BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Los_Angeles
X-LIC-LOCATION:America/Los_Angeles
BEGIN:DAYLIGHT
TZOFFSETFROM:-0800
TZOFFSETTO:-0700
TZNAME:PDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0700
TZOFFSETTO:-0800
TZNAME:PST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260402T024534Z
LOCATION:3001\, Level 3
DTSTART;TZID=America/Los_Angeles:20250625T171500
DTEND;TZID=America/Los_Angeles:20250625T173000
UID:dac_DAC 2025_sess132_RESEARCH777@linklings.com
SUMMARY:SeIM: In-Memory Acceleration for Approximate Nearest Neighbor Sear
 ch
DESCRIPTION:Chaoqiang Liu (Huazhong University of Science and Technology),
  Dan Chen (National University of Singapore), Yu Huang (Huazhong Universit
 y of Science and Technology), Wenjing Xiao (Guangxi University), Haifeng L
 iu and Yi Zhang (Huazhong University of Science and Technology), Huize Li 
 (National University of Singapore), and Xiaofei Liao and Hai Jin (Huazhong
  University of Science and Technology)\n\nApproximate nearest neighbor sea
 rch (ANNS) is crucial in many applications to find semantically similar ma
 tches for user queries. Especially with the development of large language 
 models (LLMs), ANNS is becoming increasingly important in retrieval-augmen
 ted generation (RAG) technique. An in-depth analysis of ANNS reveals that 
 its diverse operations, from extensive memory access to intensive sorting,
  are key performance bottlenecks, imposing significant strain on both the 
 memory system and computing resources. \n\nBased on these observations, we
  present SeIM, a hierarchical in-memory architecture to accelerate ANNS. S
 eIM is designed to accommodate the diverse operational characteristics of 
 ANNS. Specifically, SeIM offloads highly parallel memorybound operations t
 o the memory bank level and introduces a unified execution model to reuse 
 hardware units, requiring only lightweight modifications to standard DRAM 
 architecture. Additionally, SeIM places compute-bound sorting operations, 
 which require cross-unit data access, at the memory controller level and e
 mploys an adaptive transmission filtering technique to reduce unnecessary 
 data transfers and processing during sorting. Our evaluation shows that Se
 IM achieves 268×, 22×, and 5× higher throughput, 306×, 59×, and 4× lower l
 atency, and 3081×, 287×, and 2× higher power efficiency than state-of-the-
 art CPU-, GPU-, and ASIC-based ANNS solutions.\n\nTopics: Design\n\nTracks
 : DES2B: In-memory and Near-memory Computing Architectures, Applications a
 nd Systems\n\nSession Chairs: Steve Dai (NVIDIA) and Haitong Li (Purdue Un
 iversity)\n\n
END:VEVENT
END:VCALENDAR