Samsung's view of consumer electronics: part of the world in a picture frame

A recent discussion with Rich Yeh, a director of marketing at the System LSI division of Samsung brought up an interesting point. Discussing the markets for ARM-based, media-enabled SoCs, Yeh cited two application areas that would seem obvious: navigation systems (primarily the automotive ones, in which Samsung has a dominant position) and the emerging maybe-market of mobile internet devices. But Yeh also cited a third area that at first glance doesn’t appear that important: digital picture frames.

Now I think of digital picture frames as those $9.95 impulse-buy items hanging in their bubble packs at the check-out counters of discount stores. That doesn’t sound like a major consumer electronics application area. What Yeh had in mind, however, was something quite different: a high-resolution, maybe HD-quality display with its own graphics engine, video decompression hardware, quite possibly an open-source operating system, and, in the future, Internet connectivity. In a way, he is envisioning a cross between a photo kiosk and a small TV monitor.

But wait, there’s more. "We believe that WiFi will be a standard feature on these devices by 2011," Yeh says. So clearly this is more than a place to display a favorite frame from your $29.95 keyring digital camera. A quick Google search will indicate the range of products already out there, from vendors ranging from Kodak and Sony to the unheard-of.

Yeh apparently sees digital picture frames as becoming, in effect, an alternative to the home PC and to the large-screen HDTV for many households: in effect, a smart high-resolution monitor capable of showing the family all of their digital visual media, of picking images or video streams up from the Internet, and perhaps of becoming a home media gateway for families who aren’t comfortable with assigning this task to the cable company or to the PC, with the resulting responsibility of becoming de-facto network administrators.

It’s an interesting idea for the non-Web-infected rest of the world. And it has interesting implications for the SoC developer. For the chip developer, the digital picture frame—so conceived—would be the worst of both worlds. It would have to have the core horsepower and memory bandwidth to support at least Linux, a GUI, an open-source media stack, and an embedded browser. It would need the accelerators to handle jpeg and h.264, and the connectivity to handle streaming media through a WiFi port. And yet the device, by virtue of its back-door entry into the home media market, will face crushing price pressure—think bubble-pack at Wal-Mart, not the networking equipment isle at a computer store.

One important question is whether this can be done with a multi-chip solution. For now, Samsung appears to be staying with their multi-chip approach: a conventional-looking ARM-based SoC with accelerators on one die, WiFi interface on another, and of course external DRAM and Flash. Will the price pressure on a device trying to evolve out of a gadget market support this level of bill-of-materials cost? Or will the inside of the picture frame have to look more like an entry-level cell phone handset: with applications CPU, a lot of memory, baseband processing, and radio integrated onto a single die? If the latter, does this give a market opening to companies with RF integration experience?