Ammut Computing Unit — AmCU

Ammut Network
3 min readMay 1, 2019

--

One of the most frequently used clichés among tech experts and developers today is Moore’s law; the concept that the speed of computing systems doubles as they half in size every two years. Transistors have clearly followed the law precisely, and so has Artificial Intelligent (AI) chips. These chips are microprocessors designed with the special ability to learn and automate processes and do so more efficiently than traditional methods.

Currently, the market has huge prospects to garner investments in billions of dollars in the coming years. Based on the global need for faster, efficient, cost-effective, and generally intelligent systems and hardware, the Allied Market Research has predicted that by 2025, the AI chips industry would be worth $37.8 billion.

Ammut Network have already started to leverage on the rush for computing edge by recently securing private equity funds worth $4 million. It is indeed a wonderful era for miniaturized computing systems. Just as Moore’s law requires, it’s a never-ending race to tiny-and-more-efficient machines; so, it would not end with wonder-working general-purpose Raspberry Pi, AI companies are still searching for more portable boards blended with functionality.

Ammut Network is also to be credited for its brainchild AmCU or the Ammut Computing Unit; a high power processing hardware that has joined the number boards capable of machine learning. And currently, the company is working on a pioneering deep learning processor that would enable intelligent systems to function like data centers, working real-time, requiring low running power, portable, and cheap. After some interviews, It got accepted in the NVIDIA Inception program, a program for tech pioneers with innovative hardware products in the community . The program would enable the company gets support, network opportunities, and a promotional platform.

The Ammut Computing Unit, designed with optimized machine learning and AI capabilities, is a Single Board Computer (SBC) about the same size as a Raspberry Pi. It was built with tech pros and hobbyists in mind. With the AmCU, AI developers would not have any challenges installing open-source frameworks such as TensorFlow, Caffe, and Keras for machine learning. Via SDK, other frameworks like OpenCV and ROS respectively for computer vision and robotics development can also be used. With AmCU, the thick computing requirements for systems running deep learning models and neural networks are dissolved and accelerated; it is indeed a breakthrough in the world of computer architecture.

This whole race to optimize intelligent systems and chips did not start with Ammut or just this year, AI systems had been built in the ’80s when the whole IoT buzz was yet to stat. and just recently, Google announced its new product called “Coral” — a board designed with Edge TPU platform and custom ASIC having the same form factor and power budget as a Raspberry Pi. And what about Intel? They too have a Neural Compute Stick built to connect to a single board computer. These strides prove that there is indeed a new wave of revolution propelled by Ai, one that is not about the buzz, but the substance. As for the AmCU, it is one of the latest substances in the wave sweeping technology, and it offers two major things: portability and improved computing capacity.

--

--