NEW YORK – As Taiwanese IC suppliers like MediaTek and Mstar continue to gobble up growing chip demand among China’s consumer electronics OEMs and ODMs, the set-top box and digital TV businesses don’t exactly look exciting (or lucrative) to the semiconductor market segment any longer – at least to most chip suppliers in the West.
Look no further than struggling financials of Trident Microsystems and Zoran Corp. in recent quarters.
But of course, for Broadcom, this is a very different story. Geneva-based STMicroelectronics, keeping up its competitive position against Broadcom, also sees an upside for digital TV – including Interne-connected TVs and 3-D TVs.
During the company’s Investor and Analyst Day Thursday (May 19th), DTV was among a few specific products identified by STMicroelectronics CEO Carlo Bozzoti as keys to “the top-line evolution” of the company’s automotive, consumer, computer & communications infrastructure (ACCI) business. Others singled out by Bozzoti as driving forces for growth and profitability include: ST’s 32-bit microcontroller (based on Power architecture) for automotives; ST-developed royalty-free displayport, and computer peripheral chips.
Actions taken by ST for its turnaround efforts over the past few years include a review of the product portfolio, a focused approach on R&D and manufacturing restructuring, according to Bozzoti.
While the company’s digital convergence mantra sounds almost outdated (everyone has been saying it for the last 15 years), one thing ST has done well is to execute the genuine “digital convergence” of IPs and SoC blocks developed by different groups of the company – in particular, ST-Ericsson’s apps processor development team and ST’s home entertainment and display (HED) group.
Switching from ST40 to ARM
The HED group, headed up Philippe Lambinet, ST’s senior executive vice president, is switching its processing core from ST40 (based on SH4 core originally developed Hitachi) to ARM in its Gen. 4 platform designed for open Internet-connected TV. While the Gen 4 is scheduled for 2012 rollout, the bulk of this core work of porting to ARM, in fact, has been already carried out by ST-Ericsson, not by Lambinet’s team.
ST-Ericsson, a couple of years ahead of ST’s HED group, ported ARM core to its ASIC library. ST-Ericsson team mapped its high-level description, trimmed it and optimized it to a high-performance low-power process, thus hardening the core, explained Gilles Delfassy, president and CEO of ST-Ericsson. “We are sharing that with ST,” he added. ST-Ericsson, meanwhile, didn’t have to develop its own digital video or audio codecs, as they got those well-proven cores from ST’s HED group.
This convergence offers a huge payoff to ST. It can rely on field-proven IPs and shorten the development cycle, while eliminating the duplication of engineering resources originally needed to optimize cores or develop specific blocks.
This move by ST, right off Broadcom’s playbook, demonstrates that if a company is disciplined enough, being a large chip vendor with a diversified product portfolio won’t work against it, but actually helps it.
Lambinet revealed that his group’s Gen 4 chip, named “Newman,” for Internet-connected TV, is adopting a block used in an ST-Ericsson chip based on dual-core 1.2GHz ARM Cortex-A9, now designed in by Samsung’s Galaxy S II.
Other notable blocks integrated in ST’s Gen 4 Newman chip include: quad-cores of Mali graphic engines; Motion-Corrected Temporal Interpolation and De-interlace blocks – ST gained through the acquisition of Genesis Microchip in 2007. The chip also features 3-D video, 2-D to 3-D conversion capabilities.
'Connected TV is a set-top box with a big screen'
Asked by an analyst if there’s still money to be made as a chip supplier for TV OEMs, Lambinet answered a resounding yes.
Bozzoti noted that the set-top box is still a great business today, but its growth potential may have passed. While Lambinet agreed on this, he emphasized that ST’s strong heritage in set-top box expertise can only help strengthen the company’s Internet-connected TV solutions.
“After all, connected TV is a set-top box with a big screen,” he said. “And that’s why we think both ST and Broadcom will do well in the growing Internet-connected TV market, whereas a company like Trident with much less history in set-top will struggle.”
Lambinet also noted that in the TV world, the captive market still remains strong as CE giants like Samsung, Panasonic, Sony and Toshiba continue to design and build their own TV chips.
While Samsung isn’t likely to give up the development of their own TV chips, TV is a sector that will eventually open to the merchant market, predicted Lambinet. In some cases where vendors are severely hit by the recent earthquake in Japan and can’t get their plants restarted, they are beginning to turn to ODMs. “The move to the merchant market has already started to happen in some ways,” said Lambinet.
