Intel Gives Up On TV 89
symbolset writes "Bloomberg is reporting that Intel, on the cusp of having low-power embedded chips that can do true HD in a flatscreen, has given up on getting its chips embedded in TVs. While many might say their efforts to date have been fruitless because of energy issues, Medfield might have had a chance in this field."
Translation, please? (Score:4, Interesting)
Intel has been unable to provide a chip that offered significantly different performance from rival offerings, and failed to convince TV makers such as Samsung Electronics Co. or Sony Corp. that they needed its chips, Acree said.
OK, geeky people, what does that mean?
I interpret it as "producing chips for TVs is a commodity business and there's little opportunity to introduce anything new." Was Intel just late to the TV chip party and other chipmakers had it sewn up?
I would think even as a commodity producer, Intel would be competitive just because they have huge scale.
Intel chips. In TVs! (Score:4, Interesting)
I remember a TV many years ago, perhaps late 90's or early 2k, which booted with a common Award BIOS screen and RAM check. I think we sold exactly one (and that one was the display model).
It was a useless device. Despite having a high-res CRT display with decent color, and a line doubler (which was potentially way cool in those pre-HDTV/DVI/HDMI times), it sucked: It irrevocably upscaled the output of a PSX, and the result was double-ugly instead of double-smooth since it got the field order precisely wrong.
It had an Intel CPU.
Is it dead now?
Good.
Thanks!
[/shallow]
Re:Translation, please? (Score:3, Interesting)
Keep in mind up until 2000-something Intel was not only one of the largest chip manufacturers, but also one of the largest manufacturers of embedded controllers in the world. Some MBA dickhead under Otellini or his predecessor (the dude who fucked them to rambus for the first year of the P4's life) decided that embedded wasn't a high enough 'profit' division to hold onto and either sold it off or spun it down. Point? Intel already dominated that market many years ago, but due to trends in management are too shortsighted to retain the sort of market diversity they need to help them roll out emerging products like this into the sectors of the market that will leverage them and make them a steadily profitable while not ridiculously so market.
This is the same thing that happened to the i740 successors and larrabee (both of which sucked engineering-wise, but whose basic premise should've been kept: Intel could've been in the videocard market where nVidia/ATI are now. But some shortsighted manager decided it made more sense to cut their losses than to persevere and get the product right so the next generation and the one after that would succeed.
Re:Translation, please? (Score:5, Interesting)
Keep in mind up until 2000-something Intel was not only one of the largest chip manufacturers, but also one of the largest manufacturers of embedded controllers in the world. Some MBA dickhead under Otellini or his predecessor (the dude who fucked them to rambus for the first year of the P4's life) decided that embedded wasn't a high enough 'profit' division to hold onto and either sold it off or spun it down.
I worked for Intel during that period. Management was totally poisoned by the dot com disease. You could have a business plan that called for spending $50 million over five years to create a guaranteed $150 million a year product line with 25-40% margins and they didn't want to know. They were only interested in stuff that supposedly was going to produce a $500 million business a year in 18 months. They spent vast sums of money on second string chip companies, some of whom were already in trouble before the bottom fell out.
The other thing, they have this focus on margins that is deranged in that it's a straight percentage target that isn't adjusted depending on the market. In some markets, yes 60% is needed because you need to reinvest constantly in new designs. But there are other markets where 20% is more typical. Markets where the product life cycle is 36 months not 12. So they give up on stuff when the margins aren't their, forgetting that the proper metric is return on investment. Case, say you have a business, makes 150 mil in revenue, 20% margin, is 30 mil. If you have to invest 10 million a year in product design and what not to keep that business. Then you make 20 mil year off a 10 mil investment. 100% profit. But the way Intel sees it' it's only 20%, not enough to waste time on.
Then the dot bomb happened and they tossed overboard everything that wasn't going to turn a profit in 12 months. They also stopped development on product lines, thus killing them over the medium term. Of the dozen or so companies they bought 1998-2001 they closed all but one, and that only because being 12 and O would have been too embarrassing.
My impression is that Intel has a lot of capable people, and money to hire same. But the upper management has issues. It's like when they enter a business and find the other players are determined and competitive, customers who are used to wheeling an dealing to get the best value out of their vendors; management gets pissy and shuts everything down instead of sticking it out long enough to crave out some market share.
As a former TV Product Planner (although (Score:3, Interesting)
not for Google TV).
Intel chips are expensive and these days you would be very much be expecting a highly integrated chip with demuxes and decoders for digital broadcasts, video and audio processing elements to improve the quality. There would typically be a whole bunch of functional units for most functions all baked onto the silicon. The General Purpose Processor would typically be fairly weak but with a lot of support. Main processors may get somewhat more powerful to support browser type technology but I wouldn't expect them to reach Intel Atom speeds in most cases for some time. Which would you rather have, a TV with a fast web browser or good picture processing?
The current Sony Google TVs (the integrated screens) still carry the same main chip as the rest of the Sony range in addition to the Intel processor and graphics. I'm not certain of the extent to which this is absolutely technically required or whether it was needed to use the existing TV reception and processing software. This means that the cost of the to build Google TV was like building a normal TV and adding a bare bones Atom PC. Expectation of pure additional sales, marketing funds from Intel and an expectation of smaller margins for retailers were what made the business case I understand although I think there were also some unreasonable assumptions particularly if you had ever tried the product.
http://techon.nikkeibp.co.jp/english/NEWS_EN/20101117/187451/ [nikkeibp.co.jp]
http://www2.renesas.com/digital_av/en/mpegdec_tv/emma3tl2.html [renesas.com]
If Intel do back away from the highly cost sensitive TV chip business I would expect Google to offer support for ARM. I think most of the TV manufacturers on or moving to ARM although MIPS was is certainly used in current models. The newer high performance ARM chips are a probably significantly more expensive than the typical TV processors but probably make more sense than the Intel Atoms with the ability to custom specify the chip features and still be cheaper.
Features on such chips will be specified by major manufacturers but the feature set will probably be locked down at least 18-24 months before the TV ships ruling out some things after that date.
The TV business is hugely competitive market and there is no profit in it (possibly with the exception of companies that have their own panel manufacturing). The combination of falling prices, long parts lead times and the importance of volume to get component prices make it a very tricky business to make money in. But it is key to many companies positions in the Consumer Electronics area and can bring leverage into other businesses (by enabling retail space, offering full product suites and increasingly giving scale to over the top online video offerings.).