industry. Let's move on to our 12-nanometer partnership, our next evolution of our mature node technology centered in our fabs in Arizona. We're leveraging, in this case, UMC's strength in technology development, design enablement, and their trusted customer relationships. We're making incredible progress here, and we're already working together to plan derivatives to support specialty Foundry markets leveraging this 12-nanometer node. We're getting the production line up and running, validating, and matching our PDK. Let's hear from UMC.
Hello, everyone. I'm Min of UMC. Glad to be here. We are pleased with the progress we have made on the 12-nm collaboration with Intel.
Both teams have been working closely together to achieve the critical milestones. We truly appreciate the collaboration and have witnessed the positive changes within Intel as you progress toward becoming a leading foundry. Together, we are creating greater value for our mutual customers. Wishing you all the best and a successful Direct Connect event. Thank you very much. Thank you, Min. I can tell you all very directly, this is a great partnership for us. They're a fantastic teaching partner. Silicon, as you all know, is just part of the equation for a system foundry. Almost all of our customers are looking for ecosystem support for things like EDA tools and IP, and in some cases, even design services. But in many cases, they're looking for even more. They're looking for full turnkey solutions from initial specification all the way through to silicon and even qualified parts.
We're proud today to announce the formation of a new Value Chain Alliance, three partners enabled on Intel technologies to do just that: full turnkey design capability. We're also today announcing the formation of a Chiplet Alliance, more than a dozen companies from a variety of disciplines who are committed to ensure secure, interoperable chiplets. These two new alliances join the growing ecosystem of companies that are investing together with us for Intel Foundry's success. I'm incredibly thankful for all of these companies. And when I'm done talking today, team, I hope you will all join me outside this hall and spend some time looking to what our ecosystem partners are up to. It's extraordinary, and we're so thankful for them. Before I conclude in silicon, I do want to acknowledge one specific customer for us, a foundational customer, and that's the U.S. government and their key suppliers.
Customer feedback, whether commercial customers or the government customers, are telling us consistently they all want secure end-to-end research and development to manufacturing capabilities for leading-edge technologies right here in the U.S. The U.S. government has stepped up to the table and has committed investments to support that vision along with us. Programs like RAMP and RAMP-C allowed us to jumpstart the design and prototyping capability of our 18A technology, and the Secure Enclave program recently announced enables trusted manufacturing for key U.S. government programs, and the SHIP program extended the capabilities beyond silicon to include things like advanced packaging. Taken together, all of these investments ensure capabilities from design to silicon to system, all right here in the U.S.
Summarizing our silicon journey, team, we are production ready this year on 18A and Intel 16, and we're engaging with customers to solve their problems by developing the 18A, 18A-P, and 14A platforms and extending our partnership to 12-nanometer with UMC. Every one of these technologies will start manufacturing right here in the U.S. and all are comprehensively enabled by our ecosystem partners. A true system foundry doesn't stop at silicon. That's an incredibly important differentiator in today's world. If you think back of the product I showed you initially, that requires full system foundry capability. My colleague Navid will show us more this afternoon, but my focus today in discussing advanced packaging is to, again, bring the voice of the customer, the why, into what we're doing. Our EMIB technology is being enhanced with EMIB-T, adding TSVs to our existing silicon bridges.
Applications like HBM4, so critical for the next generation of AI applications, and UCIe, allowing chiplet interconnectivity to be enabled with the lowest possible cost without sacrificing performance. Our Foveros technologies enable lower cost with more flexibility now with Foveros-S, Foveros-R, and Foveros-B, incorporating things like integrated voltage regulators and MIM caps directly into our Foveros technology. And for us, team, EMIB is all about scale. Big, bigger, and biggest. And I do mean the biggest. Well greater than 12x reticle size, 120 by 188 millimeter package size, and dozens and dozens of stacks of memory, all enabled in a heterogeneous package solution. In our Foveros Direct 3D technology, the customer voice is driving us to improve our bump pitch, expanding the capabilities of both the top and the bottom die that we showed you earlier. As your needs are expanding, we're investing to meet those needs.
