August 16, 2004 10:32 AM PDT

Intel delays first TV chip

Intel won't be getting into television this year after all.

The Santa Clara, Calif., chipmaker said Monday it will not come out with a chip for making inexpensive large televisions this year and that it will rework the chip that has been in development for a later commercial release.

At the Consumer Electronics Show in January, Intel President Paul Otellini announced that the chip, code-named Cayley, would allow TV makers to come out with large-screen projection televisions that sell for under $1,800 by the end of the year. China's TCL had already committed to coming out with televisions based on Cayley, a liquid crystal on silicon, or LCOS, device.

"This will change large-screen television economics," Otellini said in January.

Intel was vague on why the chip is being reworked and delayed, but the company indicated it wasn't happy with the results of work to date.

"We are not going to bring a product to market this year," said spokeswoman Shannon Love. "We will go down a path that will give us clearer product differentiation and improved picture quality."

The setback is the latest in a series of delays for the company. Last month, such missteps prompted CEO Craig Barrett to issue an e-mail urging employees to do better. Among problems announced in July were delays in the release of 4GHz version of the Pentium 4 and of Alviso, a chipset for Pentium M notebooks. The chipmaker also announced that its flagship desktop processor, the 3.6GHz Pentium 4 560, was in very short supply.

The most recent announcement will probably buoy competitors such as Texas Instruments, which makes a competing chip called Digital Light Processor for companies such as Samsung. Samsung execs, as well as some analysts, have expressed doubts about Intel's plans since the beginning. The basic LCOS device--essentially a mirror that is controlled by semiconductors--has been around for years, but has not been deployed commercially with great success.

"It seems like the technology and business case for LCOS is still not there," said David Steel, vice president of Samsung's digital media business, in May.

Intel has also not had much success in branching out of the PC market. An effort to get Intel chips into cell phones has yet to seriously challenge Texas Instruments.

2 comments

Join the conversation!
Add your comment
When is Intel Going To Get It Right?
In a nutshell, When they begin to listen to the technical public. I cite their "cramming down our throats" of RAMBUS memory and most recently their decision NOT to make a 32/64 bit processor. Everybody knows they altered their course and wound up changing to DDR and they introduced a 32/64 processor but it wasn't quick enough.
Plain and simple, Intel has become so arrogant as to "know what we want and need" and that's exactly what we'll get...end of discussion.
But as long as Intel has Dell on a short leash they will continue to pop out failed projects. There is no real innovation occurring at Intel anymore and Intel (which is long-hand for Dell) has just become a commodity.
Keep it up Intel, and project after project will become only a stockholder's nightmare and a guaranteed change in management.

Fred Dunn
Posted by fred dunn (793 comments )
Reply Link Flag
Some things were correct
regarding Intel's arrogance - but I haven't seen much of that lately - it's hard to be arrogant when AMD is kicking one's butt, you know!

With regards to Dell being on a short leash, I'd like to know what kind of narcotic you must be on. Intel bends over backwards for Dell - and that is probably the only sole reason why Dell has stuck with Intel for such a long time.

The Rambus fiasco is a murkey subject, in my opinion. Here's why:
1) Parallel technology with memory architectures will eventually run out of steam. Notice how hard is it to bump up the clock speeds on memory these days? Serial has a much higher potential headroom, and the rise of Serial ATA is proof of this.
2) Any NEW technology will always have teething pains in terms of production and cost. Rambus had a third disability, RAMBUS, the company EVERYONE hates. The cost of licensing the technology was just prohibitively expensive, Rambus priced it out of the market.
3) Intel is a smart company, that realizes that you cannot make money selling faster CPUs by itself. PC architectures are holistic - YOU HAVE TO SEE THE WHOLE PICTURE. Intel has pushed and backed for new Bus architectures (PCI, PCI-Express), memory architectures (DRAM, SDRAM, etc. etc) and other advancements in the past so that the system will not be a bottle-neck to the CPU - to great success. Rambus happens to be a spectacular failure because of Rambus.

Admittedly, Intel is too conservative and stodgy at the present time. They have had EMT64 for a long time (even before AMD had it in their products - why do you think both chips have such a similar spec-sheet to them?). The problem with Intel, is that they have grown used to dictating the market speed, and now that AMD has caught up, they are scrambling to pull in their timetables.

As with the Northwood P4 and Hyper-Threading, The Prescott P4 already has EMT64 in it, but is currently disabled by Intel until their marketing department gives it a green-light. Stupid, if you asked me, but that seems to be the undoing of Intel.

With the fifth major mis-step this year, Intel is not longer (nor can they afford to be) arrogant. I think Craig Barret's letter is proof of that.

The only question left is if the new CEO, Paul Otellini will continue making the company a marketing-focused company, rather than a quality product focused one.
Posted by Tex Murphy PI (165 comments )
Link Flag
 

Join the conversation

Add your comment

The posting of advertisements, profanity, or personal attacks is prohibited. Click here to review our Terms of Use.

What's Hot

Discussions

Shared

RSS Feeds

Add headlines from CNET News to your homepage or feedreader.