BEGIN:VCALENDAR VERSION:2.0 PRODID:IEEE vTools.Events//EN CALSCALE:GREGORIAN BEGIN:VTIMEZONE TZID:America/Denver BEGIN:DAYLIGHT DTSTART:20260308T030000 TZOFFSETFROM:-0700 TZOFFSETTO:-0600 RRULE:FREQ=YEARLY;BYDAY=2SU;BYMONTH=3 TZNAME:MDT END:DAYLIGHT BEGIN:STANDARD DTSTART:20251102T010000 TZOFFSETFROM:-0600 TZOFFSETTO:-0700 RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=11 TZNAME:MST END:STANDARD END:VTIMEZONE BEGIN:VEVENT DTSTAMP:20260209T201505Z UID:36F7DB6C-E3AD-42F2-9F05-D5BE57382E97 DTSTART;TZID=America/Denver:20260206T160000 DTEND;TZID=America/Denver:20260206T170000 DESCRIPTION:Emerging resistive memory technologies provide a unique opportu nity to unify computation\, storage\, security\, and data generation withi n a single hardware substrate. This talk introduces a comprehensive hardwa re platform based on multi-oxide RRAM crossbar arrays that achieves both p recision for analog computing and stochasticity for intrinsic security and generative diversity. Leveraging 5-bit device-level analog programmabilit y and reliable 4-bit array-level weight mapping\, key primitives of memory -augmented neural networks (MANNs) are implemented\, including convolution al encoding\, locality-sensitive hashing\, and RRAM-based CAM for few-shot learning with near-software accuracy. At the same time\, inherent device variability is utilized as a high-entropy physical unclonable function for secure key generation and as a hardware-native randomness source that enh ances the diversity and perceptual realism of StyleGAN3-generated biometri c images. By integrating analog VMM\, associative memory search\, high-ent ropy randomness\, and hardware-seeded generative models within the same cr ossbar fabric\, this work demonstrates how “stochastic yet precise” me mristor arrays can provide a unified foundation for edge-intelligent\, sec ure\, and data-generating systems. Controlled multi-bit programming\, devi ce-aware learning\, and engineered randomness together enable a new class of memory-centric AI platforms that support inference\, few-shot adaptatio n\, PUF-based security\, and high-fidelity generative AI.\n\nSpeaker(s): S ungjun Kim\, \n\nVirtual: https://events.vtools.ieee.org/m/534471 LOCATION:Virtual: https://events.vtools.ieee.org/m/534471 ORGANIZER:daphnechen@asu.edu SEQUENCE:12 SUMMARY:Stochastic yet Precise: Multi-Level RRAM Crossbar Arrays for In-Mem ory Learning\, Security\, and Generative AI URL;VALUE=URI:https://events.vtools.ieee.org/m/534471 X-ALT-DESC:Description: <br /><p class="MsoNormal" style="text-align: justi fy\; text-justify: inter-ideograph\;">Emerging resistive memory technologi es provide a unique opportunity to unify computation\, storage\, security\ , and data generation within a single hardware substrate. This talk introd uces a comprehensive hardware platform based on multi-oxide RRAM crossbar arrays that achieves both precision for analog computing and stochasticity for intrinsic security and generative diversity. Leveraging 5-bit device- level analog programmability and reliable 4-bit array-level weight mapping \, key primitives of memory-augmented neural networks (MANNs) are implemen ted\, including convolutional encoding\, locality-sensitive hashing\, and RRAM-based CAM for few-shot learning with near-software accuracy. At the s ame time\, inherent device variability is utilized as a high-entropy physi cal unclonable function for secure key generation and as a hardware-native randomness source that enhances the diversity and perceptual realism of S tyleGAN3-generated biometric images. By integrating analog VMM\, associati ve memory search\, high-entropy randomness\, and hardware-seeded generativ e models within the same crossbar fabric\, this work demonstrates how &ldq uo\;stochastic yet precise&rdquo\; memristor arrays can provide a unified foundation for edge-intelligent\, secure\, and data-generating systems. Co ntrolled multi-bit programming\, device-aware learning\, and engineered ra ndomness together enable a new class of memory-centric AI platforms that s upport inference\, few-shot adaptation\, PUF-based security\, and high-fid elity generative AI.</p> END:VEVENT END:VCALENDAR