In desktop computers, laptops and smartphones, you can identify compute modules and memory. This approach is known as the "von Neumann architecture" - named after the scientist John von Neumann, one of the pioneers in the field of digital computing. In this architecture, the data is constantly moving between the memory and the computing device, and this is slow and not too effective.
All-in-one computing.
The solution to this problem can be "computational memory"? technology, also known as "memory computing". At the same time for the storage and processing of information only physical properties of computer memory are used.
The IBM research team announced a breakthrough achievement in the field of computational memory by successfully implementing the machine learning algorithm on a group of 1 million memory devices operating on the basis of the effect of phase change (ESF).
The ESF device was created from a germanium-antimony-telluride alloy, placed between a pair of electrodes. It is expected that this pilot technology will provide a 200-fold increase in speed and energy efficiency.
Suitable for artificial intelligence.
Does the ESF device perform the calculations using the crystallization mechanism? when an electric current passes, the disordered arrangement of the atoms is replaced by an ordered arrangement, that is, a crystalline. Scientists demonstrated ESF technology on two examples with time control, and compared with traditional machine learning methods.
The ability to perform faster calculations will affect the overall performance of computers. For IBM, this means more power in applications of artificial intelligence.
CMOS technology has reached the limit, and to overcome its limitations, a radical paradigm shift is called "processor-memory".
Computing memory expands the ability to process data in real time, which is very important now that companies are focusing on analytical processing of information. Industry giants such as Amazon and Google make AI a central element of their business activity, so the speed of AI computers is very much in demand.