Category: HardWare

Currently, there are more than 100 companies all over the world building ASICs (Application Specific Integrated Circuit) or SOC’s (System on Chip) targeted towards deep learning applications. There is a long list of companies here. In addition to these startup big companies like Google (TPU), Facebook, Amazon (Inferentia), Tesla etc are all developing custom ASIC’s for deep learning training and inference.
These can be categorized into two types:

1. Training and Inference — These ASIC’s are designed to handle both training the deep neural network and also performing inference. Training a large neural network like Resnet-50 is a much more compute-intensive task involving gradient descent and back-propagation. Compared to training inference is very simple and requires less computation. NVidia GPU’s, which are most popular today for deep learning, can do both training and inference. Some other examples are Graphcore IPU, Google TPU V3, Cerebras, etc. OpenAI has great analysis showing the recent increase in compute required for training large networks.
2. Inference — These ASICs are designed to run DNN’s (Deep neural networks) which have been trained on GPU or other ASIC and then trained network is modified (quantized, pruned etc) to run on a different ASIC (like Google Coral Edge TPU, NVidia Jetson Nano). Most people say that the market for deep learning inference is much bigger than the training. Even very small microcontrollers (MCU’s) based on ARM Cortex-M0, M3, M4 etc can do inference as shown by the TensorFlow Lite team.

Making any chip (ASIC, SOC etc) is a costly, difficult and lengthy process typically done by teams of 10 to 1000’s of people depending on the size and complexity of the chip. Here I am only providing a brief overview specific to deep learning inference accelerator. If you have already designed chips you may find this too simple. If you are still interested, read on! If you like it share and 👏  …..
Full article: https://towardsdatascience.com/how-to-make-your-own-deep-learning-accelerator-chip-1ff69b78ece4

Shenzhen Xunlong Software’s $19.90 “Orange Pi AI Stick Lite” USB stick features a GTI Lightspeeur SPR2801S NPU at up to 5.6 TOPS @ 100MHz. It’s supported with free, Linux-based AI model transformation tools.

Shenzhen Xunlong Software’s Orange Pi project has released an AI accelerator with a USB stick form factor equipped with Gyrfalcon Technology, Inc.’s Lightspeeur SPR2801S CNN accelerator chip. The Orange Pi AI Stick Lite is designed to accelerate AI inferencing using Caffe and PyTorch frameworks, with TensorFlow support coming soon. It’s optimized for use with Allwinner based Orange Pi SBCs, but the SDK appears to be adaptable to any Linux-driven x86 or Arm-based computer with a USB port.

 
Orange Pi AI Stick Lite

The Orange Pi AI Stick Lite is a relaunch of an almost identical Orange Pi AI Stick 2801 that was announced in Nov. 2018, according to a CNXSoft post. The previous model cost $69 and required purchasing GTI’s PLAI (People Learning Artificial Intelligence) model transformation tools for $149 to do anything more than run a demo. The new device is not only much cheaper at $19.90, but the PLAI training tools are now free. There’s no download button, however — you must contact the company to get the download link.

GTI’s up to 9.3 TOPS per Watt Lightspeeur SPR2801S is a lower-end sibling to the up to 24-TOPS/W Lightspeeur 2803S NPU, which is built into SolidRun’s i.MX 8M Mini SOM. The “best peak” performance of the 2801S is 5.6 TOPS @ 100MHz. It can also run in an “ultra low power” mode of 2.8 TOPS @ 300mW. GTI also offers a mid-range Lightspeeur 2802 model at up to 9.9 TOPS/W.

 

The 28nm fabricated, 7 x 7mm Lightspeeur SPR2801S has an SDIO 3.0 interface and eMMC 4.5 storage. It offers read bandwidth of 68MB/s and write bandwidth of 84.69 MB/s. The NPU includes a 2-dimensional Matrix Processing Engine (MPE) featuring an APiM (AI Processing in Memory) technology that uses magnetoresistive random access memory (MRAM) …..

sources: http://linuxgizmos.com/orange-pi-ai-stick-lite-taps-5-6-tops-gryfalcon-gpu/

Google has promised us new hardware products for machine learning at the edge, and now it’s finally out. The thing you’re going to take away from this is that Google built a Raspberry Pi with machine learning. This is Google’s Coral, with an Edge TPU platform, a custom-made ASIC that is designed to run machine learning algorithms ‘at the edge’. Here is the link to the board that looks like a Raspberry Pi.

This new hardware was launched ahead of the TensorFlow Dev Summit, revolving around machine learning and ‘AI’ in embedded applications, specifically power- and computationally-limited environments. This is ‘the edge’ in marketing speak, and already we’ve seen a few products designed from the ground up to run ML algorithms and inference in embedded applications. There are RISC-V microcontrollers with machine learning accelerators available now, and Nvidia has been working on this for years. Now Google is throwing their hat into the ring with a custom-designed ASIC that accelerates TensorFlow. It just so happens that the board looks like a Raspberry Pi.

WHAT’S ON THE BOARD

On board the Coral dev board is an NXP i.MX 8M SOC with a quad-core Cortex-A53 and a Cortex-M4F. The GPU is listed as ‘Integrated GC7000 Lite Graphics’. RAM is 1 GB of LPDDR4, Flash is provided with 8GB of eMMC, and WiFi and Bluetooth 4.1 are included. Connectivity is provided through USB, with Type-C OTG, a Type-C power connection, a Type-A 3.0 host, and a micro-B serial console. Gigabit Ethernet, a 3.5mm audio jack, a microphone, full-size HDMI, 4-lane MIPI-DSI, and 4-lane MIPI-CSI2 camera support. The GPIO pins are exactly — and I mean exactly — like the Raspberry Pi GPIO pins. The GPIO pins provide the same signals in the same places, although due to the different SOCs, you will need to change a line or two of code defining the pin numbers.

You might be asking why Google would build a Raspberry Pi clone. That answer comes in the form of a machine learning accelerator chip implanted on the board. Machine learning and AI chips were popular in the 80s and everything old is new again, I guess. The Google Edge TPU coprocessor has support for TensorFlow Lite, or ‘machine learning at the edge’. The point of TensorFlow Lite isn’t to train a system, but to run an existing model. It’ll do facial recognition.

The Coral dev board is available for $149.00, and you can order it on Mouser. As of this writing, there are 1320 units on order at Mouser, with a delivery date of March 6th (search for Mouser part number 212-193575000077).

source: https://hackaday.com/2019/03/05/google-launches-ai-platform-that-looks-remarkably-like-a-raspberry-pi/