Implementation of a Vector Data Model Using SIMD Registers

Abstract

The article considers the possibility of implementing a vector data model using SIMD registers. The need to solve this problem stems from the fact that humanity is on the verge of transition from conventional computers to their quantum counterparts. Emulators are not well-developed for free work on quantum algorithms. However, the need to build quantum algorithms is growing with each new step in the development of a full-fledged public quantum computer. The article proposes a method for implementing a vector data model using SIMD registers. The closest to a parallel vector implementation is a hypothetical implementation of a quantum computer, so it can serve as the basis for developing a model that meets all the criteria of a quantum model, but without using a real quantum system, which will save money and increase the efficiency of developing quantum systems. The terminology of the field of development of quantum computers is used. Today, in most sources that give a method for forming a quantum adder, we are talking about creating some kind of qubit registers for storing added numbers. That is, the number is stored in binary form in a register of qubits, each of which is in a certain quantum state, not entangled, but measured. This means that a register of this kind is inherently an ordinary bit register. This approach is relatively easy to implement, but it has no practical meaning and application, because the gain in calculation speed in this case is extremely low.

Keywords

• quantum computing
• quantum parallelism
• vector processor