Chairs

Introduction

The rapid scaling of large language models and data-intensive applications is shifting the bottleneck of modern computing systems from raw compute toward the memory hierarchy: bandwidth, capacity, data movement, and the dependability of emerging memory technologies. This special track focuses on memory-centric, reliable, and efficient computing across the full system stack, spanning processing-in-memory and near-data architectures, data-movement optimization for memory-limited accelerators, memory management for long-context LLM serving, model compression and approximation, and cross-layer reliability analysis under hardware non-idealities. We welcome contributions that connect devices and memory technologies with architectures, runtimes, algorithms, and applications, including analytical models, adaptive methods, system prototypes, and evaluations on real or representative workloads. The track brings together complementary work on near-cache and in-DIMM processing, KV-cache and attention optimizations for efficient LLM inference, LLM compression, hyper-dimensional computing, ReRAM reliability, and approximate GPU graph processing, aiming at a coherent view of performance–energy–accuracy–reliability trade-offs for next-generation AI and data-intensive systems.

Topics

• Processing-in-memory and near-data computing (PIM, in-DIMM processing, ReRAM/computing-in-memory)
• Memory-system and data-movement optimization for memory-limited accelerators (near-cache copy, GEMM data loading)
• Efficient LLM inference and serving: KV-cache management, long-context and multi-turn conversation, attention mechanisms
• LLM compression, quantization, and codebook-free coding techniques
• Cross-layer reliability analysis and fault tolerance under hardware non-idealities
• Approximate computing and adaptive accuracy–performance–energy trade-offs
• Hyper-dimensional computing and neural optimization methods
• GPU and heterogeneous acceleration for graph and data-intensive workloads

Important Date

We are pleased to announce the important dates for PDCAT 2026. Please mark your calendars!
All deadlines are based on Anywhere on Earth (AoE), at midnight on the specified date.

• Paper Submission Deadline: 31 August, 2026  21 September, 2026 (extended)
• Notification of Acceptance: 24 October, 2026
• Camera-Ready Submission Deadline: 14 November, 2026
• Conference Dates: 4-6 December, 2026