Blog

News and opinion


Back to overview

25.04.2019

Features of the VHDL real-time operating system, to enable ANS, Artificial Nervous System

Here are some figures to describe the performance of this operating system:

1. only neural nets will be processed, enabling massive parallel execution
2. aproximately 4000 neural nets can be performed simultaneously within 1 ms
3. refreshing rate of 30fps for max 8 GbE cameras, each with 1M pixel resolution


As the video processing is the most data consuming, the rest of the devices,
such as audio/pressure/etc. sensors + actors + internal organs, have less data,
so 8Gbps is the max. bandwidth to be managed, for in- + outbound devices.

There will be only one type of interface to be supported with the sensors/actors,
namely the Ethernet interface will be used to address the attached devices,
enabling a very simple way to connect devices, with UDP protocol to be used.


There is also an additional connectivity, namely to connect with each other.
This additional interface type can be used to connect to max. 256 other systems,
each between 1-10 Gbps, also only with Ethernet interface and UDP protocol.

Once many "brains" with this operating system are connected to an own network,
which on its side, is based in a data center, building a cloud-based platform,
then many devices can be connected to this system, both on-site and off-site.


As soon as the new 5G standard for worldwide communication will be operative,
it will enable upload/download rates of 1 Gb per second with a latency of 1 ms,
enabling the connectivity with the new operating system directly in real-time.

The cloud-based platform with this operating system can be used to connect
off-site Internet devices, and process the incoming information in real-time,
+evtl. send a response, to one or more connected devices, on-site and off-site.


Here are some examples of combining cloud platform with embedded systems,
implying manage/control/process from the cloud platform online and in real-time,
the Internet connected embedded systems, which all can perform worldwide:

1. network of branches, e.g. manufacturing sites, intelligent buildings, oil platform
2. fleet of automated vehicles, e.g. ships, planes, satellites
3. staff of human-alike members,, e.g. robots, holograms



Back to overview


 
About
LinkedIn