BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Denver
X-LIC-LOCATION:America/Denver
BEGIN:DAYLIGHT
TZOFFSETFROM:-0700
TZOFFSETTO:-0600
TZNAME:MDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0600
TZOFFSETTO:-0700
TZNAME:MST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260422T000711Z
LOCATION:403-404
DTSTART;TZID=America/Denver:20231116T163000
DTEND;TZID=America/Denver:20231116T170000
UID:submissions.supercomputing.org_SC23_sess163_pap580@linklings.com
SUMMARY:DGAP: Efficient Dynamic Graph Analysis on Persistent Memory
DESCRIPTION:Abdullah Al Raqibul Islam and Dong Dai (University of North Ca
 rolina, Charlotte)\n\nDynamic graphs have grown in importance for numerous
  real-world applications. To accommodate this, graph frameworks, particula
 rly their internal data structures, must support both persistent graph upd
 ates and rapid graph analysis simultaneously. Emerging persistent memory t
 echnologies, such as Optane DCPMM, offer a promising choice to simplify th
 e designs by providing data persistence, low latency, and high IOPS togeth
 er. We propose DGAP, a framework for efficient dynamic graph analysis on p
 ersistent memory. DGAP utilizes mutable Compressed Sparse Row (CSR) with n
 ew designs for persistent memory to construct the framework. Specifically,
  DGAP introduces a per-section edge log to reduce write amplification; a p
 er-thread undo log to enable high-performance, crash-consistent rebalancin
 g operations; and a data placement schema to minimize in-place updates. Ou
 r extensive evaluation results demonstrate that DGAP can achieve up to 3.2
 x better graph update performance and up to 3.77x better graph analysis pe
 rformance compared to state-of-the-art dynamic graph frameworks for persis
 tent memory.\n\nTag: Architecture and Networks, Data Movement and Memory, 
 Graph Algorithms and Frameworks, Performance Measurement, Modeling, and To
 ols, Programming Frameworks and System Software\n\nRegistration Category: 
 Tech Program Reg Pass\n\nReproducibility Badges: Artifact Available, Artif
 act Functional, Results Reproduced\n\nSession Chair: Mahantesh Halappanava
 r (Pacific Northwest National Laboratory (PNNL))\n\n
END:VEVENT
END:VCALENDAR