Link to the University of Pittsburgh Homepage
Link to the University Library System Homepage Link to the Contact Us Form

Derived ECC for Protection of On-Demand Data

Ollivier, SEBASTIEN (2019) Derived ECC for Protection of On-Demand Data. Master's Thesis, University of Pittsburgh. (Unpublished)

[img]
Preview
PDF
Submitted Version

Download (871kB) | Preview

Abstract

Traditional error correction coding (ECC) methodologies store data parity bits along with
data they protect. Subsequently, upon accessing the combined data and parity bits it is
possible discover and correct faults in either the data or correction bits. In domain-wall
memories (DWMs), a type of sequential access non-volatile memory related to spin-transfer
torque magnetic memory (STT-MRAM), it is not convenient or efficient to store data to
protect access errors data shifting during sequential access. To solve this problem, we propose
a new technique called derived error correction (DEC) for such cases where it is intractable
to record metadata for error correction. Instead, we rebuild the metadata on-demand and
store only the parity bits that protect the metadata. The DWM metadata is constructed
using a novel transverse read (TR). TR reads in an orthogonal direction of a traditional
DWM access point and can be used to calculate number of ones in a DWM. Faults in the
metadata correspond to shift-errors in the DWM. Using ECC, we can correct these faults
in the metadata and use these corrections to repair the shifting state of the DWM to ensure
correct operation. Through these techniques, we propose a shift-aware error correction code
that provides a lifetime over 10 years while reducing area by 2.6 and 3.7 times against state
of the art technique.


Share

Citation/Export:
Social Networking:
Share |

Details

Item Type: University of Pittsburgh ETD
Status: Unpublished
Creators/Authors:
CreatorsEmailPitt UsernameORCID
Ollivier, SEBASTIENsbo15@pitt.edusbo15
ETD Committee:
TitleMemberEmail AddressPitt UsernameORCID
Committee ChairJones, Alex K.akjones@pitt.eduakjones
Committee MemberJingtong, HuJTHU@pitt.eduJTHU
Committee MemberDickerson, Samuel .Jdickerson@pitt.edudickerson
Date: 23 January 2019
Date Type: Publication
Defense Date: 28 November 2018
Approval Date: 23 January 2019
Submission Date: 19 November 2018
Access Restriction: 1 year -- Restrict access to University of Pittsburgh for a period of 1 year.
Number of Pages: 38
Institution: University of Pittsburgh
Schools and Programs: Swanson School of Engineering > Electrical and Computer Engineering
Degree: MS - Master of Science
Thesis Type: Master's Thesis
Refereed: Yes
Uncontrolled Keywords: Memory, reliability, DWM, ECC
Date Deposited: 23 Jan 2019 16:25
Last Modified: 23 Jan 2020 06:15
URI: http://d-scholarship.pitt.edu/id/eprint/35523

Metrics

Monthly Views for the past 3 years

Plum Analytics


Actions (login required)

View Item View Item