Wednesday, January 27, 2016

China 2015 - 2016 Semiconductor Mergers, Acquisitions

Since 2015 there has been an accelerating trend of consolidation of semiconductor and semiconductor equipment companies. 

The trend is due to

1. The increasing cost of developing the next generation of semiconductor chips (see Feb 2013 blog SemiconductorMoore's Law Running out of Money )

2. Reduced growth rate of the semiconductor industry.

3. China’s increasing effort  to gain a strong presence in the semiconductor industry (see June 2012 blog Statusof China's Fabless Model and January 2015 Chinese$80 Billion Fabless IC )

The table below summarizes the key mergers and acquisitions of semiconductor companies announced since 2015. Before 2015 there was typically about one merger each year

In 2016 China will continue to get semiconductor presence by maybe buying portion of semiconductor companies such as Micron or Maxim directly or through a joint venture (see China Increasing IC Power )

So far, the mergers and acquisitions directly impacts wafer fabs capacity only for Altera that was bought by Intel. It is likely that China will try to increase its fabrication capacity.

Most of these deals have not yet gotten regulatory approval and could be cancelled before the deal closes. I will attempt to update the table as these transactions proceed.

March 2015
$ 12 Billion
Cypress Semiconductor/ Spansion
March 2015
$ 5 Billion
NXP RF/ Jianguang Asset Management
May 2015
$ 1.8 Billion
May 2015
$ 36.7 Billion
Intel/ Altera
June 2015
$16.7 Billion
Microchip/ Micrel
August 2015
$0.8 Billion
LAM/ KLA-Tencor
October 2015
$10.6 Billion
Western Digital/ SanDisk
October 2015
$19 Billion
Texas Instruments, Analog Devices, ? /Maxim
October 2015
$9+ Billion Market Cap
Microsemi/ PMC-Sierra
November 2015
$2.5 Billion
On Semiconductor or China Resources  / Fairchild Semiconductor
December 2015
$2.5 Billion
Dialog/ Microchip Technology / Atmel
January 2016
$3.6 Billion

Insightful, timely, and accurate semiconductor consulting.
Semiconductor information and news at -

Monday, January 18, 2016

Image Sensors and Cell phone Camera Performance

The article below demonstrates the importance of semiconductor and software developments on cell phone cameras. Semiconductor chip image sensor size, back side illumination (BSI), image signal processing, and gyroscope data processing improved image stabilization. 

Insightful, timely, and accurate semiconductor consulting.
Semiconductor information and news at -

Your Smartphone Camera Should Suck. Here’s Why It Doesn’t

SMARTPHONE CAMERAS ARE great, or at least close enough to great that you don’t notice the difference. We’ve reached the point where you’ve got to work pretty hard to find a phone with a mediocre camera, and when you do, it is an anachronism to be mocked and derided—and passed over for a phone with a better one.
It wasn’t always this way, of course. There was a time, not too long ago, when smartphone cameras sucked. They took genuinely bad photos that were underexposed or overexposed or grainy or … well, you remember. And if you don’t, consider yourself lucky. It’s taken a few years, but nowadays people take a great camera for granted. Thank companies like Nokia, which started pushing that envelope in 2007, and Apple, which gave the iPhone 4 the first camera that made people go, “Daaaaaaaamn.”
How did this happen? When you consider things like sensor size, pixel density, controls, and optics, smartphone cameras should be pretty lousy. Compared to a DSLR, they still are. But the camera in your pocket is crazy good considering the limitations manufacturers work under. And the advancements keep coming. As we look to the future, the cameras in our phones are only going to get better.
The Limits of Size
No matter what kind of camera you’re talking about, there’s a universal truth: the bigger the image sensor, the better the image. A bigger sensor will capture more detail with wider dynamic range (the detail in dark and light areas), offer superior low-light performance, and focus more sharply on moving objects. However, with few exceptions, smartphone cameras have tiny sensors
The vast majority of top-tier smartphones use Sony sensors for their main cameras and Samsung sensors for their front-facing selfie cameras. And every phone on DxOMark’s list of the 10 smartphones with the best image quality has a sensor size between 1/2.3 and 1/3 inches. In terms of surface area, the one-inch sensor in a nice point-and-shoot like Sony’s RX100 is more than six times bigger than any of the top smartphone camera sensors, while the sensor in a consumer DSLR is around 19 times bigger. Drop the cash for a pro-grade DSLR and the sensor is 50 times the size of that puny thing in your iPhone 6S.

