Huang’s law will guide the development of artificial intelligence systems for the next decade

As a rule of thumb, formulated by Intel co-founder Gordon Moore, the number of transistors per unit of processor area doubles every two years. NVIDIA is trying to set the trend for computing performance to more than double every two years.

Image source: HWRIG

NVIDIA founder Jensen Huang likes to say that the so-called “Moore’s Law” is dead. His own law, according to the authors of the publication on the pages The Wall Street Journal, is ahead of this rule in terms of the rate of increase in the performance of computing systems. Since 2012, the performance of NVIDIA components in computing used in artificial intelligence systems has increased 317 times, according to the company. In other words, the performance of such systems based on NVIDIA hardware components has more than doubled annually on average.

The purchase of Arm, according to market participants, will allow NVIDIA to strengthen progress in the field of artificial intelligence through edge computing: some of the operations critical to system performance will be performed directly by the end device, and not transferred to the server. In tandem with a developed server ecosystem, this will allow NVIDIA to make significant progress in creating new areas of artificial intelligence.

For NVIDIA, the balance of computing resources is also important, because if the graphics processors significantly outstrip the central ones in terms of speed, then the “weak link” will hold back the development of the entire system. The purchase of Arm will allow it to get hold of a valuable asset in the development of central processing units. Arm VP of Marketing Steve Roddy expects the so-called “Huang’s Law” to be exhausted in ten years, but over that time the industry will make significant progress in developing artificial intelligence systems.

If you notice an error, select it with the mouse and press CTRL + ENTER.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Recent News