Next generation processor technology and embedded computing
What's under the hood of CPUs newly arrived on the market and looking straight at embedded apps?
Jim updates us on the diverse range of embedded board and system hardware platforms utilizing the 2nd generation Intel Core processor family and how this new technology can help meet the processing, platform security, video, graphics, and power consumption requirements of today’s industrial computer systems.
We hear the message so often about how the latest and greatest processor and chipset are going to make our collective lives easier that we frequently tune out what’s being presented on specific CPU technology advancements. Our attitude could be due to a jaded view of technology, but more than likely this situation is caused by information overload. We’re all so busy that we often don’t have time to digest the real meaning of a processor technology advancement and its practical meaning for embedded computing applications. In this article we will explore the technical merits of what Intel is officially calling the “2nd generation Intel® Core™ processor family.” (See Figure 1.) Intel developed these next generation processors under the codename “Sandy Bridge.” The company introduced the first series of processors and chipsets (formerly known as Cougar Point) earlier this year. We will examine what is under the hood of these new CPUs and see how some of the features and basic processor design changes can enhance embedded computing system solutions.
There has been a lot of talk regarding the “2nd generation Intel® Core™ processor family.” Some of the information out there is pure hype. However, major design changes to the processor in the Sandy Bridge architecture should vastly improve the single board computer, embedded motherboard, and industrial computer system performance, power efficiencies, and platform security. Industrial computing solutions deployed in Mil-COTS applications, medical imaging, and industrial automation systems are well suited to take full advantage of the many Sandy Bridge micro-architecture features. There are a variety of new embedded computing boards on the market or about to be introduced that incorporate the new generation of processors and chipsets. We will examine how board factors such as SHB Express® (PICMG® 1.3), COM Express® (Figure 2), CompactPCI®, and OpenVPX® are taking advantage of the features of the 2nd generation Intel® Core™ processor family in a variety of embedded computing system designs.
Processor and chipset technology overview
There are multiple versions of the 2nd generation Intel® Core™ processors. Some are available now, and the schedule calls for some to roll out over the next 12 months. One thing to keep in mind about this new generation technology is the variety of different processors and chipsets that are available, each with unique attributes that can enhance specific embedded computing solutions. We need to have a basic understanding of the various versions of the processor and chipset available with this 2nd generation architecture because some features come at an application cost. For example, the Intel® Core™ i7-2610UE and Intel® Core™ i7-2655LE may be best suited to small form factor boards such as processor AMCs and EPIC motherboards because of their 25 W and 17 W TDP ratings, but these lower power ratings come at a cost of lower CPU frequencies, fewer cores, and less cache. If these factors are critical for your application, you need to consider that during the system design phase of your project. At the other end of the processor spectrum, the desktop Intel® Core™ i7-2600, and some future Intel® Xeon® advanced server versions, have base core speeds over 3.4 GHz, four processing cores with Intel® Hyper-Threading, as well as tremendous graphics and platform security capabilities, but these features come at a cost of TDP ratings approaching 95 W. Again, boards such as the SHB Express (PICMG 1.3) single board computer shown in Figure 3 and MicroATX motherboards that use these processor versions tend to be in larger rackmount and shelfmount embedded computing solutions that are designed to handle high-performance application requirements. Here’s a list of the base features common to all flavors of the 2nd generation Intel® Core™ processor family.
Processor die reordering and efficiency improvements
Several processing sections on the die have been reordered to more tightly integrate the memory interface, processing cores, and the traditional Northbridge functions. This reordering creates additional bandwidth while reducing latency. Efficiency upgrades related to how the processor accesses memory cache enable system improvements without necessarily generating more on-die heat. The decoded micro operations can now be buffered on the processor die and the CPU’s execution units can handle two processor loads in parallel. The net result of these changes to the 2nd generation Intel® Core™ processor architecture is a marked increase in overall system performance.