This means smartphone sensors struggle to harness light—glance at a smartphone photo taken in low light and compare it to one shot with a DSLR. It’s no contest. But a little clever engineering has made smartphones better than they ought to be. “Backside illumination” or BSI, moves some wiring to the back of the sensor, maximizing the surface area upon which photons can hit the photosites. Another trick is using a 4 megapixel sensor with a 1/3-inch image sensor. Yes, this decreased overall resolution, but also pixel density, making them better performers in the dark.
But these are workarounds. Why not just use a bigger sensor to begin with? Because there are a lot of challenges to packing one into something so small as a phone, not the least of which is heat. “With a larger sensor that takes up more real estate, this leaves less room for heat dissipation,” says Dan Unger, a Panasonic spokesman. “Add to this the current demands of the larger sensor, and heat can be a real challenge to manage.”
You can get around that, as Panasonic did, by making the phone noticeably thicker. That improves thermal management. And it’s fine for a phone designed largely for photography, but it’s not an ideal solution. If smartphones are going to use bigger senses but not fill your pocket like a Stephenson novel, there’s a lot of work to do—especially when you consider bigger sensors cost more. And if shoot video, whoa does that generate some heat.
“When you add the movie into the overall equation, then you’re putting a lot more demand on the camera in terms of dealing with heat generation,” says Chuck Westfall, a technical adviser at Canon. “The larger the sensor, the greater the heat-generation possibilities become. In a small space like a compact camera, it works against you a little bit. It’s much more so for video than it is for still imaging.”
Imaging Is a Process
The sensor simply senses light and converts it into an electrical signal. To use an analogy, it buys the groceries. Someone else cooks dinner. So while a high-quality sensor helps, it’s hardly the most important component. The lens is important, of course, but the biggest difference between a great camera and a good camera is the image signal processor—the secret sauce to any smartphone camera’s features and performance.
Hung says that the image sensor isn’t the only thing feeding information into the ISPs. A modern smartphone has several sensors at its disposal. “The gyroscope has evolved in terms of image stabilization,” he says. “A lot of the ISPs now can take the input from the gyroscope (and) combine that input with the image sensor to provide image stabilization. It’s a new kind of digital stabilization system.”
Apple and Samsung use their own image signal processors for the iPhone and Galaxy phones, respectively. However, many high-end Android handsets use the integrated image signal processors in Qualcomm’s Snapdragon system-on-a-chip, which keeps camera features relatively consistent from phone to phone. As good as it is, the company says the next-gen processor arriving early in 2016 will improve noise reduction, artifact correction, autofocus, and color reproduction.
Optics Will Stay Simple
The molded plastic lens elements in many cameras have reached the point where they’re essentially perfect. They’re also cheap. Oh sure, critics argue they don’t have an optical zoom. There’s a reason for that. Optical zooms have moving parts, which runs counter to the phone industry’s slimmer uber alles mentality. You want an optical zoom? You’ll have to accept fatter phones.
“I think most users care more about a good-looking phone and image quality than perhaps that extra bit of functionality,” Hung says. “The people that care will get lens accessories to do those things. I don’t see those more advanced things being built into many phones.”
However, some patented technologies could hasten the arrival of optical zooms. Pretty much every point-and-shoot has an optical zoom as great as 5x with lenses housed entirely within the the camera. Hell, Canon’s patented a 45X zoom folded-optics lens, but has no plans to use it anytime soon.
“The overwhelming objective on a smartphone is to keep the physical size of the device to a minimum in terms of thickness,” Westfall says. “There is going to be a limitation no matter what in terms of the quality of the lens that they’re able to put in there. Not just in terms of resolution, but in terms of focal length range and aperture as well.”
The Future of Smartphone Cameras
Earlier this year, Apple bought the image-sensor companyLinX, which uses an array of lenses to enable Lytro-like refocusing, create 3-D depth maps, and improve image quality in low light. The Light L16 camera, which uses 16 lenses and sensors to recreate the surface area and low-light capabilities of a DSLR sensor, is also due next year. The company’s founders hinted that the L16’s multi-sensor technology could show up in smartphones before long.
But as good as they are, smartphone cameras probably won’t ever match the quality of a DSLR. And they probably won’t have to. For all but the most serious photographer, the ease of a smartphone camera, and the plethora of apps that can make a crappy photo look good, is plenty. And the two approaches are complimentary. Although the smartphones have decimated the point-and-shoot segment, sales of DSLR and other high-end rigs remain strong.

