Paper by Jiménez a "Top Pick"
A paper co-authored by Daniel Ángel Jiménez, an Associate Professor in the Department of Computer Science, has been recognized as one of the most significant computer architecture papers in 2008.
The paper, "Mixed-Signal Approximate Computation: A Neural Predictor Case Study" by Renée St. Amant, Daniel A. Jiménez, and Doug Burger, was recently selected as one of the 12 papers to appear in a 2008 IEEE Micro special issue on "Top Picks from Computer Architecture Conferences." According to the IEEE Micro website, "this issue collects some of this year's most significant research publications in computer architecture based on novelty and long term impact." The initial version of the paper had been published as "Low-Power, High-Performance Analog Neural Branch Prediction" in the 2008 International Symposiumon Microarchitecture (MICRO-41). The paper demonstrates how low-power analog circuits can replace high-power digital circuits in situations where power and latency are more important than precision.
Jiménez researches computer architecture and compiler optimization.The paper continues Jiménez's branch prediction research for which he was awarded the NSF CAREER grant in 2006. The paper is co-authored with a student at the University of Texas at Austin and a researcher at Microsoft Research.