Google’s Edge TPU AI chip will run on a USB stick computer
Jul 31, 2018 — by Eric Brown 8,619 views
Google followed up on its Edge TPU machine learning chip announcement by unveiling a USB Type-C based version that you can plug into any Linux or Android Things computer, including a Raspberry Pi. There are also new details on the Edge TPU dev board.
Following Google’s announcement of an embedded friendly Edge TPU version of its Tensor Processing Unit AI chip and the related Cloud IoT Edge stack for IoT gateways, the company announced a USB stick computer version of Edge TPU that can work with any Linux or Android Things computer. It also posted more details on the upcoming, NXP-based Edge TPU development kit, including its SoC: an NXP i.MX8M.
Edge TPU chip (left) and Edge TPU Accelerator
(click images to enlarge)
The Edge TPU Accelerator uses the same stripped-down Edge TPU neural network coprocessor that is built into the upcoming dev kit. It has a USB Type-C port to plug into any Debian Linux or Android Things computer to accelerate machine learning (ML) inferencing for local edge analytics. The 65 x 30mm device has mounting holes for host boards such as a Raspberry Pi Zero.
Like the Edge TPU development kit, the Edge TPU Accelerator will enable the processing of ML inference data directly on-device. “A local ML accelerator increases privacy, removes the need for persistent connections, reduces latency, and allows for high performance using less power,” says Google.
— ADVERTISEMENT —
The Edge TPU Accelerator competes with products like Intel’s Neural Compute Stick, previously referred to as the Fathom. The USB-equipped Neural Compute Stick is equipped with the Movidius Myriad 2 VPU and neural network accelerator.
 AIY Vision Kit |
Both the Edge TPU Accelerator and dev kit are provided
by Google’s community-backed AIY Projects group, which is known for its low-cost cardboard constructed kits for Google Cloud related embedded technologies. These include the AIY Vision Kit for the Raspberry Pi Zero W and WH, which performs TensorFlow-based vision recognition. It incorporates a “VisionBonnet” board with a Myriad 2 chip. AIY Projects also provides an AIY Voice Kit with the same RPi Zero WH target that lets you build a voice-controlled speaker with Google Assistant support.
More Edge TPU dev kit details
The Edge TPU Accelerator will ship in October along with the Edge TPU chip and development kit. More details were posted on the dev kit. We now know that the computer-on-module that features the Edge TPU will run either Debian Linux or Android Things on NXP’s i.MX8M. The 1.5GHz, Cortex-A53 based i.MX8M integrates a Vivante GC7000Lite GPU and VPU, as well as a 266MHz Cortex-M4 MCU.
Two views of the Edge TPU dev kit
(click images to enlarge)
The unnamed, 48 x 40mm module will ship with 1GB LPDDR4, 8GB eMMC, dual-band WiFi-ac, and Bluetooth 4.1. The baseboard of the dev kit will add a microSD slot, as well as single USB Type-C OTG, Type-C power (5V input), USB 3.0 host, and micro-USB serial console ports.
The Edge TPU development kit baseboard is further equipped with GbE and HDMI 2.0a ports, as well as a 39-pin FPC connector for 4-lane MIPI-DSI and a 24-pin FPC for 4-lane MIPI-CSI2. There’s also a 40-pin expansion connector, but with no claims for Raspberry Pi compatibility. The 85 x 56mm board also provides an audio jack, a digital mic, and a 4-pin terminal for stereo speakers.
The Edge TPU ASIC inside the dev kit and Edge TPU Accelerator is a lightweight, embedded version of Google’s Cloud TPU chips and modules. It runs TensorFlow Lite ML models on Linux and Android Things computers. Edge TPU enables concurrent execution of multiple AI models per frame on a high-resolution video, at 30fps, says Google. Equipped with PCIe and USB interfaces, the Edge TPU is said to be optimized for small size and power efficiency.
The Cloud IoT Edge stack controls Edge TPU assisted gateways and closely links them to Google Cloud to enable cloud-integrated IoT edge computing and analytics. The combined platform lets developers build and train ML models in the cloud, and then run those models on the Cloud IoT Edge device.
Further information
An early version of the Edge TPU Accelerator and Edge TPU development kit will be available in the U.S. in October with undisclosed pricing. You can apply for early access to the Cloud IoT Edge Alpha and Edge TPU Early Access development boards from Google’s new Edge TPU Devices page, which details both the Accelerator and development kit, as well as from its Edge TPU and Cloud IoT Edge product pages.
More information may be found in the Edge TPU Accelerator announcement, as well as the original Edge TPU announcement.
That dev board strongly resembles technexion’s pico boards with the hirose connectors, but in a different arrangement. I hope its by them. I really like their stuff. The debian support sounds great too.
No google edge devices for Australia. US is treating friendly countries as threats. Google to close down access to Australia to cloud based AI services? If not then open up google edge tech to US friends.