IARC and NVIDIA’s GPU Programming and Machine Learning Workshop
MEET THE DURHAM GPU AND MACHINE LEARNING COMMUNITIES AND LEARN FROM NVIDIA TRAINERS
IARC is proud to host the first NVIDIA GPU Programming and Machine Learning workshop at Durham. Over a three-day period, trainers from NVIDIA will give hands-on demonstrations and train the Durham community to best make use of these new popular technologies.
Date: 5th – 7th December 2016
Time: 9:00 am – 17:00 am
Venue: Derman Christopherson Room, Calman Learning Centre
Target Audience: Academic Members of Staff, Post Docs and PhD Students interested in Machine Learning and GPU Programming, and students from previous IARC training sessions.
The workshop aims to establish a community of practitioners and enthusiasts at Durham. Short talks and presentations by existing Durham practitioners will interlace the workshop to give attendees a sense of what can be done with CUDA and Machine Learning. A Poster Competition will run on day 2 of the event.
Day 1 and 2 of the workshop will focus on GPU programming. The course will start off with an introduction to GPU accelerated programming and then through practical exercises attendees will learn about the hardware (architecture, memory etc.) and software environments. Day 2, again through practical exercises, focuses on streams, errors, and optimisations. A background in C programming is required for day 1 and 2.
Day 3 of the workshop focuses on Deep Learning and the DIGITS framework. A hands on exercise, with a training set, will get attendees up to speed with Deep Learning on GPU’s.
PhD/PostDoc Poster and 3-minute madness Competition: PhD Students and Postdocs are encouraged to submit posters about their research and will have 3 minutes too present. The winning submission & presentation will receive a GPU courtesy of NVIDIA. Poster contributors are not required to attend the full workshop.
Requirements: Bring your own Laptop. Familiarity with LINUX environments will be a plus but is not required. For Machine Learning no prior programming knowledge is needed. For GPU/CUDA knowledge of C Programming is required
Other Notes: Course runs for 3 days. There is a dinner/mixer on day 2. Teas, Coffees and snacks will be available throughout. Attendees will need to bring their own lunch.