A Hybrid-Domain Floating-Point Compute-in-Memory Architecture for Efficient Acceleration of High-Precision Deep Neural Networks

Documenti analoghi

• Unicorn-CIM: Uncovering the Vulnerability and Improving the Resilience of High-Precision Compute-in-Memory
  di: Li, Qiufeng, et al.
  Pubblicazione: (2025)
• SafeCiM: Investigating Resilience of Hybrid Floating-Point Compute-in-Memory Deep Learning Accelerators
  di: Bhattacharya, Swastik, et al.
  Pubblicazione: (2025)
• TimeFloats: Train-in-Memory with Time-Domain Floating-Point Scalar Products
  di: Hashem, Maeesha Binte, et al.
  Pubblicazione: (2024)
• MXFormer: A Microscaling Floating-Point Charge-Trap Transistor Compute-in-Memory Transformer Accelerator
  di: Karfakis, George, et al.
  Pubblicazione: (2026)
• Voxel-CIM: An Efficient Compute-in-Memory Accelerator for Voxel-based Point Cloud Neural Networks
  di: Lin, Xipeng, et al.
  Pubblicazione: (2024)