A broader Aixtron: Q&A with deposition specialist Aixtron AG, part two

Aixtron AG has specialized in MOCVD technology since 1983, but the Europe-based company is now brandishing its latest acquisition, the California-based Genus, which supplies both a CVD and ALD capability to complement Aixtron’s traditional emphasis on MOCVD as well as its more recently developed AVD technology.

That all adds up to a formidably broad range of technology options, leaving the company well positioned for the technology shifts that lie ahead. Small or seismic, those technology shifts will now cover a huge range of potential applications, from consumer multimedia players through to displays and area lighting.

DigiTimes.com had the opportunity to talk with Paul Hyland, President and CEO of Aixtron, and Dr. William (“Bill”) Elder, Executive Vice President at Aixtron with responsibility for their silicon interests, about the prospects for the combined company, now that it is 18 months into the acquisition process, and the business challenges and technology scenarios that lie ahead. This at-length, in-depth interview also includes contributions from Christian (“Chris”) Geng, Aixtron general manager for Greater China.

This is Part II of a five-part interview. Part I appeared on November 21. Part III follows on November 23.

Q: With the Genus acquisition, Aixtron gained both an ALD and CVD capability. How is the combined company positioning these technologies, in terms of applicability? My understanding is that ALD has never really managed to sell itself to logic manufacturers, but that it has and is seeing some success in the DRAM sector.

PH: We’re now through the first 18 months since the Genus acquisition, and we’re delighted with our progress in that first period. That’s always the most difficult period, in most mergers and acquisitions. Under Bill’s stewardship, all of our existing silicon business is developing very nicely, pulling together AVD, which Aixtron has been developing, along with CVD and ALD technology developed by Genus. These technologies are now being guided by Bill into one combined package of silicon solutions we can offer our customers. The last 12 months have been exceptionally good, and we’ve had a significant increase in revenue in that period. The contribution from Genus actually constituted just over 30% of revenue in the first half of this year. It’s making a real contribution to the business.

The whole thinking behind the strategy was that our core expertise is in gas-phase deposition, and we are now able to offer that in a very broad range of technologies: MOCVD, OVPD, ALD, AVD and CVD. We also believe that as the silicon and compound industries develop and converge, we’re very nicely placed to take advantage of these emerging opportunities.

Before I hand over to Bill, I think my previous remarks about the timing of OVPD development also apply to the anticipated emergence of ALD, in that we, as an industry, always underestimate the ability of the current technology to be extended. I am confident that ALD will deliver on its full potential, but it will take a little bit longer that was at first envisaged. It is increasingly looking like ALD will be adopted, starting at 45nm, in several different applications.

BE: Let me just elaborate on what Paul has just said. Mergers are always difficult, but I think the success we’ve enjoyed over the last 18 months, on the Genus side, which is now becoming more and more part of a single Silicon and Semiconductor technology business unit within the Aixtron organization, incorporating the atomic vapor deposition (AVD) technology that Aixtron had been developing. AVD consolidates very neatly with the ALD and CVD technology that Genus had been developing. So now, under one Aixtron-Genus flag we’ve got all three technologies really driving into the marketplace.

Over the last 18 months we’ve been very successful with all three technologies. In the first six months of 2006, revenues generated from the sale of silicon semiconductor equipment totaled €22.9 million, representing 34% of total six-month 2006 revenue. In fact if you look at the Genus product line today, revenues will be up 50% this year over last year, mostly driven by the CVD 300mm system. In the meantime, we have positioned our advanced AVD and ALD products with JDPs (Joint Development Programs) at some of our key customers.

If you look at what we’re doing this year, compared to what we did in 2005, there’s been a significant increase in revenues on the Genus side, and that’s the benefit we’re getting from incorporating the core competence that Aixtron has in Aachen with Genus’ core competence in Sunnyvale. Under my direction, we’re now bringing in some of the key people from Aachen; we’re transferring key personnel to Sunnyvale to consolidate and focus the silicon part of the organization. And from my own standpoint, I’m pleased to say that we did not lose a single senior manager or technical person from the Genus side during the merger.

We talk about acquiring technology, but at night our technology pretty much walks out the door, in that it is driven by our people. Making sure that you keep the people to ensure the success of the technology is critical. I think Paul and I have done an excellent job in making sure that we managed that transition. I believe that’s behind us now. Now, we’re really focused on the market and what the three technologies we have within silicon have to offer the customer.

PH: The core of the Genus revenue over the past 12 months, certainly the majority of it, has been in CVD, and that’s been driven by growth in both DRAM and flash applications. That is likely to continue with the imminent arrival of Microsoft’s Vista Operating System, which will increase the amount of memory in most consumer PCs to around one gigabyte. Vista will also require a certain amount of onboard flash memory, and with the continual increase in the number of applications drawing on those devices, I think that CVD in the short-, medium- and long-term, continues to look very promising.

I also believe that what Genus brought to Aixtron, in addition to ALD and CVD technology, was a significant wealth of experience with commercial customers, over the past 20 years. Over the last 12 months, Bill has been able to make major progress in placing both AVD and ALD systems in R&D applications, particularly here in Asia. That bodes well for us, particularly in terms of the next-generation products that will come through for both memory and logic applications, as well as for other devices.