Let me spend just another second talking about the why. EMIB 2.5D is the best option for many AI applications. Bridge offers advantages over large interposer technologies. It enables lower cost without sacrificing performance. It allows us to eliminate the wafer-level assembly process, which is a faster time to market. Team, this is not days. This is weeks in manufacturing cycle time. And because an EMIB bridge is a smaller die than a multi-reticle size interposer, it scales much more efficiently. And all of our EMIB technologies can be made right here in the U.S. Like everything, advanced packaging technology is not just about the technology itself. It's about an ecosystem. And I'll be direct. At Intel Foundry, we have not historically been well enabled in our advanced packaging technologies in particular with an ecosystem. We've listened to that feedback.
Our EDA and IP partners are ensuring that EMIB designs and Foveros designs can comprehensively use industry-standard tools. And all of our designs will be compatible with interface IP from our IP partners no matter where they're manufactured. We've enabled a broader spectrum of substrate suppliers with our EMIB technology so that we have supply assurance globally for EMIB technologies. And in commercial ATE team, we at Intel have historically leveraged a proprietary test platform, but our customer voice is allowing us to expand our test offerings for turnkey services with companies like Teradyne and Advantest. And finally, we have some customers telling us that they're looking for us to do EMIB packaging on wafers that we didn't manufacture. You know what? We're happy to do it. And we've enabled EMIB bumping of other foundries' fab wafers with companies like PTI. But some customers are asking us for even more.
They're asking us to enable a broader ecosystem for EMIB assembly, which up until today has been done entirely internal to Intel by Intel Foundry. Today, I'm excited to announce a partnership with Amkor. Please join me in welcoming Amkor's COO, Kevin Engel, to the stage.
Welcome, Kevin. Good to see you. Thank you very, very much. Can you tell us a little bit about what we'll be doing together? Yeah, absolutely. So Amkor is working with Intel to qualify the EMIB process. And the goal here really is to provide your customers with a seamless performance, quality, and reliable flow for EMIB, whether it's done in-house in the Intel Foundry locations or at an Amkor backend facility. Excellent. If you made a decision, Kevin, to make this investment in partnership with us, what's in it for Amkor? Why did you do this?
Yeah, so I think a couple of things. Primarily, EMIB we see as a compelling architecture and technology, especially for AI-type applications and products. So if we look at how that transitions, especially when we talk about large form factor packages as well as multi-reticle die size or multi-reticle size interposers, we really see this as a key advantage. For us, this is a strategic addition to our advanced packaging portfolio and really critical for our future. Our customers are telling us they're looking for supply chain diversification. Does this help them in that? Yeah, so I think this does, I mean, especially for EMIB-type products. If you think about Amkor in general, we have the most diverse geofootprint of any of the OSATs in the industry. We have advanced packaging in our Korea facility as well as in Portugal supporting the European market.
We've announced that we have plans to break ground in our U.S. facility in Arizona this year. So that also will diversify the supply chain into a U.S. offering and will be the first at-scale OSAT with advanced packaging in the United States. Awesome. Last question, the most important one perhaps. When? So we plan to be fully enabled by the end of 2026. We'll start in Korea with your team, and then we'll grow beyond that as the customers and you guys need.
Outstanding. Kevin, thank you so much. I really appreciate you joining us today. Have a good one. Much appreciated. You too.
There you go. Us embracing the different, team. Us embracing the different as a foundry team based on the voice of our customers. We're thrilled to have Amkor now as part of our ecosystem of partners.
So our customers are guiding us in advanced packaging to both accelerate and broaden our offering. Today, we talked about new EMIB and Foveros offerings and substantial improvement in the comprehensiveness of our ecosystem. Let me bring it home. Back to the slide I started with. I showed you this graphic at the beginning, which is your vision for the future. Intel Foundry is working extremely hard to adapt to the AI era from process technology, advanced packaging, and massive substrates. This thing really is extraordinarily impressive in a PowerPoint slide, but I promise you, when you can hold it in your hands, it will be far, far more impressive. This is an extraordinary solution defined by you, but enabled by Intel Foundry. I have to awkwardly smile. It says awkwardly smile for three seconds so they can take a picture. Let me wrap up.
My key message is, team, first, we're listening. We really are listening to your needs. Second, we're committed, committed to delivering our roadmaps, evolving roadmaps, and delivering them on time. And third, we're working very hard to work with intentionality to win your trust. We're very, very thankful for the opportunity to do so. Thanks for the time today, team.