MOUNTAIN VIEW, Calif--In a bid to tighten and hone its mobile technology focus, Intel Corp. has confirmed it has reorganized several business units to form a new mobile and communications group which will be responsible for all the firm’s smartphone, tablet and wireless communication efforts.
The reorganization fuses together the mobile communications, netbook/tablet, mobile wireless and ultra-mobility business units into one group headed by former Palm executive Mike Bell and former Infineon Wireless chief Hermann Eul.
A spokesperson said Intel was looking to “speed up and improve development of Intel-based mobile devices,” from streamlining the software development efforts for tablets and phones to fine tuning the SoCs for mobile usage.
Back in March 2011, Intel announced that Anand Chandraskher, GM of the former ultra-mobility group, would be leaving the company to "pursue other interests," with many analysts positing that he had been ousted for failing to deliver on Intel’s long held mobile aspirations.
An Intel spokesman said Bell and Eul’s role would be bigger than Chandraskher’s, with the pair boasting numerous years of wireless industry experience between them.
Though Palm was never a big player in the mobile market, Bell also has previous experience at Apple Inc. where he worked on the Macintosh and the early iPhone model. Eul, on the other hand, will be instrumental to helping Intel integrate wireless baseband onto its Atom chips, putting the firm in direct competition with rival Qualcomm.
It’s generally believed that Atom products with integrated baseband could emerge within the next two years.
Intel has said publicly it hopes to have a viable smartphone offering on the market within the first half of 2012, and has made much of its partnership with Google for Android x86 development. Whether Intel’s engineers are able to reduce Atom power levels significantly enough to compete seriously with ARM’s chip architecture, however, remains to be seen.
But everyone would agree that Intel has missed the boat for the mobile and tablet segment. So, instead of chasing behind the boat Intel should come up with something more than just power efficient chips.
Seems like folks are missing another important distinction, at least in the short run, between true mobile products (smartphones and tablets) vs. the PC ilk (laptops, netbooks and desktops). There are two key elements to drastically reducing energy needs to reach the battery life goals for mobile
- a power-sipping app processor - where today ARM still has a dramatic lead on both the dynamic and static/leakage side.
- a power sipping memory system, which today means a small (500M-1GB note 32bit addressing is sufficient) SRAM system with small energy costs for each memory transaction (apps processor and memory packaged together).
The upshot is that this also requires a specialized OS with power awareness, plus an extremely small footprint compared to PC land. So if Intel wants to make real headway in mobile in the short term, they have to succeed in a very different landscape, than PCland. A PC that uses "less power", is structurally handicapped by a 5x factor in terms of energy and battery life vs a comparable "small memory" mobile system. They have to stop producing "scaled down" hammers (Atoms) and move to making screwdrivers.
And for all the folks who call for PC "backward compatibility" in the mobile space, please realize that that is the root of the power problem - a big memory system is anathema to a mobile device today, something even Microsoft has recognized for Windows 8.
In the same way that ARM's obsession with LACK of backwards compatibility is the problem? The two platforms are rapidly converging. ARM basically sucks for anything but basic functionality but sips power. This has been OK for relatively dumb phones, but as the demands in performance and capabilities are coming up, so does the power consumption. Intel is moving in the opposite direction. Sooner or later the two are at the same place, and keep their respective markets due to incumbency.
On a somewhat related note, can anyone explain why Intel wants to sell mobile chips that NOBODY makes money on? Protecting their market from below?
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.