Paper accepted at DATE 2026: Software-Based Approximate Multiplication on Multiplierless CPUs Using Custom Instruction

Congratulations to the authors Shalu Prathmesh Rajiv and Dr. Rajesh Kedia on this excellent achievement!

We are delighted to share that the paper titled “Software-Based Approximate Multiplication on Multiplierless CPUs Using Custom Instruction” has been accepted at Design, Automation and Test in Europe (DATE 2026).

Key Highlight

This work enhances the efficiency of approximate multiplication on lightweight, multiplier-less CPUs by introducing a custom LOD (Leading-One Detector) instruction. The authors develop multiple software routines that demonstrate a clear trade-off between computation cycles and error rates, enabling efficient performance for resource-constrained systems.

Heartfelt congratulations to the authors for their valuable contribution to computer architecture research!