{"id":{"repo_id":"purdue-thes","oai_identifier":"oai:docs.lib.purdue.edu:open_access_dissertations-2960"},"canonical_url":"https://search.ndltd.org/etd/purdue-thes/oai:docs.lib.purdue.edu:open_access_dissertations-2960","repository":{"repo_id":"purdue-thes","name":"Purdue University","base_url":"https://docs.lib.purdue.edu/do/oai/"},"display":{"title":"Memory Subsystems for Security, Consistency, and Scalability","abstract":"In response to the continuous demand for the ability to process ever larger datasets, as well as discoveries in next-generation memory technologies, researchers have been vigorously studying memory-driven computing architectures that shall allow data-intensive applications to access enormous amounts of pooled non-volatile memory. As applications continue to interact with increasing amounts of components and datasets, existing systems struggle to eÿciently enforce the principle of least privilege for security. While non-volatile memory can retain data even after a power loss and allow for large main memory capacity, programmers have to bear the burdens of maintaining the consistency of program memory for fault tolerance as well as handling huge datasets with traditional yet expensive memory management interfaces for scalability. Today’s computer systems have become too sophisticated for existing memory subsystems to handle many design requirements. In this dissertation, we introduce three memory subsystems to address challenges in terms of security, consistency, and scalability. Specifcally, we propose SMVs to provide threads with fne-grained control over access privileges for a partially shared address space for security, NVthreads to allow programmers to easily leverage nonvolatile memory with automatic persistence for consistency, and PetaMem to enable memory-centric applications to freely access memory beyond the traditional process boundary with support for memory isolation and crash recovery for security, consistency, and scalability.","abstract_html":"In response to the continuous demand for the ability to process ever larger datasets, as well as discoveries in next-generation memory technologies, researchers have been vigorously studying memory-driven computing architectures that shall allow data-intensive applications to access enormous amounts of pooled non-volatile memory. As applications continue to interact with increasing amounts of components and datasets, existing systems struggle to eÿciently enforce the principle of least privilege for security. While non-volatile memory can retain data even after a power loss and allow for large main memory capacity, programmers have to bear the burdens of maintaining the consistency of program memory for fault tolerance as well as handling huge datasets with traditional yet expensive memory management interfaces for scalability. Today’s computer systems have become too sophisticated for existing memory subsystems to handle many design requirements. In this dissertation, we introduce three memory subsystems to address challenges in terms of security, consistency, and scalability. Specifcally, we propose SMVs to provide threads with fne-grained control over access privileges for a partially shared address space for security, NVthreads to allow programmers to easily leverage nonvolatile memory with automatic persistence for consistency, and PetaMem to enable memory-centric applications to freely access memory beyond the traditional process boundary with support for memory isolation and crash recovery for security, consistency, and scalability.","abstract_has_math":false,"creators":["Hsu, Terry Ching-Hsiang"],"institution":null,"degree_name":"Doctor of Philosophy (PhD)","degree_level":"Dissertation","degree_discipline":"Computer Science","degree_department":null,"school":null,"contributors":["Patrick Eugster","Xiangyu Zhang","Byoungyoung Lee"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":null,"date_issued":"","date_published":null,"updated_at":"2026-11-19T08:51:35Z","subjects":[],"languages":[],"rights":[],"rights_urls":[],"identifier_entries":[],"person_identifier_entries":[]},"links":{"outbound_url":"https://docs.lib.purdue.edu/open_access_dissertations/1734","outbound_label":"Repository record","outbound_source":"dc:identifier","record_url":"https://docs.lib.purdue.edu/open_access_dissertations/1734","record_label":"Repository record","record_kind":"record","record_source":"dc:identifier","download_url":null,"download_label":null,"download_source":null,"access_status":null,"embargo_until":null,"access_note":null},"source_record":{"url":"https://docs.lib.purdue.edu/do/oai/?verb=GetRecord&metadataPrefix=oai_etdms&identifier=oai%3Adocs.lib.purdue.edu%3Aopen_access_dissertations-2960","prefix":"oai_etdms"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Patrick Eugster","Xiangyu Zhang","Byoungyoung Lee"]},{"key":"dc:creator","label":"Author","values":["Hsu, Terry Ching-Hsiang"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.created","label":"Dc Date Created","values":["2018-05-01T07:00:00Z"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["Computer Science"]},{"key":"thesis:degree_level","label":"Degree Level","values":["Dissertation"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Doctor of Philosophy (PhD)"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["https://docs.lib.purdue.edu/open_access_dissertations/1734"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["In response to the continuous demand for the ability to process ever larger datasets, as well as discoveries in next-generation memory technologies, researchers have been vigorously studying memory-driven computing architectures that shall allow data-intensive applications to access enormous amounts of pooled non-volatile memory. As applications continue to interact with increasing amounts of components and datasets, existing systems struggle to eÿciently enforce the principle of least privilege for security. While non-volatile memory can retain data even after a power loss and allow for large main memory capacity, programmers have to bear the burdens of maintaining the consistency of program memory for fault tolerance as well as handling huge datasets with traditional yet expensive memory management interfaces for scalability. Today’s computer systems have become too sophisticated for existing memory subsystems to handle many design requirements. In this dissertation, we introduce three memory subsystems to address challenges in terms of security, consistency, and scalability. Specifcally, we propose SMVs to provide threads with fne-grained control over access privileges for a partially shared address space for security, NVthreads to allow programmers to easily leverage nonvolatile memory with automatic persistence for consistency, and PetaMem to enable memory-centric applications to freely access memory beyond the traditional process boundary with support for memory isolation and crash recovery for security, consistency, and scalability."]},{"key":"dc:title","label":"Title","values":["Memory Subsystems for Security, Consistency, and Scalability"]}]}],"canonical_facts":{"dc:contributor":["Patrick Eugster","Xiangyu Zhang","Byoungyoung Lee"],"dc:creator":["Hsu, Terry Ching-Hsiang"],"dc:date.created":["2018-05-01T07:00:00Z"],"dc:description.abstract":["In response to the continuous demand for the ability to process ever larger datasets, as well as discoveries in next-generation memory technologies, researchers have been vigorously studying memory-driven computing architectures that shall allow data-intensive applications to access enormous amounts of pooled non-volatile memory. As applications continue to interact with increasing amounts of components and datasets, existing systems struggle to eÿciently enforce the principle of least privilege for security. While non-volatile memory can retain data even after a power loss and allow for large main memory capacity, programmers have to bear the burdens of maintaining the consistency of program memory for fault tolerance as well as handling huge datasets with traditional yet expensive memory management interfaces for scalability. Today’s computer systems have become too sophisticated for existing memory subsystems to handle many design requirements. In this dissertation, we introduce three memory subsystems to address challenges in terms of security, consistency, and scalability. Specifcally, we propose SMVs to provide threads with fne-grained control over access privileges for a partially shared address space for security, NVthreads to allow programmers to easily leverage nonvolatile memory with automatic persistence for consistency, and PetaMem to enable memory-centric applications to freely access memory beyond the traditional process boundary with support for memory isolation and crash recovery for security, consistency, and scalability."],"dc:identifier":["https://docs.lib.purdue.edu/open_access_dissertations/1734"],"dc:title":["Memory Subsystems for Security, Consistency, and Scalability"],"thesis:degree_discipline":["Computer Science"],"thesis:degree_level":["Dissertation"],"thesis:degree_name":["Doctor of Philosophy (PhD)"]},"updated_at":"2026-11-19T08:51:35Z"}