SolidRun Targets Edge AI Work with New i.MX 8M Plus Computer-on-Module and HummingBoard Parts

With up to quad-core processors, a real-time core, NPU coprocessor, and a high-performance DSP, the new CoMs are impressively configured.

Embedded electronics specialist SolidRun has officially launched a new family of computers-on-modules (CoMs) based on NXP's i.MX 8M Plus system-on-chip (SoC) — and it's targeting edge artificial intelligence (AI) workloads with the integration of a Cadence Tensilica HiFi 4 digital signal processor (DSP).

"Our new i.MX 8M Plus enhanced CoMs are available in quad- and dual-core configurations and maintain the same ultra-compact 47mm by 30mm footprint and pin layout as SolidRun’s other i.MX 8M CoM products," the company explains of the new parts. "Maintaining this footprint makes it easy for developers to scale their processing capabilities up or down between our other i.MX 8M product families, including the i.MX 8M / i.MX 8M Mini/ i.MX 8M Nano. Pin compatibility and shared software resources provide a flexible and forward-thinking approach to creating future-proof products for high-performance and low power applications."

"As with all of our i.MX CoMs, the new i.MX 8M Plus boards have all necessary subsystems and connectivity options integrated. They are a great fit for simple applications to those that require high-performance processing capabilities for AI and machine learning inference at the edge. Because they’re based on NXP’s new i.MX 8M Plus processor, our CoMs are equipped with a high-performance Cadence Tensilica HiFi 4 DSP that supports voice and natural language processing chores — great for building voice-enabled solutions, as well as an integrated neural processing unit for enhanced AI and machine learning capabilities at the edge."

The new i.MX 8M Plus-based CoMs and carrier boards from SolidRun target edge AI deployments. (📹: SolidRun)

Both the dual- and quad-core variants run at 1.8GHz in standard format or 1.6GHz in the industrial-rated model, offer up to 2.3 TOPS of machine learning performance from an embedded neural processing unit (NPU) coprocessor, include an Arm Cortex-M7 microcontroller running at up to 800MHz for real-time tasks, have a Vivante GC7000UL graphics processor with OpenGL ES 3.1, Vulkan, OpenCL 1.2 FP, and OpenVG 1.1 support, have two gigabit Ethernet ports, optional 802.11ac/a/b/g/n Wi-Fi and Bluetooth 5.0 radios, and can be configured with up to 8GB of LPDDR4 memory. Storage options include onboard eMMC and external NOR flash, microSD, and PCI Express SSD, plus additional storage via two USB 3.0 ports.

To support the new modules, SolidRun has also launched new carrier boards: The HummingBoard Mate, HummingBoard Pulse, and HummingBoard Ripple, which include gigabit Ethernet connectivity, USB 3.0, HDMI, audio in and out, a four-lane MIPI CSI-2 camera input, and mikroBUS expansion along with access to the modules' 75 general-purpose input/output (GPIO) pins, I2C, PSI, QSPI, PWM, JTAG, CAN-FD, and PCI Express Gen. 3 functionality.

"Our cost-effective, feature-rich family of i.MX 8M CoMs have been instrumental in supporting developers with turning ideas into actual, working IoT products," claims Dr. Atai Ziv, SolidRun chief executive. "The addition of our new i.MX 8M Plus CoMs will unlock a near limitless world of possibilities, allowing customers to take advantage of the i.MX 8M Plus’s numerous embedded features to create exciting new IoT solutions that benefit from real-time machine learning inference and computer vision at the edge."

Pricing for the modules has been set to start at $75 with a kit including carrier board priced at $135, though these are "promotional" prices which will only be available through to March 2021. More information is available on the SolidRun website.

Gareth Halfacree
Freelance journalist, technical author, hacker, tinkerer, erstwhile sysadmin. For hire: freelance@halfacree.co.uk.
Latest articles
Sponsored articles
Related articles
Latest articles
Read more
Related articles