IT Seminar - Monday, September 24th at 11h

Felice Pantaleo — Wed, 12 Sep 2012 10:29:33 +0000

GPU acceleration of scientific applications: An update
by Dr. Peter Messmer and Mr. Edmondo Orlotti, NVIDIA
Monday, September 24th - IT Auditorium (31-3-004) at 11h
https://indico.cern.ch/conferenceDisplay.py?confId=206360

Real-time Low Level Trigger using GPUs

Felice Pantaleo — Mon, 23 Jul 2012 10:41:16 +0000

Introduction

This pilot project within NA62 aims at integrating GPUs into the central L0 trigger processor, and also to use them as fast online processors for computing trigger primitives. Several TDC-equipped sub-detectors with sub-nanosecond time resolution will participate in the first-level NA62 trigger (L0), fully integrated with the data-acquisition system. In order to reduce the readout rate of all sub-detectors to 1 MHz, multiplicity information asynchronously computed over time frames of a few ns, both for positive sub-detectors and for vetos, will be exploited. The online use of GPUs would allow the computation of more complex trigger primitives already at this first trigger level.

Goals

Create a system that can replace FPGAs in the early trigger stages
Design and implement fast algorithms and the necessary data structures
Maximize the throughput keeping latencies low
Determine the architectural characteristics that can better dial with high throughput I/O and little time-out for data transfers and computation
Quantify the latency and throughput stability of the systems

Material

Update on Real-time Trigger using GPUs, F. Pantaleo, October 10th 2012
Real-time use of GPUs in NA62 experiment, F. Pantaleo, May 9th 2012

Forum on Concurrent Programming Models and Frameworks - CUDA