CE3100: Intel squeezes yet another chip out of the old Dothan core

Santa Clara (CA) – During my most recent visit to Intel’s headquarters, I was able to catch up with Eric Kim, the company’s senior vice president of the digital home group and Intel’s third push into the consumer electronics market. This time, Intel focuses much more on its core strength than before and in fact tries to learn from the CE industry, rather than trying to change it.

The CE3100 Media processor is a critical product for Intel, as the company is branching out in another market that may be increasing the firm’s long-term chip sales. Similar to what the Atom series did for netbooks, Intel hopes that the CE3100 will conquer set-top boxes and digital TVs. And while this chip was announced back in August 2008, Kim just told us that it is based on the 90 nm Dothan core. Dothan was introduced in May 2004 as the second-generation Pentium M processor, which changed Intel’s course from a focus on Gigahertz to lower power consumption.  

Kim revealed that the CE3100 system-on-a-chip (SoC) is based just on this processor, but has received some upgrades, such as a DDR2 memory controller as well as a security and graphics engine and support for peripheral ports. In its core, it is, however, a 90 nm Dothan CPU. Kim said that Intel will transition the processor to a 45 nm Atom core in the second half of this year. Interestingly enough this is the same strategy that built Intel’s success in the netbook market: The A100/A110 processor used in UMPCs  in 2005 and 2006 were essentially underclocked 90 nm Dothan CPUs (600 and 800 MHz) and were replaced by the Atom processor in late 2006.
It isn’t Intel’s first move into consumer electronics. There was a first shy try in 2000 and 2001 with an electronic microscope and the Pocket Concert MP3 player, both of which were positioned to establish the Intel brand on CE retail shelves, but failed. Then we had Viiv in 2006, which was closer to Intel’s core knowledge of platform hardware, but this brand is virtually invisible on the market today as well. The CE3100 sticks to what Intel knows best – developing and selling chips to companies that integrate the CPU into their CE products. And this third try may be, in the end, the most conclusive way to success for Intel.

Some readers may remember that AMD has purchased a similar and actually flourishing business through its acquisition of ATI in 2006. However, the media processor business never worked for AMD and recently has been sold off to Broadcom. Intel’s approach is a bit different, as it involves a much more complex software platform, including support for browsers, Java, DRM and Flash. Despite some may claim that Intel made some progress in the CE industry in the past, it is clear that companies such as Intel and AMD will always have a difficult time to think the way a CE company does. That is where Kim comes in: He joined Intel from Samsung to get the company’s CE business on track and build a bridge to the Sony’s and Panasonic’s of this world.  
Kim confirmed that Intel has a lot of work to do in this respect and he believes it will be “important for the company to be humble” to understand the CE business.

To me, this sounds the most promising CE approach of Intel so far. It will be interesting to see if we all will buy TVs with x86 processors in the future.