A comprehensive technology agnostic RRAM characterisation protocol
A comprehensive technology agnostic RRAM characterisation protocol
Resistive switching memories, also known as memristors, have exhibited an immense potential for a wide array of applications, ranging from non-volatile memories to neuromorphic computing and reconfigurable circuits. As the scope of these applications expands there is an increasing need for a comprehensive characterisation methodology.
Towards that goal we present a characterisation routine that covers a broad range of device aspects. Our testing routine employs our in-house developed memristor characterisation tool. The proposed workflow starts with a pre-electroforming I–V in order to deduce the dominant transport mechanisms. This is followed by the electroforming process which can be carried out using either current-compliant I–V curves or pulsed voltage ramps for a compliance-free approach. After establishing a base resistance we reevaluate the transport mechanism since both the core material and the interfaces have been altered with respect to their pristine state.
Switching performance of the device is benchmarked with endurance and retention testing either in room or elevated temperatures to extrapolate the lifetime of the memory window. We then proceed to evaluate the switching dynamics of the devices by applying a biasing scheme optimiser. This allows us to determine the switching behaviour under external bias, as well as the switching polarity and operating range of the device. After these parameteres have been established we evaluate the maximum number of operationally relevant states using a bespoke routine. Finally, an analytical model8 of the response of the device can be readily extracted.
Stathopoulos, Spyros
98d12f06-ad01-4708-be19-a97282968ee6
Michalas, Loukas
25d00d54-5900-485e-bd52-d3505fe881a7
Khiat, Ali
bf549ddd-5356-4a7d-9c12-eb6c0d904050
Serb, Alexantrou
30f5ec26-f51d-42b3-85fd-0325a27a792c
Prodromakis, Themistoklis
d58c9c10-9d25-4d22-b155-06c8437acfbf
July 2018
Stathopoulos, Spyros
98d12f06-ad01-4708-be19-a97282968ee6
Michalas, Loukas
25d00d54-5900-485e-bd52-d3505fe881a7
Khiat, Ali
bf549ddd-5356-4a7d-9c12-eb6c0d904050
Serb, Alexantrou
30f5ec26-f51d-42b3-85fd-0325a27a792c
Prodromakis, Themistoklis
d58c9c10-9d25-4d22-b155-06c8437acfbf
Stathopoulos, Spyros, Michalas, Loukas, Khiat, Ali, Serb, Alexantrou and Prodromakis, Themistoklis
(2018)
A comprehensive technology agnostic RRAM characterisation protocol.
International Conference on Memristive Materials, Devices & Systems<br/>, China National Convention Center, Beijing, China.
03 - 06 Jul 2018.
Record type:
Conference or Workshop Item
(Other)
Abstract
Resistive switching memories, also known as memristors, have exhibited an immense potential for a wide array of applications, ranging from non-volatile memories to neuromorphic computing and reconfigurable circuits. As the scope of these applications expands there is an increasing need for a comprehensive characterisation methodology.
Towards that goal we present a characterisation routine that covers a broad range of device aspects. Our testing routine employs our in-house developed memristor characterisation tool. The proposed workflow starts with a pre-electroforming I–V in order to deduce the dominant transport mechanisms. This is followed by the electroforming process which can be carried out using either current-compliant I–V curves or pulsed voltage ramps for a compliance-free approach. After establishing a base resistance we reevaluate the transport mechanism since both the core material and the interfaces have been altered with respect to their pristine state.
Switching performance of the device is benchmarked with endurance and retention testing either in room or elevated temperatures to extrapolate the lifetime of the memory window. We then proceed to evaluate the switching dynamics of the devices by applying a biasing scheme optimiser. This allows us to determine the switching behaviour under external bias, as well as the switching polarity and operating range of the device. After these parameteres have been established we evaluate the maximum number of operationally relevant states using a bespoke routine. Finally, an analytical model8 of the response of the device can be readily extracted.
Text
Submitted abstract
- Accepted Manuscript
More information
Submitted date: 3 April 2018
Accepted/In Press date: 3 May 2018
Published date: July 2018
Venue - Dates:
International Conference on Memristive Materials, Devices & Systems<br/>, China National Convention Center, Beijing, China, 2018-07-03 - 2018-07-06
Identifiers
Local EPrints ID: 422746
URI: http://eprints.soton.ac.uk/id/eprint/422746
PURE UUID: f6de0398-0ba2-49fa-bae5-447edc310bb2
Catalogue record
Date deposited: 01 Aug 2018 16:30
Last modified: 15 Mar 2024 19:53
Export record
Contributors
Author:
Spyros Stathopoulos
Author:
Loukas Michalas
Author:
Ali Khiat
Author:
Alexantrou Serb
Author:
Themistoklis Prodromakis
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics