Ronnie05's Blog

Intel tries hard, very hard catching up on Tablets and Smartphones (The Roadmap thru to 2014)

Posted in Computing and Operating Systems by Manas Ganguly on July 1, 2011

Intel is finally getting its roadmap in place to address the Tablet and Smartphone markets, where it has not really established itself. With Tablets heavily eating into the laptop category, it is time that Intel marshalls its considerable resources to address new markets. Its first system-on-a-chip for tablets and smartphones–codenamed Medfield–will be a crucial component of that strategy, though chips that follow may be more important commercially. Medfield will arrive in the first half of 2012, followed by Clover Trail technology in the second half of 2012

Medfield marks Intel’s move to a 32-nanometer system-on-a-chip Atom processor for tablets and smartphones. At long last leaving 45-nanometer Atom processors behind. Generally, the smaller the chip geometry, the faster and/or more power efficient the chip is. Medfield is likely a single-core processor and the system-on-a-chip Clover Trail variety a dual-core chip

Clover Trail chip would be a good match for Windows 8 tablets and convertibles and should be in time for the new Microsoft operating system is expected to be released to consumers sometime in the second half of 2012.
• Medfield: first half 2012, single core, 32-nanometer
• Clover Trail: second half 2012, dual-core, 32-nanometer
• Silvermont: 2013, new Atom architecture, 22-nanometer
• Airmont: 2014, 14-nanometer

The only thing here is that while Intel dicusses deal core, many of its competitors like Nvidia, Qualcomm, and Texas Instruments are already discussing quad-core chips based on the ARM design. Nvidia for sure is bringing forth Quad core chips for tablets and high-end “super phones” by end of this year. Qualcomm and Texas Instruments are shipping dual-core ARM-based designs into products now with plans for quad-core later in 2012. And Apple’s next-generation A6 is rumored to be quad-core too. All this actually puts Intel at least one generation behind its competitors.

Is Intel falling behind ARM in the war of Smart Computing?

Posted in New Technologies, The Technology Ecosystem by Manas Ganguly on May 14, 2010

Microprocessors and Intel is synonymous. I am the generation that has grown on the 236, Pentium I,II,III,IV generation of microprocessors and had been re-assured of the computing experience with the “Intel Inside” logo.

However, Intel’s story has been very different in mobile computing. It started late, was not fleet footed and hence let ARM run over with a near monopoly @ 90% market share and tie ups with giants such as Marvell, Nvidia, Qualcomm, Samsung, ST Ericsson, Texas Instruments and even Apple.

Intel’s reaction to ARM came late through the Atom series and platforms Menlow (its first attempt at a processor platform) and Moorestown (the latest one). Moorestown for all that its worth doesnot include design wins in high volume device categories and is thus unlikely to reach high volume products globally. Thus the appeal for Moorestown is rather limited and can at best be Intel’s “opening the door” or “beach-head” in the mobile devices computing space. It will possibly get the smartphone game started for Intel and cement relationships with tier one device manufacturers. The Moorestown platform will also be important from the perspective of joint development exercises such as Meego and Android based products.The next evolutionary step of Intel, called the Medfield will possibly be the best solution for the advanced mobility devices which thus will take Intel’s fight to ARM.

On the other hand, ARM is gaining from strength to strength. First a 90% Monopoly on Mobile and Handheld computing devices and then the partnerships with chip makers such as Nvidia, Qualcomm, Marvell, Samsung, Texas Instruments and Apple. That is some arsenal out there. Not to leave it alone at the markets and their partnerships, ARM is rapidly evolving its platforms to multiple core supports which thus enable very high computing power. This is something that Intel is missing in its arsenal sorely and there is no Intel product in its roadmap which takes the multi-core to higher levels. Thus ARM with its products and technology may exceed Atom based processors in raw computing terms by 2011.Intel thus needs a multi-core architecture plan for Atom processors within this year or risk being out-computed.

What it also means is that ARM will take the threat to Intel’s strong hold i.e Computing devices very effectively. So while Intel tries to enter ARMs market with Moorestown, ARM is planning to invade into Atoms home turf with multicore processors.

The future @ Intel appears to be hinged on its own software on its processors and its partnerships with larger manufacturers. Intel’s acquisition of Wind River (for software, middleware and embedded systems), Meego and Nokia Qt platforms are critical to Intel’s success even with a less advantageous Moorestown for the time being.

Last word: Intel seems to be going the Nokia- Microsoft way who failed to read the winds and were swayed away by the changed that ensued. Intel being Intel however, it will be interesting to watch how it goes from here.

%d bloggers like this: