New Chips Highlight the Advancements in Memory Technology for Industrial ApplicationsApril 28, 2020 by Amos Kingatua
By nature, memory devices are built to last. But when it comes to industrial applications, these parts are expected to far outlast memory devices for consumer products.
Unlike consumer products with shorter life cycles, the memory storage for industrial applications should be accessible and available for several years. That said, electronic systems in an industrial space need reliable RAM and storage solutions with high-level memory operations and data security.
Other desirable features for industrial applications include low power consumption, small footprint, compatibility with other technologies, and long service life.
For real-time applications, fast and reliable access to the read access memory (RAM) by a microprocessor is vital in ensuring greater performance. To reduce reliance on external memory, manufacturers are increasingly integrating larger amounts of fast memory inside the microprocessor chips.
We can better assess how memory technology is advancing in each of these areas by discussing a few recently-released memory components for industrial applications.
A Chip Focused on Extended Memory Life
Hyperstone describes its SATA NAND flash controller, the X1 chip, as a power-efficient, reliable, and secure flash controller for demanding industrial applications. The chip is comprised of 32-bit dual-core microprocessors. It also offers advanced memory management that is said to enhance the storage drive performance and service life.
The X1 chip is designed for industrial applications. Image used courtesy of Hyperstone
The X1 also includes advanced flash technologies, such as FlashXE, hyReliability, hyMap, and others. In addition to its ability to withstand high temperatures, vibrations, and radiation, the controller chip has built-in health monitoring and security features that provide end-to-end data path protection, according to Hyperstone.
The company says its controller chip supports the majority of the common flash memory devices for demanding industrial applications and embedded systems while offering high sequential read and write speeds of 550 MB/s and 500 MB/s, respectively.
Block diagram of the X1. Image used courtesy of Hyperstone
Some of the applications Hyperstone mentions for the XI include data centers, in-vehicle electronics, autonomous vehicles, robots, industrial automation communication, and more. It may also find its way into aerospace, military, and healthcare devices.
MCUs for Speedy Read Access
Some other memory devices geared for industrial applications come from Renesas—namely, their RX72N and RX66N group of MCUs. These 32-bit microcontrollers are said to provide combined equipment control and networking functionalities in addition to their high-capacity, fast built-in memory.
Block diagram of RX72N. Image used courtesy of Renesas
Each MCU has 4 MBs of on-chip flash memory and a fast (120 MHz) read access, in addition to 1 MB of SRAM. This eliminates the need to connect the chip to external memory, which usually has slower read speeds, and facilitates the proprietary RXv3 CPU core. The RX72N group has two Ethernet channels and a maximum of 240 MHz while the RX66N has one channel and a max speed of 120 MHz.
In addition to their high performance, the RX72N and RX66N are each housed in a single chip, which in turn may reduce design complexity, the number of external components, and space requirements. Other notable key features include high application and data security, hardware acceleration, easy integration with other technologies, and shorter development cycles.
Block diagram of RX66N. Image used courtesy of Renesas
By providing equipment control and networking from a single chip, Renesas feels the high-memory MCUs enhance the real-time performance for industrial applications such as robots, PLCs, IIoT, automation, and others.
Memory When Space Is an Issue
Smart Global Holdings, Inc. recently released a tiny form factor, high-density DDR4 Module-in-a-Package (MiP) for space-constrained applications. A typical 16 GB SMART's MiP can reach DDR4 3200 MHz speeds and is usually available in two configurations, either as a 1G x 64 package or the two-channel x 32 package.
Smart Global Holdings asserts that the small footprint of this MiP allows designers to maximize memory capacity while using minimal board space. For example, they can replace the multiple down-board DRAM or SO DIMM with a single MiP.
Diagram of the DDR4 MiP with two Channels of x 32 and signal transferring to and from the memory controller. Image used courtesy of Smart Global Holdings
Aside from saving space, Smart Global Holdings explains several additional benefits to the DDR4 MiP, including:
- Simple circuit design: The MiP has built-in passive components in addition to the thermal sensor, which reduces the routing complexities and external components.
- Improved performance: This type of memory eliminates the need for a mating connector. It also improves signal integrity and reduces flight time.
The small form factor module may be useful for designers working with embedded computing, IIoT, mobile routers, broadcast video, graphics cards, and other high memory applications with space limitations.
The Hefty Place of Memory in Industrial Spaces
You probably see phrases like "data availability," "integrity," "retention," and "performance" floating around discussions of memory devices. But when you throw in the operating conditions of industrial environments—high temperatures, vibrations, shock, etc.—those features of memory technology become ever more valuable.
What features do you prize most in a memory device? Does it depend on the application? Share your thoughts in the comments below.