LONDON – Strength in mobile devices and weakness in PCs will characterize the coming year and drive global semiconductor market growth of 4.9 percent in 2013 after a flat 2012, according to market forecast firm International Data Corp. (Framingham, Mass.).
This growth – to $319 billion in 2013 – will come after the market expanded by less than 1 percent in 2012 to reach $304 billion, according to the firm. In the longer term the global semiconductor market will show a compound annual growth rate (CAGR) of 4.1 percent from 2011 to 2016 and reach $368 billion in 2016.
Weakness in PC demand with ongoing impacts on DRAM together with some inventory rationalization in the face of global macroeconomic uncertainty characterized 2012, in IDC's estimation. Smart phones and tablet computers, set-top boxes and automotive electronics were bright spots in the market and IDC expects that to be the case in 2013 and over the next few years.
IDC expects semiconductor inventories to come into balance with demand in the second quarter of 2013 with growth to resume in the second half of 2013.
IDC has broken down its semiconductor predictions by segment.
So for chips sold into the computing industry segment 2013 will show year-over-year growth of 1.7 percent, in line with a 2011 to 2016 CAGR of 1.7 percent. Semiconductor revenues from mobile PC demand will register 5.5 percent year-over-year growth in 2013, after declining 7.7 percent in 2012.
The market for chips for communications is set to expand by 6.5 percent in 2013 with a five-year CAGR of 5.5 percent. However, semiconductor revenues for 4G phones will experience annual growth of 140 percent in 2013 and a CAGR of 103.4 percent for 2011 to 2016.
Consumer growth, driven by tablets, e-readers and set-top boxes, will show 9.8 percent growth in 2013 and fiver-year CAGR of 6.0 percent.
Hi Bruce (and Peter) - contrary to your belief on the difficulty in making sense of the various sales growth expectations I "put out," I feel that the model I have developed is, in fact, straight forward and provides each month a good proxy of what the sales growth might be each month as the year plays out. I should point out that Linear Regression Analysis that the model employs operating on historical monthly sales history of the industry tracked and published by the WSTS going back over the past 28 years is pretty rigorous on a relative basis. Since the model is purely mathematical no assumptions and/or emotion are invoked in arriving at the forecast results.
Bruce - for more details on the "workings" of the forecasting approach please provide me your e-mail address (@ mikedcowan(at)verizon(dot)net and I will send you my monthly report with all the details re. the model's approach and latest results.
I agree with you, predicting the future of the semi industry is complex and chaotic. The graph showing the retreat of the forecasters from rosy 2012 predictions is stunning. However, if you looked at the North American Semi Equip Book to Bill and global semi sales in May of 2012 it was obvious we were pointed in the wrong direction, yet there was a lag before the models responded. I know these are old school benchmarks and everyone is trying to anticipate the turning points. I am curious what general trend Objective Analysis uses to "be right".
Nice spot Mike
Of course, the article serves Objective Analysis well. Why else would Jim Handy contribute it?
But it does show how analysts struggle with the complex, chaotic nature of forecasting markets.
Kris - check out the graphic put together by Jim Handy at the following URL = http://www.forbes.com/sites/jimhandy/2012/12/28/a-review-of-semiconductor-forecasts/.
Come the first week in February the final (actual) 2012 sales number will be published by the WSTS and you can check out yourself how the various prognosticators portrayed in the above invoked graphic "faired" relative to the accuracy of their respective predictions throughout 2012.
Hi Peter - the latest run of the Cowan LRA forecasting model is more BEARISH on 2012's predicted sales growth outcome than IDC's prediction and, correspondingly, more BULLISH on what the model sees for 2013's sales growth expectations. In particular, the model is projecting a 2.3 percent DECLINE in global semiconductor sales for 2012 compared to 2011's actual sales (per the WSTS) of $299.52 billion. For 2013 the model's sales growth forecast estimate is expected to register an INCREASE of 6.4 percent over 2012's projected sales forecast estimate of $292.76 billion. Both years' sales growth predictions put forth by the model are gleamed from its “operating on” the WSTS's HBR (Historical Billings Report) sales numbers through October 2012 which were posted by the WSTS at the beginning of December. Final year-end sales numbers for 2012 as gathered and tracked by the WSTS will be published on their website at the beginning of February 2013.
Therefore, stay tuned for the "wrap up" of 2012 sales results as well as updated Cowan LRA Model forecasts for 2013's outlook each month as the New Year "plays out."
Happy New Year to ALL.
Mike Cowan (independent market watcher and developer of the Cowan LRA Model for forecasting global semiconductor sales)
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.