Blog

AI benefits from GPU, not CPU advancements

|
September 21, 2018

A quick follow-up to our blog posts about AI

The name of the game is no longer Moore’s Law where we see processors getting exponentially faster. AI technology is driven not by computing processes of the past, but from an evolution beyond central processing unit (CPU) advances to graphics processing unit (GPU)-based processors. These graphics chips used by gamers are being used by AI for their massively parallel-processing capability. As Talla commented, “We now have the equivalent of a super computer on a single chip. This allows image recognition to make a huge leap forward.”

Now, with AI, deep-thinking image identification is faster and more ubiquitous. By 2020, NVIDIA estimates there will be 1 billion cameras deployed for surveillance worldwide. But why do we build them? For public safety, parking, or customer experience of Disneyland? Where do we store this data, and how do we use the data?

What do you think? Join the discussion and comment below!