« January 2011 | Main | March 2011 »

4 posts from February 2011

02/16/2011

How would you feel if your new Blu-ray player got a virus?

Probably not very happy.  This may not have been an issue when you got your first VHS player, but it is an unavoidable consequence of the perpetual, ubiquitous connectivity of today’s embedded systems.  Over 50% of respondents in VDC’s 2010 Software and System Development Survey indicated that they expected a project similar to their current one to include a web component within the next two years - a 20% increase over current project values.

Clearly, embedded system engineering no longer implies work on isolated, autonomous systems.  Embedded engineering teams now need to be cognizant of potential security threats, in addition to mitigating possible deterministic, memory, and/or processing requirements.

Today, Wind River, a leading embedded operating system vendor, and McAfee, a leading security solution provider, announced a formal collaboration to develop a tightly integrated security solution for non-PC (embedded) devices.  Although the announcement of this strategic partnership may not be a huge surprise since both companies are now subsidiaries of Intel, we believe that it addresses a growing gap in the marketplace.Security

Although our research indicates that a majority of engineers are at least acknowledging that security could be an issue for their current projects, the jury is still out on how best to ensure it.  Runtime solutions, such as that suggested through Wind River and MacAfee or even a standalone hypervisor used to create a “securely” partitioned guest OS, are one approach; however, we believe that it is equally important for security to be “designed into” the application through the judicious implementation of development and coding standards combined with the use of automated test solutions, such as static analysis tools, to minimize the amount of vulnerable code that could possibly be deployed within an end device.

One thing is for certain, the evolution of embedded and mobile systems is already causing many embedded engineering firms to reevaluate their existing solution sets. 

How has your company changed to address this growing issue?

Stay tuned - VDC will be covering embedded system security as part of our Safety and Security report from our 2011 Embedded Software and Tools Market Intelligence Service.

 

Related Posts:

02/15/2011

Embedded Engineers Experience with Multicore and/or Multiprocessing Designs

Suppliers, both software and hardware, continue to announce new products and services to support, in particular, multicore device/system development. As a matter of fact it is a keyword similar to Android, virtualization, hypervisors, security, etc. that are a must in any marketing campaign to improve SEO and grab attention. But how ready are embedded engineers to actually use multicore/multiprocessing in their development?

According to VDC’s 2010 Embedded Engineering Survey slightly less than half cited they have no experience with multicore/multiprocessing designs. No big surprise considering 62% of current projects being developed at the time of the survey cited single processor architectures. Also no big surprise that engineers developing mobile phones, consumer electronics, and telecom/datacom devices/systems cited the highest levels of current experience.

However, flash forward two years and the same engineers expect that almost 62% of projects under development at that time will require multicore and/or multiprocessing architectures. Quite a shift expected in a short period of time!

For suppliers, this migration will present challenges to many embedded engineers that will translate to commercial opportunities to support customers with system engineering services, training, etc.  Are you ready?

Multicore Experience 

02/14/2011

Enea Expands Multicore Strategy and Support for Linux

What Happened?

Enea recently announced a strategy to expand and strengthen their multicore support for heterogeneous multicore systems. The strategy includes an alliance with Timesys and membership in the Linux Foundation.

VDC’s View

Enea has long supported Linux as a solution for their customers and multicore, especially the telecom segment. We expect that for a time this support was under the radar, visible mostly to their customers looking to deploy both Linux and OSE-based embedded devices and systems.  Enea later established their Linux Competence Center (and subsequently its Android Competence Center) as a means to commercialize and offer services for Linux development and integration.

Enea’s support was agnostic to the various Linux distributions customers were looking to integrate whether it is Wind River’s, MontaVista’s or other commercially available distributions or publically available Linux distributions. Flash back to 2009 when we all know that both Wind River and MontaVista Software were acquired. For Enea, Linux is critical to their long term multicore strategy and a partnership with Timesys offers continued access to Linux through Timesys LinuxLink as well as through their Linux expertise.

