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:20260422T000712Z
LOCATION:503-504
DTSTART;TZID=America/Denver:20231112T162000
DTEND;TZID=America/Denver:20231112T164000
UID:submissions.supercomputing.org_SC23_sess429_ws_shpc101@linklings.com
SUMMARY:Analyzing the Performance Impact of HPC Workloads with Gramine+SGX
  on 3rd Generation Xeon Scalable Processors
DESCRIPTION:Shinobu Miwa (University of Electro-Communications, Japan) and
  Shin'ichiro Matsuo (Georgetown University)\n\nBoth 3rd generation Xeon sc
 alable processors and Gramine 1.0, which potentially improves the performa
 nce of Intel SGX, were released in 2021. In this paper, we provide the fir
 st performance analysis of HPC workloads with Gramine and SGX on 3rd gener
 ation Xeon scalable processors. Our analysis starts with some microbenchma
 rks and is then extended to various HPC workloads. Our experimental result
 s show that Gramine+SGX shows a small performance overhead (4-17%) for bot
 h compute-intensive and memory-bandwidth-sensitive workloads but a bit lar
 ge performance overhead (up to 170%) for a memory-latency-sensitive worklo
 ad. In addition, we show that the combination of Gramine and a 3rd generat
 ion Xeon scalable processor shows a slowdown of 1.5x on average (up to 4.4
 x) for many HPC workloads. This number is an order of magnitude smaller th
 an that reported in the previous work using the combination of the former 
 generation SGX toolchain and processor.\n\nTag: Distributed Computing, Sec
 urity\n\nRegistration Category: Workshop Reg Pass\n\nSession Chairs: nael 
 abu-ghazaleh (University of California, Riverside); Kevin J. Barker (Pacif
 ic Northwest National Laboratory (PNNL)); Yang Guo (National Institute of 
 Standards and Technology (NIST)); Andres Marquez (Pacific Northwest Nation
 al Laboratory (PNNL)); and Sean Peisert (Lawrence Berkeley National Labora
 tory (LBNL); University of California, Davis)\n\n
END:VEVENT
END:VCALENDAR