Google parent company Alphabet (GOOGL) has unveiled a new artificial intelligence (A.I.) supercomputer that it says is faster and more powerful than any other competing system.
Google is a major A.I. pioneer, having developed some of the biggest advancements in the field over the past decade. However, some analysts say the tech giant has fallen behind in terms of commercializing it’s A.I. inventions.
Rival Nvidia (NVDA), for example, now dominates the market for A.I. model training and deployment, with over 90% market share. Google has been racing to catch-up and has begun designing its own A.I. semiconductor and microchips called “Tensor Processing Units” (TPUs).
Google has now announced that it built a system with over 4,000 TPUs designed to run and train advanced A.I. models.
Google claims that its TPU-based supercomputer, called TPU v4, is now the fastest A.I. system in world.
The amount of computer power needed to train and run generative A.I. platforms such as ChatGPT is expensive, and the tech industry is increasingly focused on developing new semiconductor and microchips that reduce the amount of computer power needed.
News of the Google’s new A.I. supercomputer comes as a growing number of countries take action to slow the development or halt the proliferation of the technology.
In recent days, the Italian government announced that it is banning ChatGPT over privacy concerns, while regulators in Canada have launched an investigation into the technology.
The German government has also threatened to ban chatbots and the European Union is discussing labeling A.I. technology based on risk levels.
Alphabet’s stock is down 24% over the last year and trading at $104.47 U.S. per share.
