Reality bites: Google replaces Intel with Marvell for Google TV

Earlier today, The Wall Street Journal reported that Marvell has replaced Intel as the lead chipset supplier for Google TV. The deal is non-exclusive, but the bigger news is that Google has replaced Intel's X86 architecture with ARM, which is supported by Broadcom, nVidia, Samsung and Texas Instruments, along with Marvell and others. When the first Google TV devices were released by Logitech and Sony, it was clear that they were far too expensive for the market; for example, while Apple was selling Apple TV for $99 (U.S.) and Roku's set-top boxes were priced at $99 or less, the Logitech Revue was launched at $399. In order for Sony and Logitech to be competitive, they had to drastically cut prices and, in Logitech's case, take huge losses. (Logitech subsequently abandoned Google TV.)

By switching from the Intel architecture to ARM, Google TV's licensees will gain a less-expensive, lower-power platform that can compete with set-top boxes from Apple, Roku and others on both price and performance. They'll also get a choice of multiple processor vendors; for example, even though Marvell is the lead partner, there's nothing keeping Samsung from using its own ARM-based processors in its HDTVs, Blu-Ray players and set-top boxes.

In short, this is the move that Google should have made from the beginning. With lower-priced set-top boxes, the ability to run apps and an operating system based on a more modern version of Android, Google TV 2.0 should be significantly more successful than the original version. At the very least, it has a chance for survival, instead of being "dead on arrival".

Here comes your $99 eBook reader/tablet

Earlier this week, Marvell displayed engineering prototypes of "Moby", a tablet powered by its Armada processor that will run either Android or Windows Mobile. Moby will come will built-in WiFi, Bluetooth and GPS, the ability to encode and decode 1080p HD video in real time and 3D graphics support. The tablet will support Flash, and Marvell claims that end-user devices based on its reference design will be able to be sold for $99. As Technologizer pointed out, the tablets shown at this week's Future of Publishing Summit were rough engineering prototypes, so Marvell's claim that production versions ready for customer delivery will be ready by the end of the year is probably correct.

One interesting aspect is that Marvell is targeting the education eBook market with Moby; the company's press release is full of references to the educational market, and Marvell even plans to give Mobys away to every student at an at-risk school once the device goes into production. There have been companies targeting the low end of the eBook reader market, but $99 is an awfully low price point to beat. The advantages of full color, video and Flash will outweigh any disadvantages of using an LCD display instead of e-Ink for educational applications.

In any event, the tablet market is getting very crowded, very quickly, and Marvell is positioning itself to be one of the arms merchants.