Ultratech's Investments Pay Off After Lean Time; Memory chip market, though down, ironically is a key to bright outlook

Sales growth at Ultratech, a small chip equipment maker, is running ahead of a lot of its larger peers.

Ultratech's products include machines that chipmakers use to take photos of chip designs and imprint them onto wafers. Chipmakers then etch these designs into the wafers and cut them into chips.

Ultratech got into trouble in 2005. Rivals came out with new products earlier than it did. Its sales fell for the next two years, and it lost money from 2005 to 2007.

Chief Executive Art Zafiropoulo started investing in new tool designs back in 2005. In took a while to turn the company around, but in late 2007, sales and earnings started growing again.

Sales last quarter rose 7% from the year-earlier quarter to $32.1 million. Analysts expect sales of nearly $132 million this year, which would be up 18% from 2007. Research firm Gartner forecasts an overall 21% decline in sales this year for the chip gear industry. Ultratech's stock has risen more than 80% since mid-March, to 15.65.

Zafiropoulo recently spoke with IBD about Ultratech's prospects for the second half of the year.

IBD: What's your outlook this year?

Zafiropoulo: I feel good about the company. As time goes on, it's reinforcing the strategy we put in place and investments we put in place. We're going opposite the trend in the equipment industry.

IBD: What is the trend?

Zafiropoulo: I see all these bad reports coming out. There's good news and bad news there for us. The good news is, we're not focusing on memory chips at a time when sales in that market are down.

The bad news is, we don't focus much on memory chips at all. Today, more than 50% of equipment spending is by memory makers.

We're beginning to penetrate memory companies. We feel this is a good time for us. When recovery of memory takes place, we should see acceleration of our business model. We feel good about that.

IBD: When will memory prices recover?

Zafiropoulo: Earlier, I had thought we would see some recovery in the fourth quarter. It's becoming a little less likely that will occur. We see memory chip inventory increasing. I just don't think much is going to happen between now and the end of the year.

IBD: You mostly sell to chip foundries and to makers of processors, such as Intel. How are those markets doing?

Zafiropoulo: We see our sales in the foundry and logic (processor) areas as being relatively strong. After Q1 we had indicated we saw a strong year, and there's nothing out there that tells me it's not going to happen.

IBD: Gartner put out a report on July 8 that forecast equipment industry sales falling 20.6% this year, after rising 6.4% in 2007. Do you agree?

Zafiropoulo: Yeah. Understand something, though. Marketing forecasts are always wrong. That's why people keep updating them.

IBD: Gartner looks for 9.8% industry growth in 2009. Does that sound right?

Zafiropoulo: I think that's quite reasonable. If the recovery occurs sooner, some time in 2008, I think it would be too high (because more sales would have occurred in 2008). If recovery doesn't occur until the second quarter of 2009, I think it would be too low (because pent-up demand would increase sales even more).

IBD: Given the weak market now, do you look for mergers and acquisitions in the equipment industry?

Zafiropoulo: We keep saying there has to be some consolidation, but it hasn't happened yet.

This is a very unusual industry. They're into a lot of synergy, strategic fits. If you're a high-margin company, do you want to take on a low-margin company? There's a lot of consideration of investor value.

IBD: There's a lot of buzz about 3-D chip design. What is it, and why is it important?

Zafiropoulo: Today we build transistors on a single plane of a chip. We interconnect them with wiring.

At some point the technology will reach a limitation. At that point the question is: What can we do to continue Moore's Law? (Intel co-founder Gordon Moore said chipmakers would roughly double the performance of processors every 18 months.) With 3-D design, after you build the first layer of transistors you put a thick layer (of insulation) on top of that and then another layer of transistors. You can actually accelerate Moore's Law with 3-D design, without shrinking feature size. This is critical to where the industry is going.

IBD: What type of chips would most benefit from 3-D design?

Zafiropoulo: The first application will be in Nand flash (the common type of memory in cell phones, digital cameras, digital music players and other such products). That's where the biggest price pressure is.