Shulou(Shulou.com)11/24 Report--

CTOnews.com, June 8 (Xinhua)-- AlphaDev, an artificial intelligence developed by Google's DeepMind, has created a new way to sort data, 70% faster than algorithms designed by human programmers, a discovery that could make millions of software run faster around the world.

Data sorting is one of the basic functions of a computer, which is used to arrange data in alphabetical order or numeric size. At present, there are many different sorting algorithms, but after decades of optimization, it is difficult for human beings to innovate.

The design goal of AlphaDev is to discover new algorithms to accomplish a given task and to go beyond existing methods as far as possible. Instead of fine-tuning the current algorithm, it is built from scratch. It uses an intermediate language called assembly code, which is closer to computer binary instructions than human code and makes it easier for AlphaDev to create more efficient algorithms.

AlphaDev generates one instruction at a time, then tests its output correctly, and requires the shortest algorithm to be generated. DeepMind says the task is very difficult because of the large number of possible instruction combinations.

When asked to create a sorting algorithm, AlphaDev amazingly found a way to sort five pieces of data 70 per cent faster than the best algorithm. Even if 250000 pieces of data are sorted, it can be 1.7% faster than the best algorithm.

"at first we thought it was wrong or there was some bug, but when we analyzed the program, we realized that AlphaDev had really found a faster way." Said Daniel Mankowitz of DeepMind.

As the sorting algorithm is widely used in a variety of commonly used software, this improvement may have a significant impact on global computing. DeepMind has made them open source and has added a common code base called Libc++, which means anyone can use them immediately, the first update to the sorting algorithm part of the code base in more than a decade.

Mankowitz says Moore's Law is coming to an end because of physical limits, but AlphaDev may be able to make up for that by improving efficiency.

Mark Lee of the University of Birmingham in the UK says AlphaDev is interesting and even a 1.7 per cent speed increase is useful. But he says he is skeptical that this approach can make up for the end of Moore's Law because it is unlikely to make the same progress in more complex software.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.