Nvidia AI: bringing machine learning to gaming, healthcare

Nvidia is relatively antique for a technology company, having been founded in 1993 by a trio including CEO Jensen Huang.
To the public, the company is perhaps best known for its graphics cards used in PC gaming, a market which it dominates alongside its compatriot AMD. That in itself is big business, with Nvidia stating that it draws on a market of 2.7 billion gamers worldwide (albeit most of them not gaming on PCs) and participates in an esports scene that is set to generate $1.1bn of revenue this year.
In more recent times, the company has branched out from purely focusing on gaming to areas such as automotive (where it partnered with Continental and others on an autonomous vehicle alliance) and, increasingly, artificial intelligence.
SEE ALSO:
-
SAS to make $1bn investment and partners with NVIDIA in artificial intelligence drive
-
Nvidia to buy Israeli chipmaker Mellanox for $6.9bn in data centre push
-
Bosch, Nvidia, VW, Aquantia and Continental form autonomous vehicle alliance
Nvidia has emphasised the synergy of its products in these different areas. Deep learning, the subset of machine learning using artificial neural networks, tends to rely on the usage of graphics cards, which are adept at highly parallel tasks. In turn, Nvidia brings deep learning methods to bear on gaming, with its Deep Learning Super Sampling technology allowing for the resolution upscaling of games while they’re being played.
Another area in which the company’s technology is gaining traction is in healthcare. A number of companies participating in the company’s Inception startup accelerator programme have interests in this direction. One, Deep 6 AI, focuses on using machine learning to analyse medical records and identify potential candidates for medical trials. Another, Aiforia, is using AI to automate the analysis of tissue samples.
Kaisa Helminen, CEO of Aiforia, said: “With deep learning AI, we are able to extract more information from a patient tissue sample than what’s been previously possible due to limitations of the human eye.”
(Image: Nvidia)