5G: Off the Fence Rail and into the Ring at IMAPS DPC 2018

5G: Off the Fence Rail and into the Ring at IMAPS DPC 2018

The cynic inside often thinks “All hat, no cattle” about the newest new, new thing (right, Michael Lewis?). But if you were to consult the Gartner Hype Cycle listings from 2008 as I just did you would find, perhaps to your surprise, that Cloud Computing, 3-D Printing, Solid-State Drives, and Augmented Reality are now things, having made it past 2008’s inflated expectations, and the disillusionment trough trap, to emerge, in 2018, as some pretty solid businesses.

Right, Amazon Web Services? Right MakerBot? Right SanDisk? Right Apple?

5G is nowhere to be found on the 2008 Hype Cycle chart; heck, basic smartphones had just come out, it was still a 2G-3G-LTE wireless world (at least, for LTE, in some more advanced parts of the world), and we were happy for it.

Gartner today categorizes 5G as being at the “Innovation Trigger” stage in its development, but in my recent travels, say to SEMICON Korea 2018 last month and now to IMAPS DPC 2018 in Fountain Hills, Arizona, two weeks ago, I say that 5G has come off the fence rail and entered the ring in a serious way.

Let me help you understand why that is.

Time won’t change you
Money won’t change you
I haven’t got the faintest idea
Everything seems to be up in the air at this point
I need something to change your mind
Mind, by Talking Heads

First, consider the titles of these DPC 2018 talks and who it was giving them: “Packaging and Integration Strategy for mmWave Products” (Urmi Ray, STATS ChipPAC); “Innovative Wafer Fan-out Technologies -Heterogeneous Integration for a Connected World” (Curtis Zwenger, Amkor Technology); “Advantages of through glass vias (TGV) for RF Front End” (Pramodh Bangaloremadhuranath, Corning Incorporated).

And then consider a little bit more detail from Urmi Ray: “The fifth generation (5G) mobile communication era is expected to address the insatiable need for data communication by introducing mmWave technology and protocols … The key to enabling this architecture is packaging and system integration, especially involving an effective antenna structure and RFIC communication in cost-effective, small form-factor packages.”

Materials and process needs may have special 5G wrinkles, for example, the characteristics of dielectric materials, new or existing, at 5G – mmWave frequencies, and the interaction copper interconnect roughness has on signal transmission, but these are wrinkles that respond to our engineering irons, without resorting to developing fundamental science.

Oh, and fan-out packaging, as in eWLPB? Well, it might just be ideal for 5G, like it is proving itself to be today for automotive radar.

Or consider this from Aric Shorey, Corning: “As the wireless industry continues to address consumers’ need for more mobile data, one thing is clear: The 5G networks of the future will require more spectrum. Managing the explosion in frequency bands in an efficient and cost-effective manner will be a major challenge for the successful deployment of next-generation networks.”

Here, again, materials and their properties are important parts of the 5G equation; about RF switches for wireless front ends, Eric suggested that “… most previous efforts have struggled to achieve minimum reliability performance due to inferior materials and struggled to achieve the necessary cost structures due to process complexity and packaging limitations.”

The answer? Glass substrates. (I’m thinking this is something Professor Venky Sundaram, Georgia Tech, embraces too.)

I’m not worried.

Phil Garrou knocked the ball out of the park with his excellent DPC 2018 opening keynote presentation on “3D-IC: Past, Present, and Future.” There’s a whole blog itself  devoted (mostly) to Phil’s talk, but the takeaway from Phil for me that applies to my thoughts about 5G here is this: there’s a basic disconnect with making market predictions with a ten-year horizon for technologies that aren’t yet fully developed today.

Phil was referring to what turned out to be overenthusiastic predictions about commercial 3D-IC business forecasts (there were some, were there not) and, while 3D-IC components are in just about every smartphone camera module today, and while HBM memory now clearly supports the monster GPUs we’re seeing from NVIDIA and AMD, there just hasn’t been as much 3D-IC as we were hoping.

The market researcher’s dilemma is that no matter how good you are at forecasting you’re never going to get the numbers exactly right.

Which is why I think Gartner, and its Hype Cycle framing of market potential for new technologies, has a better way. Sure, some technologies (electronic paper, Gartner Hype Cycle 2008) seem to have fizzled out, but that’s the nature of the beast – fizzle some will.

5G sitting in the innovation trigger stage of life doesn’t seem quite right to me. With our insatiable need for data and high bandwidth delivery of same, both for consumer IoT and for industrial IIoT, much less for the real-time vehicle-to-vehicle and vehicle-to-cloud communication channels every autonomous vehicle will bank its life (and yours!) on, 5G seems inevitable.

Sure, 5G standards aren’t completely in place yet, and we still have two years to go before “official” rollout at the 2020 Tokyo Summer Olympic Games, but 5G was in the air at DPC 2018, already, and in a big way.

Personal fuel cells (Gartner, 2002) never made it off the fence rail, but 5G has clearly entered the ring.

From Santa Clara, California, thanks for reading. ~PFW

p.s. For a copy of my IMAPS DPC 2018 talk on “Linear Transport Degas, Pre-Clean, and PVD Processes for RDL Barrier/Seed Formation in Fan-Out Packaging” please contact me at pwerbaneth@intevac.com.