Bert22306 your comment is right in the money with the exception in my opinion, that a large portion of broadcasting video consumers like the flexibility of a thin-client box, but are reluctant tech adopters.
Unless a GUI is intuitive and easy to navigate, there always going to be a segment that prefers simplicity of legacy boxes.
Well, perhaps not Internet TV's, but at least the STB portion. The STB is classically been a shim between the delivery method/format, and the "dumb" TV. As it becomes more cost effective to migrate the smarts, menus, features, video decoding, mpeg, digital demodulation, into an SoC solution in the smarter TV, the market for the external STB will continue to shrink. Of course there will continue to be a market as a technology shim for legacy TV's.
Developing markets can leap-frog to newer, more cost effective, transmission standards for terrestrial or satellite. Just look at the explosion of wireless telephony, vs the infrastructure costs of deploying wired solutions.
Have to agree with you there Larry. The government needs to wield a big stick and get this stupidity in check.
These SAME companies have been bleating for decades about the promise of *interactive* TV. Unfortunately the model people like, is the one *without* the interference and meddling, the cable/telco companies love so much.
After selling us a bill of goods for Internet delivery, they want to cherry pick what content they actually want to carry, price some bits differently, and despite charging at both ends, they want more. And yet, they continue to resist the idea that consumers might want to cherry pick "a-la-carte" what content/services they wish to purchase from them.
I have to agree with Larry. I am hearing an increasing push for bandwidth time and speed limitations and usage fees. If these are utilized to invest in the future web infrastructure, that's one thing. But if used to stifle technology progress and competition, that's not a desirable route to go.
Actually, you woud be surprised to hear that some do predict Internet-connected TVs may show up in BRIC sooner than digital TVs (because a lot of countries have no mandate to phase out analog signals...)
Does the Internet TV's are going to replace the set top box and other regular TV broadcasting technologies. I don't think it will happen sooner than 5 years in the developed nations. And in the BRIC countries it would take at least 10 more years due to lack of high speed internet access required for such internet TV.
The spoiler for the wave of Internet-connected TVs and other devices is the toll booth being set up by the ISPs. This is a textbook example of how monopolies deter innovation, and unless something is done at the government level progress in all of these IP-based services will slow or stall.
Why do you see MStar being stuck in Asia? Is the penetration into Western/US market being prevented by things like patents, IP, and MPEG-LA issues?
Personally I don't see Trident as being viable moving forward. They took the entire NXP STB team and pretty much canned them all. Such pools of talent will likely materialize as competition elsewhere.
I have a Vizio 22" TV with the Apps+Ethernet+WiFi, not that impressed, but it is likely that the entire STB market is going to seriously struggle with the commodification of the space with this functionality being dumped directly into the TV. The market is thick with streaming boxes, including those from big names like Seagate, Western-Digital, and SONY. A SONY SMPN100 box with 802.11n+Ethernet, 1080p, which will network with BD players at home, has a street price of around $88, and is very slick compared to the Vizio implementation.
Then there is GoogleTV, and playing for other Android devices. I don't see this as a high margin space moving forward. The likes of Intel are going to have to adjust their margins if they even want to play in this space.
NetFlix and RedBox would seem be bigger winners in the current game.
I have always seen set-top boxes as an interim solution anyway. I mean, the plain-jane kind, not things like DVD players.
It makes all the sense in the world to build in the required reception hardware and software, once it has become well established and relatively stable. The cable companies have been blocking such built-in solutions only because it was more money in their pockets to rent out the proprietary STBs, month by month. And people caved in like lemmings.
In the Internet case, my current "STB" is simply a nice, slim, very quiet, dedicated PC, with RF remote keyboard and mouse. Total flexibility, beholden to no single service other than a broadband link.
My suggestion to these CE manufacturers is, build in what amounts to standard PCs, even if you limit these to mostly a web browser and Flash player. What some might call, a "thin client." Give the users the ability to set up their own "favorites." Try, oh please try, to avoid the teptation to collude with unnecessary third parties.
That's a damn good point.
In the end, all these smartness -- required in set-top boxes -- are quickly moving into TVs and non-traditional set-tops. I actually loved when Lambinet said, "Connected TV is a set-top box with a big screen"!
Join our online Radio Show on Friday 11th July starting at 2:00pm Eastern, when EETimes editor of all things fun and interesting, Max Maxfield, and embedded systems expert, Jack Ganssle, will debate as to just what is, and is not, and embedded system.