Talking 3D with Manish Ranjan

I’ll never forget the first time I interviewed Manish Ranjan of Ultratech. It was during my first SEMICON West experience in 2006. I was still wet behind the ears, faking my way through most interviews. The easiest ones were with the companies who just wanted to talk about their latest product launches. Not so with Manish. He started talking to me expert-to-expert, offering market analysis, discussing technology trends, and using lingo I hadn’t yet fully added to my vocabulary. I remember feeling completely out of my league. Panic set in. I had to get out of there quick, before I made a complete fool of myself. When he asked if I had any questions, I just smiled and said, “Nope. I think you’ve pretty much covered it. Thank you very much. See you next time.” Exit, stage right.

Between 2006 and 2007, I learned a lot, and by the time we met again, this time as part of the AP Roadshow, I could hold my own in the interview. Those annual chats have become a personal benchmark of how much I have learned. But I also look forward to them because Manish provides a great supplier-side perspective of what’s going on in the back-end sector of industry. Not to mention that he also provides good sound-bites. Our meeting this year at SEMICON West 2009 was no different.

Since Ultratech’s lithography steppers originally served the front-end market, but re-emerged to focus on back-end lithography needs, Manish offers a unique vantage point on the scaling vs. 3D debate. He’s seen it from both ends, and this is what he says: 22nm, front-end scaling is no longer effective. Companies must make a choice. Either invest in front-end technologies, or turn to packaging. With that “a cost function shifts to a value-add function,” notes Manish. He says the push from the end-customer is getting stronger  as the value equation is justified. Like many others, he points to market adoption of usingTSVs for CMOS image sensors as an example of this, and says baseband chips, microprocessors at sub-22nm, and DRAM are also justifiable applications. It basically comes down to economics vs. performance. Historically for memory, performance was second to cost and with logic, cost is second to performance. But cost be damned, with DRAM, the density issue will drive the need for TSVs.

Ultratech made the choice to invest in wafer-level packaging and 3D technologies well before the current hype, and certainly before I knew much about it. The company distinguishes itself by its optics technology, and had TSV applications in mind when it developed its dual-sided alignment system. As a result of being an early innovator, the company holds 87% of the market share for the markets it serves (flip chip, gold bump, and TSV applications. Manish says when it comes to niche markets, “You have to be #1 to survive.”

Manish also remembers the first time I interviewed him. We always laught about it. When I think about it, that visit to Ultratech is what first sparked my interest in 3D. Thanks Manish, you gave me some great ideas! See you next time. — F.v.T