DeepMind assures that the corporation will not remain in the shadow of the Stargate supercomputer.
Google intends to invest $ 100 billion in artificial intelligence, said DeepMind CEO Demis Hassabis. At the TED conference in Vancouver, he answered a question about rumors about the $ 100 billion Stargate supercomputer, which Google's competitors Microsoft and OpenAI are said to be planning to create. Hassabis did not disclose specific figures, but stressed that Google is investing "more than this amount over time."
The head of DeepMind did not specify what exactly Google will spend a huge budget on and in what time frame, but assured that the company's computing power exceeds the resources of Microsoft and other market players. One of the areas of investment will probably be new hardware, including Axion processors of its own design on the Arm architecture.
According to Google, Axion is 30% faster than "the fastest Arm processors for the mass market" and 50% faster and 60% more energy efficient than Intel's and AMD's x86 chips. However, these data have not yet been confirmed by independent tests.
Part of the budget will probably be allocated to DeepMind, which develops software for AI. Despite some unsuccessful projects, such as material research using AI and a weather forecasting model, the company has made progress in teaching AI social skills, which is a significant step towards creating an even smarter AI.
Hassabis stressed that achieving the company's goals requires "a lot of computing power, "which was" one of the reasons for working with Google in 2014.
While Google may not invest $100 billion in a single supercomputer, the company is definitely increasing its computing resources, for example, with the upcoming construction of a $ 1 billion data center in Kansas City.
Google intends to invest $ 100 billion in artificial intelligence, said DeepMind CEO Demis Hassabis. At the TED conference in Vancouver, he answered a question about rumors about the $ 100 billion Stargate supercomputer, which Google's competitors Microsoft and OpenAI are said to be planning to create. Hassabis did not disclose specific figures, but stressed that Google is investing "more than this amount over time."
The head of DeepMind did not specify what exactly Google will spend a huge budget on and in what time frame, but assured that the company's computing power exceeds the resources of Microsoft and other market players. One of the areas of investment will probably be new hardware, including Axion processors of its own design on the Arm architecture.
According to Google, Axion is 30% faster than "the fastest Arm processors for the mass market" and 50% faster and 60% more energy efficient than Intel's and AMD's x86 chips. However, these data have not yet been confirmed by independent tests.
Part of the budget will probably be allocated to DeepMind, which develops software for AI. Despite some unsuccessful projects, such as material research using AI and a weather forecasting model, the company has made progress in teaching AI social skills, which is a significant step towards creating an even smarter AI.
Hassabis stressed that achieving the company's goals requires "a lot of computing power, "which was" one of the reasons for working with Google in 2014.
While Google may not invest $100 billion in a single supercomputer, the company is definitely increasing its computing resources, for example, with the upcoming construction of a $ 1 billion data center in Kansas City.
