Intel CEO Pat Gelsinger is in Taiwan for the Intel Vision event, which is aimed at high-level executives and managers from customers and ecosystem partners. In a Q&A with the Taiwanese press, Gelsinger stated that Intel is re-engaging with the foundry business, driving and defining crucial standards for the future utilizing open technologies, and revitalizing product execution to restore competitiveness in the market.
The following is a synopsis of the Q&A session.
Q: What is the purpose of this trip to Taiwan?
A: Many members of my executive team are in attendance at the Intel vision event for customers and ecosystem partners. And with that, we hope to assist them in understanding the Intel Vision as well as how we may effectively interact. And, as you may be aware, Intel and the Taiwan ecosystem have a long history of collaboration. And I like to tell people that I met many of the OEM leaders when they were young and just starting out, so I've had decades of fantastic relationships with the Taiwan leaders.
Q: As you mentioned, Intel collaborates extensively with its partners and customers in Taiwan. But given the geopolitical uncertainties, do you believe Taiwan's semiconductor sector is in jeopardy as a whole? If that's the case, how does Intel handle it?
A: From Intel's perspective, we believe that the world needs more balanced and resilient supply chains. And that's been a theme since I took over as CEO, and something we've been working on with governments around the world because, as I like to say, technology supply chains are more important than where the oil reserves have been for the last five decades. Everything is going digital, so we need to think about our technology supply chain resilience for the future in a very strategic way.
We think that applies to the entire supply chain. Globally, we need more fabs. We need more systems to be created, we need to consider the subcomponents, we need to consider packet defense against unexpected attacks, and they need to be balanced and resilient globally.
We believe that Taiwanese businesses would benefit by being less reliant on a single source of goods or location, as customers would likewise expect them to have a more resilient supply chain. Naturally, it has been an important component of our build out of our production plants around the world for us. We therefore think it is the best course of action for good business, good customer requirements, and good geopolitics.
Q: Micron product sales are being banned in China. Qualcomm and Broadcom are said to be the next companies on the list. And, from what I understand, Intel collaborates with all of these businesses. Intel is very well-represented in China. So, what is Intel's take on this?
A: When I was in China, maybe three or four weeks ago, I spoke with a lot of Chinese political and commercial figures. Intel has been active in China for 39 years, and they gave us a really warm welcome. We were also reminded of the crucial role we played in China and how our technologies and products have helped its people for more than 40 years.
Despite some political disagreement, we regard the business community as a bridge between the US and China that can help forge a lasting connection between the two countries.
Additionally, we reaffirm our dedication to helping them and being a good partner to our customers in China. Even though there may be greater conflict in the partnership as a whole, for them. We aspire to be known as a business that simply upholds the interests of its Chinese clients, the Taiwan ecosystem, and its global partners. We are pleased with how well this strategy is working for us.
We do hope that more normal interactions between the United States and China will take place, and that the various Cabinet members and their Chinese counterparts can engage in more productive dialogue. Again, part of my upcoming trip to China in July will be to reinforce the value of good dialogue.
Q: How do you navigate the Chinese market now? Since there is a great deal of legal uncertainty. Second, Intel is seeking backing from the US government for the US Chips and Science Act. So, how can you figure out a balance between the two forces?
A: We are a US company, so we must follow US laws. However, we have been operating in China for almost 40 years and have substantial customer meaningful revenue there. So it's an important market for us that's growing. It's necessary to be on both sides of the fence when watching. And, as I already stated, manage with caution.
Q: Intel has proposed a four-year timeline for the accelerated development of five process nodes. Will it be too challenging to complete?
A: Intel is confident and is currently adhering to the timeline. The fourth-generation Xeon Scalable processor, codenamed Sapphire Rapids, which uses the Intel 7 process, was released in early 2023; the fifth-generation processor, codenamed Emerald Rapids, will be released in the fourth quarter of 2023, using Intel 7; Sierra Forest will debut in the first half of 2024, using Intel 3; and Granite Rapids will be released at the end of 2024.
Following Granite Rapids in 2024, an Intel 18A-based server CPU codenamed Clearwater Forest will be released. It is expected to be introduced in 2025, and relevant verification will be carried out at the end of 2024.
