Differentiable Content Addressable Memory with Memristors. (2nd March 2022)

• Record Type: Journal Article
• Authors: Pedretti, Giacomo; Graves, Catherine E.; Van Vaerenbergh, Thomas; Serebryakov, Sergey; Foltin, Martin; Sheng, Xia; Mao, Ruibin; Li, Can; Strachan, John Paul
• Is Part Of: Advanced Electronic Materials. Volume 8:Number 8(2022)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2022-03-02
• Subjects: analog computing -- content addressable memories -- in‐memory computing -- memristor; Materials -- Electric properties -- Periodicals; Materials science -- Periodicals; Magnetic materials -- Periodicals; Electronic apparatus and appliances -- Periodicals; 537
• DOI: 10.1002/aelm.202101198
• Languages: English
• ISSNs: 2199-160X

Abstract

Abstract: Memristors, Flash, and related nonvolatile analog device technologies offer in‐memory computing structures operating in the analog domain, such as accelerating linear matrix operations in array structures. These take advantage of analog tunability and large dynamic range. At the other side, content addressable memories (CAM) are fast digital lookup tables which effectively perform nonlinear Boolean logic and return a digital match/mismatch value. Recently, nonvolatile analog CAMs have been presented merging analog storage and analog search operations with digital match/mismatch output. However, CAM blocks cannot easily be inserted within a larger adaptive system due to the challenges of training and learning with binary outputs. Here, a missing link between analog crossbar arrays and CAMs, namely a differentiable content addressable memory (dCAM), is presented. Utilizing nonvolatile memories that act as a "soft" memory with analog outputs, dCAM enables learning and fine‐tuning of the memory operation and performance. Four applications are quantitatively evaluated to highlight the capabilities: improved data pattern storage, improved robustness to noise and variability, reduced energy and latency performance, and an application to solving Boolean satisfiability optimization problems. The use of dCAM is envisioned as a core building block of fully differentiable computing systems employing multiple types of analog compute operations and memories. Abstract : A differentiable content addressable memory (dCAM) is presented that performs operations in between analog crossbar arrays and digital output content addressable memory. By using analog nonvolatile devices, the dCAM acts as a soft memory with analog outputs and enables learning and fine‐tuning of the memory operation and performance. Four applications of the dCAM are shown to highlight the capabilities.