Roving Reporter: Adding Graphics and Storage to Intel® Quark™ SoC-based Solutions

Widespread deployment of headless IoT gateways that allow cloud or back end control of machine-to-machine (M2M) interaction with edge devices and sensors is fueling the growth of the Internet of Things (IoT). However, many industrial IoT applications require a local human machine interface (HMI). Additionally, they may require storage for data that may need to be retrieved or used at some future point in time.

With Intel IoT® Gateway designs, the requirement for a visual interface and storage usually means stepping up from a headless Intel® Quark™ SoC-based gateway to one using an Intel® Atom™ processor or Intel® Core™ processor. But for certain power- and cost-constrained applications, this may exceed design parameters.

Fortunately, Silicon Motion, a General member of the Intel® Internet of Things Solution Alliance, has a solution. Their Osprey Visual IoT platform combines an Intel Quark SoC-based Intel IoT Gateway solution with their SM750 graphics controller and Ferri-eMMC* embedded memory (Figure 1). This solution can accelerate product design for these power- and cost-constrained applications. It also opens up new opportunities to create many smart infrastructure devices using a low-power industry-standard Intel® platform.

Figure 1. Silicon Motion’s Osprey Visual IoT platform is Intel® Quark™ SoC-based Intel® IoT Gateway design that includes their SM750 graphics controller and Ferri-eMMC* embedded memory.

At first it might seem hard to beat the low power requirements of a processor such as the Intel® Atom™ processor E3815. This processor features a thermal design power (TDP) of just 5W. But add a board and the total operating power goes up. It can range from 6 to 10W.

Silicon Motion’s Osprey Visual IoT platform (Figure 2) starts with an Intel Quark SoC running at just 2.2W. Adding board, graphics controller and embedded memory takes the operating power to just 4.1W running dual displays or under 4W running just one display.

Figure 2. Silicon Motion Osprey Visual IoT platform (107mm x 115mm).

Several factors account for this low operating power. With the Silicon Motion SM750 graphics controller (Figure 3), the Intel Quark SoC needs to function only as a command and control processor, managing the graphics but not actively taking part in them. What’s more, Silicon Motion’s on-board frame buffer memory is extremely efficient in offloading processor memory from display refresh duties.

Figure 3. Photo and diagram of Silicon Motion SM750 graphics controller.

With some workloads, operating power can be even less. Silicon Motion’s ReduceOn intelligent power management feature algorithmically varies the SM750 graphics controller’s clock and disables unused units, resulting in dramatic reduction in power usage without compromising quality or performance. In addition, the controller’s low power idle mode puts the frame buffer memory in self-refresh for fast response (instant-on) on wake-up.

The combination of the SM750 graphics controller’s ability to offload any graphics management from the processor and the controller’s built-in power management features enables Silicon Motion to achieve what might be the lowest power display solution for Intel processors available today. Silicon Motion measurements show the SM750 uses less the 1.5W even when driving two full HD displays.

The cost is low too. Silicon Motion estimates the cost of its Osprey Visual IoT platform at approximately two thirds the cost of an Intel IoT Gateway based on the Intel Atom processor E3815.

While the Osprey Visual IoT platform keeps power needs and costs low, it doesn’t skimp on performance. The SM750 graphics controller delivers dual full HD 1920x1080 display output capability for both analog RGB (VGA) and digital display (HDMI/DVI). Additionally, for machine vision, video capture, and test analysis applications, the platform has video input ports to capture video data directly to the frame buffer, offloading the processor. Using an A/D video converter, the SM750 can input YUV (color encoding that reduces bandwidth) directly into the frame buffer. Video can be converted to RGB on the way in, or left in YUV format for the SM750 to convert to RGB on the back-end display output. The processor can access the video data in the frame buffer over a PCIe interface if required for image processing.

Silicon Motion’s graphics products are packaged in multi-chip modules (MCM) with embedded video memory to maximize cost effectiveness, minimize power consumption, and ensure prolonged device longevity. The integration of the graphics fundamentals provides outstanding reliability and flexibility to address the requirements of a broad range of applications.

Highly Reliable Solid State System Storage

The storage component, Ferri-eMMC, is a cost-effective and highly reliable single-package embedded memory solution fully compliant with JEDEC-standard eMMC 4.5 protocols for industrial and commercial applications (Figure 4). Based on Silicon Motion's advanced NAND management technologies including error correction, bad block management, and NAND health monitoring, the Ferri-eMMC provides robust solid state system storage for industrial embedded applications. It includes:

• Advanced ECC management, wear leveling and data refresh features for exceptional data retention and protection against read disturbance
• Advanced system level protection against unstable power for reliable data protection

Figure 4. Diagram of the Silicon Motion Ferri-eMMC* embedded memory storage solution.

Ferri-eMMC is available in 2GB to 32GB configurations. It comes in both commercial (-25°C to 85°C) and industrial temperature grades (-40°C to 85°C).

Turning Edge Data into Value with Intel® IoT Gateways

Intel IoT Gateways provide a flexible and scalable infrastructure from the edge to the datacenter. Application ready, with the flexible architecture to ingest various types of data and the performance to meet varying workload requirements, Intel IoT Gateways help organizations scale to meet evolving needs and continue to future-proof their operations.

Osprey Visual IoT platforms are based on Intel IoT Gateway designs that include pre-integrated and pre-validated hardware and software from Intel, McAfee, and Wind River. Using them, companies can develop, prototype, and deploy intelligent, secure gateways and maintain interoperability between new infrastructure and legacy systems, including sensors and datacenter servers. Intel IoT Gateways include security technologies that tightly integrate the hardware-based security of Intel® processors with operating system and application software security to enable seamless, secure data flow from the edge to the cloud, protecting data in flight or at rest.

Intel’s Lower Power SoC

The Intel Quark SoC is Intel’s lowest-power secure SoC. Its integration of I/O interfaces, clocks, and voltage regulator in a 15mm x 15mm package simplifies design and reduces external components required on the platform. Features include:

• BGA packaging with a 0.593 ball pitch that allows low-cost PCB designs in cost-sensitive applications
• Rich I/O that provides two on-chip Ethernet* interfaces, PCI Express,* USB 2.0, SD/SDIO/eMMC, SPI, UART, and I2C/GPIO
• Intel® Pentium® processor instruction set architecture (ISA) that enables applications to scale from the Intel Quark SoC to platforms based on Intel® Atom™ and Intel® Core™ processors without recompiling code
• Compatibility with 32-bit Intel® architecture solutions and silicon that reduces time-to-market for new products while maintaining interoperability with legacy solutions
• Extended temperature options (-40° C – +85° C) that support thermally constrained and harsh operating environments.

Conclusion

Osprey Visual IoT platforms deliver display capabilities and storage in an Intel Quark SoC-based Intel IoT Gateway design for very little power and cost. Our Solutions Directory includes other low power Intel IoT Gateway designs offering graphics, though none will be this low.

Silicon Motion is a General member of the Intel® Internet of Things Solutions Alliance.

Mark Scantlebury

Roving Reporter (Intel Contractor), Intel® Internet of Things Solutions Alliance

Editor-in-Chief,Embedded Innovator magazine