Samsung is typically seen as one of the rivals that has forced Sony’s television businessdeep into the red. Today’s remarks that the company is considering spinning off its LCD business, therefore, are something of a surprise. Sony ended its involvement in a joint LCD production venture with Samsung earlier this year, but the Korean company’s television business is still seen as being in much better overall health than Sony’s. That estimate, however, is strictly a matter of degree — Samsung’s display division lost $891M in 2011, compared to Sony’s $2.2B.
“We are internally reviewing various options but nothing has been decided yet,” Samsung spokesman James Chung said. There are rumors that Samsung might integrate its OLED and LCD divisions rather than selling off its LCD business, but the glory days of LCD production are seen as over. The chart below neatly captures why; it shows operating profit margins on LCD panels from 2003 through Q2 2011.
The volatile surges and troughs in margins are explained by computer OEM adoption, consumer uptake, the advent of HDTV broadcasting, the economic downturn of 2008, and the amount of time it took to ramp new production facilities as demand swelled. This last decline has been driven by a weak economic recovery and consumer indifference towards 3D television.
The problem for display manufacturers is that their ability to ramp new products is distinctly limited. Sharp’s so-called “Quattron” technology, which uses an RGBY color scheme, is useless because television signals aren’t encoded in an RGBY format. Higher resolution displays above 1080p are more difficult to manufacture and aren’t supported by any formal standard. Up-sampling is a theoretical workaround, but upsampled content will never look as good as native — and you need a huge screen or a very close viewing distance before the difference between 720P and 1080P are apparent, nevermind the gap between 1080P and a hypothetical 1440/2180P screen.
The other problem Samsung faces is the rise of mainland Chinese production. What’s happening in Korea and Japan now is extremely similar to what happened in the US in the 80s and 90s, when Japanese manufacturers put a number of US television companies out of business. Analysts are making noise about OLED as the possible savior of Samsung’s display business. Don’t believe it. We’re not going to rehash the entire OLED situation at the moment, but rest assured, if Samsung knew how to make OLED displays as cost-effectively as LCD televisions, it would do so. A slow, long-term, half-decade ramp on OLED capabilities might make the technology workable on large screens, but it’s not going to deliver the kind of meteoric turn-around that woos investors and makes highborn ladies fan themselves.
Samsung executives do have one thing going for them — they can pat themselves on the back and say: “Thank God we aren’t at Elpida.” That company faces “uncertainty” over its own ability to continue as a going concern.
Elpida, established in 1999, is Japan’s last major DRAM manufacturer. It was formed from the merger of NEC and Hitachi’s DRAM businesses, took over Matsushita’s DRAM division in 2003, and has struggled to remain profitable. The problem for any company in the DRAM business is that prices and demand are extremely volatile.
The chart above is from 2009, but it illustrates the point well. In August of 2008, before the beginning of the Great Recession, DRAM prices were above cost — but just barely. By December of 2008, six months later, it literally cost more to make DDR2 1Gb chips than they were selling for. The market didn’t return to August 2008′s contract price range until November of 2009. Manufacturers try to buffer themselves against such price shocks by introducing more expensive, higher-margin parts, but the consumer market often doesn’t cooperate. Even when it does, external shocks, like the hard drive floods in Thailand, can have a ruinous impact on paper-thin margins.
At this point, Japanese investors are betting that the government will choose to bail out Elpida and help the company meet its upcoming debt obligations. Earlier this year, the government agreed to back a joint spinoff venture between Renesas (NEC), Fujitsu, and Panasonic; the new arrangement will combine the SoC manufacturing and research divisions of the three companies as well as maintain their contract manufacturing capabilities. One of the major challenges such companies face is the advent of SoC designs; semiconductor manufacturers have historically relied on ASICs (application-specific integrated circuits) to drive growth and margins. FPGA’s supplanted, but never entirely replaced the ASIC business; SoC’s could finish the job.
These sorts of shifts are part of the same pattern that’s led AMD to move towards SoC design and pitted Intel against ARM manufacturers for the future of mobile products. As transistors become smaller and countries like China improve their home-grown manufacturing, the list of competitors is evolving rapidly. It’s a shakeout that’ll eventually produce a new semiconductor power base in mainland China, but the transition will be rough for the manufacturers in Japan and East Asia.