BE: You know, there’s been much written about a reduction in expectations for ALD, the “push out” and so on, but what’s not been written about is how with the push out in new materials, the incumbent technology enjoys a much more robust life, for a longer period of time, and that’s what we’re seeing today. Where we thought that our CVD for memory products would be beginning to subside, in fact, we’re seeing additional life in those technologies and that’s what is really driving the increased revenue we’ve been seeing on the silicon side, this year. Meantime, we’ve been placing JDPs with our key customers, both for memory and for logic applications, where we’re really bringing the strength of AVD and ALD to the market, and we’ll be ready for the 45nm and 32nm technology nodes that will require new materials and new processes.

Q: In the meantime, with Vista and the PC, it sounds as though we’ll be seeing developments in BIOS technology plus developments in hard disk drive technology.

PH: One aspect of Microsoft Vista is the ability to have onboard flash speed up the boot-up process. You can also have regularly used programs and data codes stored in flash, and that will enable the system to run quicker and more efficiently. Vista should also allow for more comprehensive graphics and data management, and we’re already seeing two of the leading disk-drive manufacturers, Samsung and Seagate, bringing out hybrid hard disk drives, with flash onboard. I believe initially that will be on the order of 128MB, but undoubtedly that will increase over time, as hybrid technology becomes more commonplace. So you’ve now got not only convergence between silicon and compound. We’re now beginning to get convergence between conventional data storage and flash storage, and I think when you look at the advances that some of the big players are making in that space, it’s almost inevitable that the applications will grow.

You can already see Vista-ready PCs already being sold in the market, so I believe some of the increase in investment and volume driven by Vista is already with us, although the consumer-level Vista product from Microsoft won’t be appearing until the New Year. After that, as more people take up the operating system, I think it’s likely we’ll see an increase in the specifications of the average PC on the market.

BE: Over the last couple of weeks, Paul and I have been talking to many of our key customers in the memory space, and what we hear is a predicted continued robust capex in 2007, because all of these new applications are driving next-generation technology nodes, both in DRAM and in NAND flash. So 2007 looks as though it will be another healthy year in that sector if these predictions are true. As Paul said, it’s being driven by Vista and the ability of the technology to identify new applications as you increase density and speed.

Q: And what about flash memory in the camera market, including cameraphones? Do you anticipate increased demand there?

PH: Well, currently the largest CompactFlash density you can get is 16G-bit, but it won’t be long, I’m sure, before they deliver something even denser. And as digital camera pixels increase and become more capable – I think 17 megapixels is currently the highest rating – the need for data storage is just going to continue to grow.

The arrival of 3G mobile phones and download services is also going to increase the need for data storage and processing speed in mobile phones and other portable devices. People are going to download music, TV and even movies to some of these new portable devices. There seems to be an insatiable demand for portable storage. Even so, there will be cycles, I’m sure, as these types of applications go through different phases of development.

Q: And as the retail pricing of flash continues to come down, there will be a tendency for volume purchases to pick up?

PH: Yes, almost certainly. That same dynamic also exists in the LED industry; an increasing volume curve and a reducing price curve. There’s always a direct relationship between the two.

BE: That’s what drives the additional applications, that increased density combined with lower cost. Our customers have increased the density of NAND flash tremendously, and that always drives new applications. A year from now, you’ll be able to download both video and MP3s to your cell phone. That will potentially create a tremendous market further downstream.

PH: If we were to look at the semiconductor industry 10 years ago, and the percentage of products that actually found their way into consumer hands, it was certainly significantly lower than it is today. In the very early days of the industry, the end market character had a big influence on the frequency of the cycle in semiconductors. In the very early days, the frequency of investment largely reflected the capital investment cycle in PCs and servers. Today, the percentage of end-market or consumer-driven applications is considerably higher, and the impact that has had is to increase the frequency of the cycles. We now have shorter cycles, and that in turn has led to a need for greater flexibility in the industry, as well as some consolidation. Personally, I believe that that consolidation process will continue, both in the LED industry and also in some of the semiconductor markets we are serving. It’s a natural market dynamic. (Note: Epistar announced their acquisition of Epitech and Highlink less than one month after this interview was conducted.)

About 12 months ago we happened to be in Korea, when Samsung announced they would double their R&D for the next 5 years. The intent is very clearly laid down for focused regional R&D. I think Bill is absolutely right. The volume and the capacity and the capability of these technologies are driving the end applications.

BE: You know, Samsung today is in a very unique position, a position where they can directly drive the market because they have both the front-end manufacturing technology, whether its display technology or silicon, and the end-product-markets. I believe they are number three in cell phones today. They’re driving the market for backlights for TV; they’re playing a major role in driving the move to mobile technology. They’ve got this whole complement of technologies defining the market, and the investment they’re putting into R&D, across all sectors, is just incredible. Equally important is that the momentum created by such large single players creates huge opportunities for those companies operating in the same food chain, which is particularly relevant here in Taiwan.

Paul Hyland

Paul Hyland, President and CEO, Aixtron: “What we have today is great challenges and great opportunities – in equal measure. I wouldn’t want it any other way.”
Photo: Company

Dr. William (“Bill”) Elder, Executive Vice President, Aixtron

Dr. William (“Bill”) Elder, Executive Vice President, Aixtron: “We talk about acquiring technology, but at night our technology pretty much walks out the door. Making sure that you keep the people to ensure the success of the technology is critical.”
Photo: Company

Aixtron’s Lynx-iXP 300mm CVD System (front)