This announcement continues to focus on a Linux approach by Enea where they look to complement their core offerings (both RTOS and worldwide services) by offering their customers a choice through partnerships rather than invest in assembling and supporting their own Linux distribution. For Enea it also continues to demonstrate the ability to deliver an integrated solution that includes runtime software (RTOS, Linux, hypervisor, etc.), middleware, and software development tools to address the challenges faced by engineering teams in fast moving multicore silicon architectures. Lastly, membership in the Linux Foundation will allow the company the opportunity to participate in the various working groups and technical projects and be part of the discussion and future direction of Linux as a means to ensure the representation of the telecom segment’s needs.

We expect that the continued evolution of the commercial Linux market to now focus more on services  and will cater to Enea’s core competencies – which should now be bolstered by Timesys’ experience and expertise. As with Enea’s previous ELPF, one of the key challenges will be for whether there is sufficient customer demand in Linux commercial products to fuel continued R&D investment. This may be the major obstacle to their partnership since Timesys has positioned itself as more of an aggregator of Linux builds then a market innovator.

Related Posts:

02/10/2011

We all know Android is Linux, but is it the New Linux?

Android has been all over the news lately, after having supposedly eclipsed the iPhone and approaching the Blackberry in share of smartphone shipments in the fourth quarter.  Although it’s arguable that the recent launch of the CDMA iPhone for Verizon might dent Android’s growth in the US mobile phone market, the Linux-based OS has enjoyed a meteoric-like rise in OEM and consumer adoption since it was first announced in 2007 – as demonstrated by the string of Android-focused announcements at last month’s CES as well as the anticipated product launches at the upcoming MWC.

In the same fashion as much of the other Android-related news in recent months, the alleged memo circulated from Nokia’s CEO Stephen Elop lit up industry news wires yesterday as he declared Nokia behind the times and stuck on “burning” platforms (Symbian and MeeGo).  While the underlying message conveyed by Elop may have been surprising to few, his candor certainly was.  It was only last year that MeeGo was proclaimed by the company as another viable open source contender to Android for mobile devices.  Now it appears as if it will suffer the same fate as LiPS and LiMo before it, unable to muster sustainable market interest and development momentum.

While Android has emerged as the defacto mobile Linux platform choice, it has only begun to make headway within the broader embedded systems market.  Yesterday’s announcement of the first open source port of the platform to the PowerPC (PPC) architecture, however, may accelerate its penetration into other embedded device classes. 

Android was, of course, initially developed as a consumer-facing, mobile phone platform.  As a result, it made perfect sense for Android to be developed for the ARM architecture, which is by far the predominant processor family used in the domain.  Although ARM’s penetration into other device classes has continued to accelerate over recent years, the landscape of the broader embedded market has (and will) remained much more heterogeneous due to the diverse nature of the different system requirements across different industries – from consumer electronics to aerospace to telecommunications infrastructure, etc.Android linux

Our 2010 embedded system engineering results, however, already showed Android’s use in a number of other application classes beyond mobile phones - from RFID readers to residential utility monitoring/control to automotive infotainment, among others.  Although real-time constraints will certainly limit the applicability of Android in certain applications (e.g. telecom infrastructure, military/aerospace, etc.), we expect that its porting to another widely used embedded architecture (yes, we know it is already ported for MIPS…) will begin to cannibalize the non-consumer-facing markets already utilizing Linux in greater frequency.  The investment, ecosystem and rate of innovation around Android are now far too compelling to OEMs looking to the benefits of a Linux open source solution.

Three years ago, we did expect that the evolution of open source offerings would continue to pressure the traditional ecosystems and business models around Linux in the mobile market.  I don’t think any of us expected that it would have occurred this quickly, that it would be completely dominated by Android, or that it would have had as profound an impact in other embedded markets. 

The bigger question now is will the non-Android Linux commercial market still be able to sustain itself or will the market revert itself to a new center of gravity around Android and a state similar to the early 2000s that was focused on a much smaller subset of mainline distributions, before the continued evolution of specific vertical market requirements led to the development of a myriad number of distributions targeted at the different industry-specific applications?

VDC is currently conducting its 2011 research on the embedded and mobile operating system landscape.  For more information about our upcoming 2011 Embedded System Engineering Survey or our Embedded/Real-time Operating System, Android and Linux in Embedded Systems, or Mobile Operating System reports, please click here.

 

Related Posts: