元器件 > 高级半导体博客



Some Thoughts From the International Microwave Symposium

by Eric Higham | 7月 21, 2017

Apologies for the lack of a blog entry in June, but I’ve been immersed in the ever-changing world of 5G. As I’m told, being adaptable in the face of change is a very valuable skill. I put this to the test as I “adapted” to a seat in a plane for the approximately 12 hours of flying to get to Hawaii and the International Microwave Symposium (IMS 2017). Of course, the only “change” I faced was to a different plane in Los Angeles. Now, since it was Hawaii, I followed another piece of advice and I got out of my hotel room a bit more than I normally do. I took many pictures and sent them back to loved ones on the mainland, but I devoted most of my time to the events going on at the conference.

IMS has historically been the largest microwave components technical conference/tradeshow. It has evolved over the years from strictly a defense-oriented show to embrace the growth of commercial opportunities and markets. This year, not surprisingly, a big focus of the technical sessions and the exhibitors in the tradeshow was the continuing evolution of the 5G market. We contributed to this vibe at our annual “lunch and learn” session with “Which Semiconductor Technologies will Successfully Solve the 5G Conundrum?”, a presentation containing our thoughts about how the semiconductor industry is responding to the requirements of 5G.

To illustrate the growing influence of 5G on this technical community, the IEEE organized a two-day 5G Summit at the beginning of Microwave Week. Unfortunately, the second day of the summit conflicted with our lunch session, but the presentations on the first day were very interesting. The very first thing that I noticed was the composition of the panels and speakers. There were presentations from “silicon players” Intel and Qualcomm, operators like NTT DoCoMo, test equipment companies and a variety of small companies at various stages of capitalization and funding. This presented a startling contrast to the defense companies and compound semiconductor companies that dominated this conference in the past.

In an unabashed plug for Strategy Analytics’ grasp on the evolution of the 5G market, much of what was presented meshed well with the trends and forecasts we’ve been discussing for the past couple of years. There was a lot of discussion of how important MIMO (Multiple Input Multiple Output) and beamforming would be to achieve 5G data rates, the possibilities and benefits of enhancing the wireless capabilities of vehicles to 5G levels, the importance of spectrum, the challenges faced in test and measurement and how the technical community is addressing all these challenges.

One thing that definitely surprised me was the amount of 5G IoT (Internet of Things) discussion. IoT represents one of the pillars of the 5G vision, along with extreme mobile broadband and ultra-reliable low latency communications, so it is not surprising from that standpoint. The interesting point is the IoT pillar probably demands the least amount of compound semiconductor technology and probably has the lowest technology “wow” factor (and I apologize to anyone working on IoT applications). The extreme mobile broadband pillar promises data rates in excess of 10Gbps that will make for near-instantaneous video downloads and when coupled with the low latency pillar, will enable things like autonomous vehicles, virtual reality and remote control of all sorts of things in what has become known as the “tactile internet”. What the IoT applications do have going for them, however, is quantity measured in the billions and the revenue from these applications measured in the $trillions! I think that does a very good job of explaining the interest in the IoT pillar…at any type of conference.

One of the other things that I have been talking about for some time is the challenge of power dissipation. While the millimeter wave 5G networks will likely lag behind the volumes of the <6 GHz networks initially, we are forecasting much faster growth for the millimeter networks once they start deploying. This is because of the propagation characteristics of the signals. Higher path loss at millimeter wave frequencies means more base stations to serve a given area. More base stations mean more power amplifiers, but the efficiency of these millimeter wave PAs will be much lower than current cellular power amplifiers. This challenge will manifest itself in a couple of ways; first, it means more operating expense for service providers. Secondly, efficiently removing the excess heat becomes an important design challenge, especially in devices that will be much smaller because of the frequency range.

 

The following chart from a Berkeley Wireless Research Center presentation is very interesting:


Berkeley Wireless PA Efficiency

 

It shows that in the millimeter wave frequency range, CMOS and SiGe have a hard time achieving more than 20% PAE. There is an accompanying slide in the presentation that shows a “millimeter wave” Class A PA that achieves 15% Power Added Efficiency (PAE) at saturation. This PAE drops to 2% at 6 dB back off. Now, I realize that class AB is a more common scheme for efficiency and there is a lot of effort aimed at improving PAE, but 5G back off also looks like it will be closer to 10 dB, so this poses a significant challenge.

Several thoughts spring to mind from this slide. I’m reminded of the expression “if it can be done in silicon, it will be done in silicon”…and the close corollary; “not everything that can be done in silicon should be done in silicon”. I’m also imagining the look on an operator’s face when they realize that they are paying for a design where 95% (give or take) of the DC energy is lost to heat. I’m also imagining that the look of disgust from the operator is only exceeded by a similar look from the designer that has to figure out what to do with all that dissipated heat. Interesting challenge and a big opportunity that all the semiconductor technologies must adapt to in response to the changes brought about by 5G requirements.

We have done a considerable amount of research into 5G, so if you are looking to position your products and strategies, understand how the market is evolving and how to take advantage of 5G opportunities, contact us and we can discuss how we can help with your challenges.

Well, I wish I could tell you that this blog is a make-up for the one I missed in June, but I’m at headed off to vacation in a couple of days. I’ll have a nice family reunion, not as big as in the past, but that’s another example of having to adapt to change. I will check in August and we can commiserate about the end of the summer in Boston!

  • Eric
Previous Post: Is that InP Knocking on the 5G Door? | Next Post: GaN, 5G and the End of Summer
Leave a comment