Are there still people who are clueless about 3D? Apparently there are. Lots of them. Despite all the discussion and attention 3D integration gets on the conference circuit and through media outlets such as 3D InCites, there are still IC design and process engineers working in semiconductor manufacturing who are still unaware about 3D technologies and the potential impact it has on the future of the industry.
What triggered this particular thought today (and inspired me to write this blog post) was a simple question posed by Ann Stefora Mutschler in a blog post on Semiconductor Manufacturing and Design: “but what exactly is 3D?” My immediate internal response was “do we really still have to ask that question?” But then I realized the answer is most definitely, yes. How do I know? Because they often ask ME what it is.
I realize my perception is skewed. I eat, breathe and sleep all things 3D. I go to as many conferences as I can, hoping for something new, and more and more I hear a lot of the same things being presented and discussed all over again. Whenever I am inclined to criticize the content being presented, I take a step back and think about the audience. It is quite possible, I realize, that this particular batch of attendees hasn’t heard all this before. They may be first-timers. Unlike me, they don’t spend their days scouring the Internet for nuggets of new information regarding 3D from transistors to systems to share with a community of 3D enthusiasts. They spend most of their time actually designing and building stuff. They may attend one conference a year, and probably don’t have much time to read up on technology in between. I am a talker. They are doers.
Who’s taking the risk? The aforementioned post titled “The Sky isn’t Falling”, discusses the shake-up to the supply chain that comes with manufacturing 3D ICs from the perspective of EDA vendors and customers. The supply chain topic is not a new one, although getting the perspective of the EDA vendors puts an interesting spin on it. (I’m not going to summarize it here, but I do recommend reading it to hear what Ahmed Jarraya (CEA Leti) and Steve Smith (Synopsys) have to say on the topic.)
Reading Mutschler’s post brought forth other thoughts I’ve been mulling over regarding the differences in the way the EDA vendors entered the 3D IC market vs. the way the equipment manufacturers and materials suppliers did. Both were required to make enormous investments in R&D to make 3D IC a reality. However, because “A” stands for Automation in EDA, there’s an excepted school of thought that the need for 3D specific EDA tools is unnecessary until the potential for high volume manufacturing (HVM) is established. In the meantime, the general consensus is that existing tools are “good enough” to get started.
On the flip side, the equipment manufacturers and materials suppliers invested years ago to develop the tools and processes necessary to bring 3D ICs as far as they’ve come. To do so is an enormous risk, and they must have done their due diligence to be sure they would realize a return on investment. Additionally, “good enough” does not exist in the process/materials/equipment world, where high yield is the name of the game. Nothing moves forward until they get it right. So does that make them the greatest risk takers in the whole 3D effort? Maybe and maybe not. It’s quite possible they know something the rest of us don’t.
Doers don’t necessarily talk about it. Consider the following: while there are those of us who are talking about it, hashing, rehashing and debating the same issues, you can be sure that there are those already out there doing it and not talking about it. It doesn’t take a brain surgeon to know that the manufacturing giants like to gain the competitive edge working in stealth mode, and only sharing knowledge when they require the participation of others to meet their goals. Those “others” have to include all of the aforementioned suppliers at some point in development, who are of course sworn to secrecy by virtue of an NDA. These guys are talkers AND doers.
But we still need to keep the conversation going. While doing it is the ultimate end goal, talking about 3D is still important because there’s an educational learning curve. Before it can be done in HVM across the entire industry, we need to get to that place where we longer hear the question “What exactly is 3D?”. ~ F.v.T.