Last March, Google took the wraps off of Coral, a collection of hardware development kits and accessories intended to bolster the development of machine learning models at the edge. It launched in select regions in beta, but the tech giant today announced that it’s graduating to a “wider” and global release.
All Coral products — including the $150 Coral Dev Board, the $74.99 Coral USB Accelerator, and the $24.99 5-megapixel camera accessory — are available for sale at electronics retailer Mouser and for large-volume sale through Google’s sales team. The company says that by the end of the year, it’ll expand distribution into new markets including Taiwan, Australia, New Zealand, India, Thailand, Singapore, Oman, Ghana, and the Philippines.
Coinciding with Coral’s general availability, the Coral website — which now lives at Coral.ai — has been revamped with better organization for docs and tools, testimonials, and “industry-focused” pages. Additionally, it links to a new set of examples aimed at providing solutions to common AI problems, such as image classification, object detection, pose estimation, and keyword spotting.
Lastly, Google says it’ll soon release a new version of the Mendel operating system that updates the system to the latest version of Debian (Buster), and it says it’s hard at work on updates to the Edge TPU compiler and runtime that’ll improve the model development workflow.
“We’ve received a lot of feedback over the past six months and used it to improve our platform,” wrote Coral product manager Vikram Tank. “Coral is also at the core of new applications of local AI in industries ranging from agriculture to health care to manufacturing … [It’s] already delivering impact across industries, and several of our partners are including Coral in products that require fast ML inferencing at the edge.”
For the uninitiated, the Coral Dev Board is a miniature computer featuring a removable system-on-module with one of its custom tensor processing unit (TPU) AI chips. As for the Coral USB Accelerator, it’s a USB dongle designed to speed up machine learning inference on existing Raspberry Pi and Linux systems.
TPUs are application-specific integrated circuits (ASICs) developed specifically for neural network machine learning. The first-generation design was announced in May at Google I/O, and the newest — the third generation — was detailed in May of last year.
The TPU inside the Coral Dev Board — the Edge TPU — is capable of “concurrently execut[ing]” deep feed-forward neural networks (such as convolutional networks) on high-resolution video at 30 frames per second, Google says, or a single model like MobileNet V2 at over 100 frames per second. It sends and receives data over PCIe and USB, and it taps the Google Cloud IoT Edge software stack for data management and processing.
Edge TPUs aren’t quite like the chips that accelerate algorithms in Google’s data centers — those TPUs are liquid-cooled and designed to slot into server racks, and have been used internally to power products like Google Photos, Google Cloud Vision API calls, and Google Search results. Edge TPUs, on the other hand, which measure about a fourth of a penny in size, handle calculations offline and locally, supplementing traditional microcontrollers and sensors. Moreover, they don’t train machine learning models. Instead, they run inference (prediction) with a lightweight, low-overhead version of TensorFlow that’s more power-efficient than the full-stack framework: TensorFlow Lite.
Toward that end, the Dev Board, which runs a derivative of Linux dubbed Mendel, spins up compiled and quantized TensorFlow Lite models with the aid of a quad-core NXP i.MX 8M system-on-chip paired with integrated GC7000 Lite Graphics, 1GB of LPDDR4 RAM, and 8GB of eMMC storage (expandable via microSD slot). It boasts a wireless chip that supports Wi-Fi 802.11b/g/n/ac 2.4/5GHz and Bluetooth 4.1, a 3.5mm audio jack, and a full-size HDMI 2.0a port, plus USB 2.0 and 3.0 ports, a 40-pin GPIO expansion header, and a Gigabit Ethernet port.
The Coral USB Accelerator similarly packs an Edge TPU and works at USB 2.0 speeds with any 64-bit Arm or x86 platform supported by Debian Linux. In contrast to the Dev Board, it’s got a 32-bit Arm Cortex-M0+ microprocessor running at 32MHz accompanied by 16KB of flash and 2KB of RAM.
On the subject of the camera, which is manufactured by Omnivision, it has a 1.4-micrometer sensor with an 84-degree field of view, 1/4-inch optical size, and 2.5mm focal length, and it connects to the Dev Board over a dual-lane MIPI interface. In addition to automatic exposure control, white balance, band filter, and blacklevel calibration, it features adjustable color saturation, hue, gamma, sharpness, lens correction, pixel canceling, and noise canceling.