# Anti-Turing Machine

## TL;DR

The paper proposes the concept of an Anti-Turing machine, a data-centric and decentralized computing paradigm that processes data in-place within a 2D or 3D memory space, moving beyond traditional CPU-centric models.

## Contribution

It introduces the Anti-Turing machine paradigm, emphasizing in-memory data processing and a new programming approach that focuses on data transformation goals rather than data structure.

## Key findings

- Enables in-place data processing in 2D/3D memory spaces.
- Shifts focus from data structure to data transformation goals.
- Proposes a new computing paradigm beyond Turing-based models.

## Abstract

The invention of CPU-centric computing paradigm was incredible breakthrough of computer science that revolutionized our everyday life dramatically. However, the CPU- centric paradigm is based on the Turing machine concept and, as a result, expensive and power-hungry data transferring between the memory and CPU core is inevitable operation. Anti-Turing machine paradigm can be based on two fundamental principles: (1) data-centric computing, and (2) decentralized computing. Anti-Turing machine is able to execute a special type of programs. The commands of such program have to be addressed to the 2D or 3D persistent memory space is able to process data in-place. This program should not define the position or structure of data but it has to define the goal of data processing activity. Generally speaking, it needs to consider the whole memory space like the data transformation space. But the data placement, particular algorithm implementation, and strategy of algorithm execution are out of scope of the program.