As long as that’s the case—as long as DSLR cameras take a better picture, you can bet the companies making sensors and processors for smartphone cameras will continue pushing the technology further.

Tuesday, January 12, 2016

Semiconductor Revenue Growth, Ranking

Worldwide semiconductor revenue declined 1.9% in 2015 (see article below) while fab capacity increased 6% .  Samsung and Hynix benefited from the iPhone 6s demand. 

The ranking will change in 2016 due to recent trouble of Toshiba and Western Digital buying SanDisk . As in past ranking SanDisk revenue is not included, which make this table inaccurate (see April 2012 blog Top 25 2011 Semiconductor Sales Ranking )

Insightful, timely, and accurate semiconductor consulting.
Semiconductor information and news at -

Gartner Says Worldwide Semiconductor Revenue Declined 1.9 Percent in 2015

Mixed Results in All Segments Drove Slow Growth
Worldwide semiconductor revenue totaled $333.7 billion in 2015, a 1.9 percent decrease from 2014 revenue of $340.3 billion, according to preliminary results by Gartner, Inc. The top 25 semiconductor vendors' combined revenue increased 0.2 percent, which was more than the overall industry's growth. The top 25 vendors accounted for 73.2 percent of total market revenue, up from 71.7 percent in 2014.
"Weakened demand for key electronic equipment, the continuing impact of the strong dollar in some regions and elevated inventory are to blame for the decline in the market in 2015," said Sergis Mushell, research director at Gartner. "In contrast to 2014, which saw revenue growth in all key device categories, 2015 saw mixed performance with optoelectronics, nonoptical sensors, analog and ASIC all reporting revenue growth while the rest of the market saw declines. Strongest growth was from the ASIC segment with growth of 2.4 percent due to demand from Apple, followed by analog and nonoptical sensors with 1.9 percent and 1.6 percent growth, respectively. Memory, the most volatile segment of the semiconductor industry, saw revenue decline by 0.6 percent, with DRAM experiencing negative growth and NAND flash experiencing growth."
Intel recorded a 1.2 percent revenue decline, due to falls in PC shipments (see Table 1). However, it retained the No. 1 market share position for the 24th year in a row with 15.5 percent market share. Samsung's memory business helped drive growth of 11.8 percent in 2015, and the company maintained the No. 2 spot with 11.6 percent market share.
Rank 2014
Rank 2015
2014 Revenue
2015 Estimated Revenue
2014-2015 Growth (%)
2015 Market Share (%)
Samsung Electronics
SK Hynix
Micron Technology
Texas Instruments
Infineon Technologies


"The rise of the U.S. dollar against a number of different currencies significantly impacted the total semiconductor market in 2015," said Mr. Mushell. "End equipment demand was weakened in regions where the local currency depreciated against the dollar. For example in the eurozone, the sales prices of mobile phones or PCs increased in local currency, as many of the components are priced in U.S. dollars. This resulted in buyers either delaying purchases or buying cheaper substitute products, resulting in lower semiconductor sales. Additionally, Gartner's semiconductor revenue statistics are based on U.S. dollars; thus, sharp depreciation of the Japanese yen shrinks the revenue and the market share of the Japanese semiconductor vendors when measured in U.S. dollars."
The NAND market continued to deteriorate throughout the year. As a result, revenue grew only 4.1 percent in 2015, fueled by elevated supply bit growth that resulted in an aggressive pricing environment. The tumultuous NAND pricing environment rippled through most of the NAND solutions, particularly solid-state drives (SSDs), which continue to encroach on hard-disk drives (HDDs). The ensuing price war in SSDs further pressured the profitability of the NAND flash makers amid the biggest technology transition in flash history — 3D NAND. While 3D NAND commercialization was modest, it was limited to only one vendor — Samsung. Modest revenue gains have not stopped investment in NAND flash and 3D technology, with all vendors continuing to spend aggressively in the technology and most with new fabs.
After 32.0 percent revenue growth in 2014, the DRAM market hit a downturn in 2015. An oversupply in the commodity portion of the market caused by weak PC demand led to severe declines in average selling prices (ASPs), and revenue contracted by 2.4 percent compared with 2014. The oversupply and the extent of ASP declines could have been significantly worse if Micron Technologies' bit growth had performed in line with its South Korean rivals. Fortunately for the market, the company saw negative bit growth due to its transition to 20 nm, sparing the industry from an even more severe downturn.
Additional information is provided in the Gartner report "Market Share Analysis: Semiconductors, Worldwide, Preliminary 2015 Estimates."