BROWSE BY TECHNOLOGY



RTC SUPPLEMENTS


TECHNOLOGY IN CONTEXT

On-Board Storage

SSDs Take Hold in Embedded Applications

With lower costs, higher reliability, greater capacity and faster read/write speeds, solid-state storage devices are moving into embedded applications leading both to more alternatives as well as a move toward standardizing interfaces.

GARY DROSSEL, SILICONSYSTEMS

  • Page 1 of 1
    Bookmark and Share

Amazing as it might seem, there has never been a “specific” solid-state drive (SSD) form factor designed for onboard storage for the embedded system OEM market. Rather, designers have had to use available SSD form factors and interfaces compatible with the chipsets originally designed for consumer applications. The SSD market has been driven by innovation and consumer demand for smaller, more powerful consumer electronics devices such as iPods, digital cameras and smart phones, and it has begun to evolve for both consumer and industrial-grade applications. Huge consumer adoption has led to decreases in cost per gigabit in NAND storage devices, making consumer-based SSD devices affordable to the masses.

Likewise, SSDs for consumer applications began to find their way into a wide range of industrial applications due to lower costs. However, embedded system OEMs quickly discovered that they did not deliver the high performance, high reliability and long product life required in critical system applications. As a result, systems deployed in the field have experienced high failure rates because of a variety of endurance, power anomaly or environmental issues. The demand for SSDs also spread in embedded system applications due to the affordability of higher capacities in newer form factors such as 1.8-inch and 2.5-inch drives, beginning the evolution of hard disk drive replacement.

Early adoption of SSDs was primarily made by high-end industrial and military system designers who needed a storage solution that would overcome the performance and reliability issues of hard drives in rugged and mission-critical applications. Initial applications included flight and mission data recorders, troop and field computers, GPS communication systems and industrial applications that mandated high shock, vibration and extreme temperature tolerances. As SSD technology and the industry matured, SSDs became a more attractive solution for single board computing developers, which expanded onboard integration to include everything from edge routers and switches in netcom applications, industrial automation and control equipment to ATM machines and other forms of interactive kiosks and medical devices.

CompactFlash initially emerged as the form factor of choice for embedded systems due to its small footprint, performance, reliability and its adoption as a well-understood industry standard. The initial cost per gigabyte for CompactFlash was very high, limiting adoption to embedded applications where it was crucial to have no moving parts and the ability to withstand extreme temperatures and high shock and vibration. As the cost decreased, demand exploded due to the merits of solid-state storage in many applications. However, looking to the future, there is a downside to using CompactFlash because of its limited interface compatibility with Parallel ATA (PATA).

Demand also grew for SSD solutions as primary storage in embedded applications that did not require ruggedization or were not deployed in harsh environments. For these applications, SSDs were attractive for their lower power consumption and long product life in usage models that often required 24/7 “always on” capability. SSDs in 2.5-inch and 1.8-inch form factors became true drop-in replacements for hard drives in embedded applications that required high endurance, multi-year product deployments of up to 10 years, faster data transfers, less frequent product requalifications and protection from the effects of power anomalies.

Onboard Storage Meets Embedded

Demand for onboard storage that meets designers’ performance, reliability, size, low power, interface, scalability and security requirements has resulted in new SSDs with integrated technologies to match embedded system application requirements. Industrial-grade SSDs are available in a wide range of form factors from 2.5-inch and 1.8-inch drives, to CompactFlash and PC Cards and 40 and 10-pin modules plus new ultra-small modules with no compromise in performance, reliability, scalability and product life. Now that SSDs are viable for onboard storage, embedded system designers can choose from SSD technology that has evolved with faster read/write speeds and that supports popular USB and SATA interfaces, as well as traditional PATA, SD and MMC interfaces (Table 1). Important, too, the latest SSDs are available in capacities that match typical embedded system usage models. These tend to use SSDs for operating system storage, fast boot capabilities, event/error logging, look-up tables, etc. Optimal storage capacity for these functions is well within the capacity range of most SSDs: 512 Mbytes to 32 Gbytes.

Because they only require a fraction of the system power of hard drives, SSDs are an appropriate choice for onboard storage even in the most space-constrained designs. Adding to the power and reliability advantages of SSDs, are integrated technologies that prevent drive or data corruption from power anomalies. The number one cause of storage system field failures is drive corruption from an ungraceful power-down, brownout, power spike or unstable voltage level. When power goes out, the result is often a corrupted drive and ruined data, resulting in costly unscheduled downtime as field technicians reformat drives, reinstall operating systems or return products. New integrated voltage detection technology in SSDs eliminates drive corruption in the event of a power disturbance adding significantly to system reliability.

Contributing to its high endurance and reliability, solid-state storage technology features robust wear-leveling and error correction code (ECC) algorithms. In addition, integrated technologies that act as an early warning system to forecast usable life are now standard features on advanced SSDs to virtually eliminate the chances of storage system wear-out. This allows OEMs to set usage model thresholds to schedule field maintenance and drive replacement without incurring expensive unscheduled downtime.

Security breaches are on the rise for embedded systems, which means that robust security must also be a crucial feature for SSDs. In the past, system design challenges due to the small footprint and low-power requirements for storage systems used in many embedded system applications prevented storage security options from extending beyond basic encryption technology. Today’s leaders in solid-state storage technology offer an array of advanced user-selectable security options that prevent IP theft, protect application data and software IP from theft, corruption, and accidental or malicious overwrites. Applications such as data recorders, wearable and field computers, medical monitoring and diagnostic equipment, POS systems and voting machines are now requiring more advanced levels of security that make sure the SSD is usable only on the originally intended host system.

New Ultra-Small Form Factor SSD Option

The embedded industry has started to move away from parallel interfaces in general, preferring the faster serial interface. There is also a growing trend in embedded computing to migrate to smaller form factor boards such as Computer-on-Modules (COMs) where CompactFlash is actually too big for the computing platform. Addressing embedded system demand for an ultra-small advanced storage form factor that supports faster interfaces, SiliconSystems has developed the company’s SiliconDrive II Blade product family, which combines integrated advanced storage technologies with the robustness and locking mechanism of the SiliconBlade Socket (Figure 1).

The SiliconDrive II Blade was designed as an alternative to SD and MMC cards in netcom, embedded, industrial, military and medical applications. More robust, secure and a quarter of the size of a CompactFlash card, it offers improved shock and vibration due to the “lock down” mechanism of the connector. Another benefit of the SiliconDrive II Blade is that it is upgradeable in the field for long deployments, unlike chip down solutions that are soldered to the PCB. The new form factor is available in capacities from 512 Mbytes to 4 Gbytes.

Responding to market demand for open standards and the need for second sourcing of products, SiliconSystems has contributed its SiliconDrive II Blade specification to the Small Form Factor Special Interest Group (SFF-SIG) for the purpose of creating an official governing standard. With its growing membership of embedded computing suppliers who are interested in advancing small form factor designs and adoption, the SFF-SIG has formed a working group to develop a specification under the “MiniBlade” name that will define a wide array of storage, communications, GPS and other I/O products. Members of the SFF-SIG will have full access to the MiniBlade specification. The small form factor embedded system market will now have a standardized ultra-small mass storage solution.

More Options Match Market Requirements

There are a plethora of SSD options available today enabling OEMs to choose the onboard storage solution that matches their chipset, host interface, performance, reliability, power, usage model and security requirements. OEMs have the flexibility to design a wider range of products that match the needs of a variety of markets.

Designers no longer need to work around form factors and interfaces compatible with the chipsets originally designed for consumer applications. Today, there is a good variety of SSD form factor options for onboard storage that offer the high performance, high reliability and long product life needed for the rigors of embedded system applications.

SiliconSystems
Aliso Viejo, CA.
(949) 900-9400.
[www.siliconsystems.com].