Unlock instant, AI-driven research and patent intelligence for your innovation.

Reduced instruction set processor based on a memristor

A technology that simplifies instruction sets and processors, and is applied in instruction analysis, electrical digital data processing, instruments, etc., to achieve the effect of saving power consumption and shortening processing time

Active Publication Date: 2019-04-05
HUAZHONG UNIV OF SCI & TECH
View PDF9 Cites 6 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] Aiming at the defects of the prior art, the object of the present invention is to provide a memristor-based RISC processor, which aims to solve the power consumption and time overhead caused by frequent data movement between modern computer memory and processor The problem

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Reduced instruction set processor based on a memristor
  • Reduced instruction set processor based on a memristor
  • Reduced instruction set processor based on a memristor

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0080] The operation sequence of embodiment 1 is as follows, all bit line switches and connection switches are defaulted to off state before the operation starts:

[0081] (1.1) The row selection module selects the 6 CMUs with corresponding addresses respectively, and sets their selected bits p to 1.

[0082] (1.2) Carry out a round of marking position setting:

[0083]

[0084]

[0085] (1.3) The connection switch connects A1 and A2 according to the values ​​of connection bits c1 and c2, connects A3 and A4, and A5 and A6 are not connected with any unit; the column selection module cooperates with the row selection module to close the 11-0 bits of A1 and A2 Bit line switches, close all 32 bit line switches of A3 and A4, and close the 31st to 16th bit line switches of A5.

[0086] (1.4) After the voltage application module finds all units with p=1, it applies the voltage V to A1 and A3 according to the values ​​of the voltage judgment bits v1 and v2 COND , apply voltage...

Embodiment 2

[0106] The operation sequence of embodiment 2 is as follows, all bit line switches and connection switches are defaulted to off state before the operation starts:

[0107] (2.1) The row selection module selects the CMU whose on-chip address is 0x31D according to the value in the program counter (for convenience of description, it is recorded as unit A), the auxiliary unit MOV1 of the data movement auxiliary block in the auxiliary area, and the address of the instruction cache 0x000 ( For the convenience of description, it is recorded as unit I), and their selected bits p are all set to 1.

[0108] (2.2) Carry out a round of marking position setting:

[0109] CMU

c1

c2

v1

v2

A

0

0

1

1

MOV1

0

0

0

0

I

0

0

0

0

[0110] (2.3) The column selection module cooperates with the row selection module to close all 32 bit line switches of MOV1 and I.

[0111] (2.4) The voltage application module applies volta...

example 2

[0121] The instruction fetching stage shown in Example 2 is not only an operation step of data movement, but also an operation step of the instruction fetching stage of all instructions.

[0122] From Example 1 and Example 2, in addition to the initial address generation, selection, and final deselection, a (or concurrent) implication / clearing / setting operation will go through the following steps: mark position setting→closing Make / Bit Line Switch → Apply Voltage to Include / Clear / Set → Open Make / Bit Line Switch. In the subsequent instruction execution examples, the control steps such as setting the mark position, closing the bit line / connection switch, and disconnecting the bit line / connection switch are omitted, and only the initial address generation, selection, and completion of the command function are given. Elapsed imply / clear / set operations, and finally deselect operations.

[0123] After the instruction fetching stage, enter the instruction decoding and execution stag...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a reduced instruction set processor based on a memristor. The memristor is a nonvolatile device which uses a resistor to store '0' 1 'logic. Meanwhile, the memristor can realize'implication logic 'by applying a pair of voltages VCOND / VSET, and various data operations, logical operations and arithmetic operations can be realized based on the implication logic. Therefore, thememristor is a computing storage fusion device with great potential, and the computer processor based on the memristor also becomes the research direction of the next generation of computer processor. The computer processor based on the memristor is designed according to the characteristic of storage, calculation and fusion of the memristor, the novel processor is different from a framework thata traditional computer must use a special memory and an arithmetic unit, and the computer processor is the processor integrating calculation and storage. Compared with a traditional computer, the speed, the parallelism degree and the power consumption of the computer processor based on the memristor are greatly improved.

Description

technical field [0001] The invention belongs to the field of computer processors, in particular to a reduced instruction set processor based on memristors. Background technique [0002] All modern computers follow the design principles of the von Neumann architecture, in which computing and storage modules are separated and connected by a bus. While the computer is running, instructions and data are continuously transported between the processor and the memory via the bus. However, when the computer is running, the overhead for moving data between the processor and the storage system is far greater than the actual computing overhead of the processor, and these overheads include power consumption, time, and space. This frequent and expensive data movement is considered to be the "von Neumann bottleneck". As long as the computer still adopts the von Neumann architecture, the von Neumann bottleneck will always exist. It can be predicted that as time goes by, the von Neumann b...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06F9/30
CPCG06F9/30134G11C13/0007G11C11/005G06F9/30145G11C13/0026G11C13/0028G11C13/0038
Inventor 刘群王业旺缪向水鲁宏伟李坚
Owner HUAZHONG UNIV OF SCI & TECH