Token-Picker: Accelerating Attention in Text Generation with Minimized Memory Transfer via Probability Estimation

Documenti analoghi

• PIM-MMU: A Memory Management Unit for Accelerating Data Transfers in Commercial PIM Systems
  di: Lee, Dongjae, et al.
  Pubblicazione: (2024)
• IANUS: Integrated Accelerator based on NPU-PIM Unified Memory System
  di: Seo, Minseok, et al.
  Pubblicazione: (2024)
• SAL-PIM: A Subarray-level Processing-in-Memory Architecture with LUT-based Linear Interpolation for Transformer-based Text Generation
  di: Han, Wontak, et al.
  Pubblicazione: (2024)
• HURRY: Highly Utilized, Reconfigurable ReRAM-based In-situ Accelerator with Multifunctionality
  di: Shin, Hery, et al.
  Pubblicazione: (2024)
• DiSC: Resolution-Scalable Acceleration of Diffusion Models by Exploiting Sparsity and Cached Token Reuse with Hash-based Distribution
  di: Yoon, Jieon, et al.
  Pubblicazione: (2026)