Future Computers Will be Radically Different – Veritasium

Future Computers Will be Radically Different – Veritasium

  • Posted by MGUG Admin
  • On March 27, 2022

Computers as we know in the modern era are almost entirely digital, and we’ve based their entire architecture around the bit system. When it comes to processing imagery and data there are aspects that, in fact, do have disadvantages against analog methods.

In terms of the Data Sci and Geospatial domain, the relevant takeaways here are that image processing in the context of training is at a bottleneck due to costly computational steps, such as matrix multiplication, as part of the machine learning process – and also many many proponents demanding now purchasing and using Cloud resources. We are at a new phase of tuning the tech pipeline, which will have larger implications on the delivery of products / analytics, trend forecasting; and change management.

 

Tune in to see the full story!

https://www.youtube.com/watch?v=GVsUOuSjvcg

 

Timestamps:

Analog Computing Mechanics

AI Primer

Applications, Institutional Interest and Tech Improvements

Modern Training Methods, GPU Demand

Analog Computing Part 2