Movers and SHAKERS
Image Credit: Inspiration4 Photos (Flickr)
The Most Effective Artificial Intelligence Will Never Exist in the Cloud
As SpaceX sends amateur astronauts into orbit and a related company struggles to get its EVs autopilot perfect, we can understand the increased need for edge computing. Or, as some would call edge, “AI on the fly,” the challenge of bringing artificial intelligence as close to the need for output or intervention as possible. The reality of AI computing is, the speed of WiFi, radio waves, and even light are not instant; bringing the decision-making computer as close to the outcome as possible allows for more successful outcomes.
Artificial intelligence (AI) applications include language processing, military decision making, robotic control, face recognition, aviation, hands-free driving, and a long list including much that has not yet been thought of. Many of the mentioned applications are now confined to data centers with a geographic address and significant storage. For many applications, AI decisions are best if onsite or at the “edge” close to the user situation that benefits from the support. Communication out to a network would slow down the response. This is what edge computing provides, onsite decisions with minimal delay in live situations.
Breaking the Climate Controlled Ceiling
Bringing AI closer to the situation with systems that are transportable and durable enough to survive the elements outside of a controlled environment is critical to AI advances. The technological considerations are massive, they include computational ability, durability, heat survival, dampness, size, weight, and power consumption.
A company meeting this challenge is One Stop Systems (OSS). They’ve developed high-end GPU-level processing with massive solid-state storage and minimal network delay. One of the “onboard” challenges being met by their products includes making computations using the surge in information from new sensors and data capture technology. These lead to even more raw information and a growing number of situations with inputs that must be resolved quickly. In real-life situations, there is no time for complex combinations of data to be sent to the Cloud to help with an imminent situation like a vehicle collision or a sudden combat event. Simply, edge computing technology is the future of AI technology - without it, the effectiveness of AI stalls at some point.
Growth Trend of Edge Computing
In 2018, only 10% of generated data was created and processed on board or at the edge. By 2025, it is expected to grow to 75%. Processing simply must be on location to meet the growing needs. Transporting information to an outside server environment unnecessarily could be considered reckless. Transportable AI that has the computational bandwidth and strength is the future of non-human decision-making. Beyond environmental durability, upcoming systems would also have to provide cybersecurity at a level that isn’t possible with Cloud or server solutions (meet FIPS 140-2).
The guts of a machine that takes four amateur astronauts into space, or the onboard computer in a car that avoids an accident after the driver falls to sleep will be the unsung heroes in our amazing technological strides forward. The companies behind the scenes, building out edge and other technologies that make these achievements possible are experiencing increased demand. Investors would be wise to pay attention to these underlying technologies and the providers.
At 1 pm ET today (Sept. 16) One Stop Systems (OSS) will be live with a virtual roadshow available only through Channelchek. There is no cost to attend virtually and listen to the management presentation and submit questions. Edge computing is a very interesting and growing field; learn more by attending the virtual event.
Edge Computing Importance to AI Applications
Trading Technology Continues To Level the Playing Field
This is What Could Slam the Brakes on EV Growth
The SpaceX Inspiration4 Will Bring Space Tourism Closer to Reality
Stay up to date. Follow us: