PROJECT TITLE :

Highly Parallelized Pattern Matching Hardware for Fast Tracking at Hadron Colliders

ABSTRACT:

A high-performance “pattern matching” implementation primarily based on the Associative Memory (AM) system is presented. It's designed to solve the important-time hit-to-track association downside for particles made in high-energy physics experiments at hadron colliders. The processing time of pattern recognition in CPU-based mostly algorithms will increase rapidly with the detector occupancy because of the limited computing power and input-output capacity of hardware obtainable out there. The AM system presented here solves the problem by having the ability to process even the most complicated hadron collider events created at a rate of one hundred kHz with a median latency smaller than . The board built for this goal is in a position to execute petabyte comparisons per second, with peak power consumption below 250 W, uniformly distributed on the large area of the board.