元器件 > 高级半导体博客



GaN Hits the RF Jackpot!

by Eric Higham | 1月 30, 2015

For those of you who don’t know, Strategy Analytics has an office just outside of Boston and the Boston metro area is recovering from a blizzard of “historic proportions” (according to the weather forecasters). My house is in a swath of the region that hit the “jackpot” with 36” of snow, so, I’m taking a break from shoveling to recuperate and contemplate. I’m contemplating other activities that would still get the blood pumping and give me a workout, but would be more satisfying than shoveling. Not surprisingly, those thoughts are gravitating toward warmer climates where the sand is as soft as the snow I’m shoveling, just much warmer! These musings are a moot point, because the workout that follows will be of the intellectual variety as I discuss some thoughts on the RF GaN market.

 Talk about hitting the jackpot; that’s just what the RF GaN market has done! I just postedGaN RF Market Update: 2014 - 2019, along with the accompanying data tables GaN RF Market Update: 2014-2019 (Data Tables) to the Strategy Analytics web site. The most significant finding of this latest research is the magnitude of the commercial market, particularly the wireless infrastructure segment. Based on the results, commercial applications accounted for slightly more than half of the total GaN RF device revenue in 2014. This is exciting news, but it’s like the band that has been working 15 years to become an “overnight sensation”. This has been in the works for a long time!

 The consensus in the industry has been that GaN-based devices would hit production in RF commercial applications “next year”…for several years. As a result, some RF GaN forecasts (not from Strategy Analytics) had revenue well into the hundreds of millions or even billions of dollars by now. When this didn’t materialize, the forecast would push out a year. GaN has had strong developmental support and usage in military applications from its inception, but commercial adoption has been much slower.

 Over the past several years, GaN processes and products have been solving the issues of reliability and manufacturability and building up a history of successful usage in a number of military and lower volume commercial applications. As the process has matured, price has dropped into a range where it becomes competitive with the incumbent technologies. What has slowed GaN adoption, certainly in wireless infrastructure applications, is the resilience of LDMOS technology.   This technology still accounts for the lion’s share of wireless infrastructure power revenue and developers release a new generation of LDMOS, with better performance and lower cost whenever competitive technologies threaten to capture more market share. The performance of LDMOS falls short of GaN in several areas, but more often than not, it is “good enough” and the price of LDMOS becomes the main differentiator. So, what has changed?

Our research shows that the biggest bump for GaN in wireless infrastructure applications in 2014 came from the ongoing large LTE macrocell deployment effort in China. Chinese network operators are investing significant resources to establish a countrywide LTE network. The scope of these deployments gave GaN and LDMOS a big revenue boost. While the buzz in this segment is around the jump in GaN revenue, it is worth noting that Freescale and NXP, the two largest LDMOS providers, both pointed to challenges meeting demand for LDMOS in 2014. This may provide some opportunities for GaN to capture market share, but I think it points out that LDMOS is still the power technology of choice at many equipment manufacturers. The boost from China illustrates the old adage that a rising tide lifts all boats. This rising tide has given GaN a boost in the wireless infrastructure market segment that manufacturers of the technology hope to build on.

 We anticipate that increasing GaN adoption in wireless and CATV/broadband infrastructure applications will continue to push GaN revenue higher. In addition, early adoption of GaN RF devices in applications like microwave and millimeter wave radios for backhaul, commercial VSAT and the emerging industrial segment will add to this revenue in the future. However, we must not forget military applications. As the graph below shows, we estimate that revenue for military applications will grow faster than commercial applications and account for more than 50% of the GaN RF device revenue by 2016.

 

 

 The fast revenue growth for GaN devices in military applications reflects the growing trend toward more electronics content in military systems. Military budgets are growing slowly, but they remain large and the networks rely on improving electronics capability to perform increasingly sophisticated missions. This philosophy fits well with performance advantages of GaN. Because of this, we expect to see widespread adoption of GaN in radar, communications and electronic warfare (EW) applications.

 Revenue for commercial RF applications of GaN devices will continue to grow at a healthy rate, just not as fast as the military segment. Before China commenced their LTE deployments, revenue in the wireless infrastructure power market was declining. High power macrocell coverage was saturating, with most of the new deployments revolving around lower power “small cells” to backfill networks for data coverage. The Chinese LTE deployment looks like a short-term boost to the power market. As operators begin meeting their initial coverage goals, they will begin shifting new equipment deployment from macrocells to lower power base stations. This will slow the overall rate of growth of the infrastructure power market. Cost improvements and performance advantages will allow GaN to continue to capture market share from LDMOS, but the total market will see slow growth, at best. This will slow the overall growth of commercial GaN revenue.

I don’t want to end on a negative note, because the takeaway from all of this is positive! GaN has hit the jackpot, as revenue growth in military applications is strong and commercial adoption is finally beginning to fulfill its oft-discussed potential. This is a very dynamic market with a few potentially disruptive developments going on behind the scenes. I won’t address these in this blog, but I will tease that the future may hold some big changes. It will continue to be an interesting time in the GaN market, so stop by this blog and our webpage, or become a client, to get information on the latest developments with this technology and these market segments. 

- Eric

Previous Post: What does another "G" mean to the semiconductor industry? | Next Post: The Butterfly Effect; A Falling Snowflake Leads to a Massive Semiconductor Opportunity
Leave a comment