32nm process technology
This new processor technology packs twice as many transistors on the CPU die as compared to previous generation Intel processors. While this leads to improved processing capability, it can also lead to a significant increase in on-die heat generation. Several new power-handling improvements overcome this basic law of physics to produce some excellent performance-per-watt system test results.
Processor power consumption reduction
The 2nd generation Intel® Core™ processors have a new power control unit on the CPU die that redistributes power to different cores, including the new graphics core, based on the system’s computational processing and graphical demands. Intel also redesigned the CPU’s scheduler code to reduce power consumption while performing additional instruction execution.
Single component chipset design
The term “chipset” is a bit of a misnomer these days given that many of the memory interface and system interfaces reside in the processor die. Residence in the processor die enabled the previous generation processors to combine the two-component Northbridge and Southbridge chipset into one low-power component called the “Platform Controller Hub.” This single component, low-power “chipset” approach is also used in the latest PCH versions to enhance overall power savings in new system designs. The PCHs available today for use with the 2nd generation Intel® Core™ processors are the Intel Q67 and QM67. Later this quarter the Intel C206 will be available for use with the advanced server versions of the CPUs.
Additional processor features
The following list of features is available on most versions of the 2nd generation Intel® Core™ processor:
- Enhanced graphics controller supports multiple video interfaces.
- Intel® Advanced Vector Extension (AVX) doubles vector width, which should provide a 2x performance increase in many applications.
- 256-bit extension to SSE is also available.
- System platform security and remote management capabilities have been increased with Intel® AMT 7.0, Intel® VT and Intel® TXT.
- ECC memory support is now included in the advanced server versions of the 2nd generation Intel® Core™ processor.
- More available PCIe 2.0/1.1 links available out of the processor and the PCH.
Putting technology to work
So now you ask something like, “Great, now why do I care?” or more precisely, “How does this new processor technology help me in my embedded computing system applications?” Both are valid questions, and the simple answer in some applications is that your specific system requirements may not need the latest processor technology. If your current system solution is running just fine, then leave it alone. However, application software, platform security, network interfacing, and video/graphics demands are constantly placing more of a strain on the embedded board’s available resources. And of course everybody wants to make more efficient use of power in his or her system designs. As we saw in the previous section all of these concerns are addressed to one degree or another by the 2nd generation Intel® Core™ processor family.
Now the question might be rephrased as, “Is there an embedded single board computer board form factor available with these new processors for my specific system design?” There are many computing system form factors available today, all demanding embedded boards that optimally match the application. The good news is that there are number of different board form factors (3U CompactPCI for example – see Figure 4) available from a variety of vendors that implement the 2nd generation Intel® Core™ processor and related platform controller hub. Some boards support lower-end/low-power processor versions. Others support higher-end CPUs that provide more functionality while consuming more power.
The size, shape, and industry specification adherence run the gamut from small COM Express modules and OpenVPX cards (Figure 5) up to SHB Express single board computers and MicroATX motherboards.
Table 1 lists some of the available embedded boards currently on the market that support the 2nd generation Intel® Core™ processor.
It’s clear that the Sandy Bridge processor architecture has some compelling features and capabilities for embedded computing applications. The ability to support multiple video and graphics displays, extensive PCI Express interfaces that connect directly to the processor, power savings, and data throughput increases implemented in this new CPU architecture all bode well for future industrial computer system designs. Many different single board computer form factors are available to drop into a wide variety of system designs. Applications in Government & Defense, Communications, Medical, and Industrial Automation will all benefit from this quantum leap in processor and embedded computer board technology.
Trenton Technology firstname.lastname@example.org www.trentontechnology.com
Intel, Intel Core, Intel Xeon, and Intel Advanced Vector Extensions are trademarks or registered trademarks of Intel Corporation.
PICMG, SHB Express, COM Express, and CompactPCI are registered trademarks of the PCI Industrial Computer Manufacturers Group.
PCI Express is a registered trademark of the PCI-SIG.
All other product names are trademarks or registered trademarks of their respective owners.