Functional Demonstration of a Memristive Arithmetic Logic Unit (MemALU) for In‐Memory Computing

Abstract

The development of in-memory computing has opened up possibilities to build next-generation non-von-Neumann computing architecture. Implementation of logic functions within the memristors can significantly improve the energy efficiency and alleviate the bandwidth congestion issue. In this work, the demonstration of arithmetic logic unit functions is presented in a memristive crossbar with implemented non-volatile Boolean logic and arithmetic computing. For logic implementation, a standard operating voltage mode is proposed for executing reconfigurable stateful IMP, destructive OR, NOR, and non-destructive OR logic on both the word and bit lines. No additional voltages are needed beyond "V P " and its negative component. With these basic logic functions, other Boolean functions are constructed within five devices in at most five steps. For arithmetic computing, the fundamental functions including an n-bit full adder with high parallelism as well as efficient increment, decrement, and shift operations are demonstrated. Other arithmetic blocks, such as subtraction, multiplication, and division are further designed. This work provides solid evidence that memristors can be used as the building block for in-memory computing, targeting various low-power edge computing applications.

90% yield production of polymer nano-memristor for in-memory computing

“…Functional demonstration of the arithmetic-logic module, which is the core unit of a computer CPU, is also made possible by the combination of the NAND, OR, AND, and NOR operators with a 4 (bit line, BL) × 3 (word line, WL) PBDTT-BQTPA memristor array 44 , 45 . Figure 5 a, b show the design and memristor array representation of the parallel 1-bit full adder circuit involving only five polymer devices.…”

Memristive Stateful Logic for Edge Boolean Computers

“…The next step is cascading the gates for computing. Many studies have demonstrated various logic circuits, such as an adder, [34,35,37,40,44,47,49,55,57,[60][61][62][63][64][65][66][67][68][69] multiplier, [37,55,[70][71][72] multiplexer, [73][74][75][76] other combinational logic circuits, [49,[77][78][79] and sequential logic circuits. [29,80,81] These operations can be used to implement image processing, [82] neural networks, [83][84][85] and hyperdimensional computing [43,46,86] in in-memory computing.…”

Stretchable and Twistable Resistive Switching Memory with Information Storage and Computing Functionalities

“…[6][7][8] Taking consideration of both the conventional silicon memory cells and the newly invented information storage technologies, resistive switching memory (also known as resistive random access memory or RRAM in short) can be rationally marked as the most promising stretchable information storage technology due to the structural simplicity of a resistor and the prodigious range of materials selection, in addition to the performance merits of high speed, low power, and intrinsic nonvolatility. [9][10][11][12][13][14][15] Moreover, the logic-in-memory functionality of resistive switching memory enables data computation directly in the memory array, [16,17] thus holding the great potential to simplify the circuit architecture and reduce the energy consumption of the entire wearable electronic system.…”