In July 2021, Intel announced its goal to construct five process nodes in four years. The industry was stunned. It's been less than two years, and the development has been rather straightforward, and it's still being made step by step.
Q: Intel Foundry Services (IFS) has a modest revenue scale and continues to incur losses. Would you share your most recent IFS strategy with us?
A: We have clearly put out a concept where we must have proper firewalls for our IFS business so that customers can have confidence in us and that when we make guarantees on capacity corridors and their designs, those are protected. So we believe strongly that we must be able to provide such commitments, and as our foundry company expands, their trust develops.
My internal product teams regard Intel's manufacturing as the internal foundry. And we've defined it as putting internal and external foundry customers on an equal playing field, so that we can truly benefit from external foundries while also letting my internal product teams to benefit from other foundries. And in doing so, we effectively create a divide between design and manufacture.
We have also described our goal as a systems foundry, including wafers, packaging, chip standards, and software assets that we present to our foundry customers. And this is a strategy that has piqued the curiosity of some of the major foundry customers, who see Intel as bringing a deeper set of services to those potential partnerships.
In fact, some of the most promising foundry opportunities in our pipeline right now include combining Intel chiplets with their designs on advanced packaging. In this regard, they do not seek separation; rather, they want to be able to profit from Intel for both goods, chiplets, and their ideas into a single SoC. So we believe that this is a differentiator for the whole foundries approach, and that it actually provides us with more opportunities.
I believe it to be the end of June we will describe in greater detail to the financial and industry analysts how Intel's internal foundry model creates essentially two organizations within a single entity with the ability to deliver more comprehensive solutions from Intel's systems foundry.
Q: What are IFS's medium and long-term objectives?
A: By the first quarter of the next year, I will deliver the P&L for Intel in this manner, with the IFS and manufacturing being presented as a separate P&L for Intel that is different from the product businesses of Intel.
It takes a lot of work to restate the company in this way and be able to deliver a comprehensive P&L for manufacturing and IFS. And once we start doing so, we'll give you specific instructions on how we plan to boost the company's finances over time.
We've stated that our goal is to be the second-largest foundry by 2030. Aside from that, we don't currently have any stated public KPIs, but our goal is to become the world's second largest foundry by 2030.
Q: What are your thoughts regarding the recent market decline? Do you believe now is the right time to expand?
A: We believe there is still enough bad news about the global economy, but we anticipate some growth in the second half of this year. It includes making adjustments to the inventory. By the second half, trade will be fully restored at least in the PC industry and possibly in the data center sector. Generally, the second half is stronger than the first.
Given the market slump, we feel we must trade operational costs while continuing to make long-term capital investments with investment cycles of four to five years. And, as you know, we anticipate that the semiconductor industry will roughly double by the end of the decade. As a result, both TSMC and Samsung intend to expand, despite the fact that we must be cautious about short-term financial concerns. Long-term strategic capital investments cannot be abandoned.
Q: In the year 2000, Intel dominated its industry. At the time, you were the CTO of Intel. At that time, it was also an honor for the Taiwanese media to attend the IDF, but Intel abruptly canceled the event. Then with the rise of AMD, Apple, and other rivals, Intel's market share has begun to fall. As a result, you have assumed leadership of Intel at a crucial juncture in its history. Which viewpoint do you hold? And why should you have thought you could restore Intel to its former glory when you were CTO in 2000?
A: The points you mentioned are ones that saddened me, you know. I had started IDF for Intel and to me, the idea of bringing the ecosystem of innovation standardization together was a powerful one. Now we're rebuilding, and you may recall some of those industry engagements. And I believe that the era when Intel ceased engaging with the industry was one of the factors behind its current struggles.
Intel has also made inappropriate decisions about process technology planning. Intel pioneered EUV technology but subsequently abandoned it, whereas TSMC has focused on EUV development and is reaping the benefits.
So what I'm doing is reengaging with the industry, right with vision and creativity, to be that significant partner with the industry once again. Second, we've adopted EUV and will be the first company in the next generation of up to return to process leadership, where we create the greatest transistors in the industry, and we're reinventing our development and execution.
We also need to retrieve the product execution mechanism, commonly referred to as tick tock. In many respects, therefore, what was once old is once again new in terms of offering reliable leading technology with an eye on the larger ecosystem.
Article translated by Jessie Shen
