Wewnętrzne spotkanie konsorcjum PRACE-Lab.
Organizator lokalny: PCSS
Narzędzie: Webex
Dzień 1 9:00-14:00 KS 15:00
Dzień 2 9:00-14:30
Welcome for 4 days of hands-on sessions on Parallel Programming, Performance Optimization, Artificial Intelligence, and Machine & Deep Learning delivered by Intel engineers.
Explore exciting topics including scientific and technical computing, computer vision, image and pattern recognition, machine learning, optimized deep learning and big data analytics.
No special requirements and pre-work. Please bring and use your own – Intel®-based – laptop. We will provide all required access to software and technology during the session.
2020.10.26 14:00-18:00 Warsaw time2020.10.27 14:00-18:00 Warsaw time
2020.10.28 14:00-18:00 Warsaw time
2020.10.29 14:00-18:00 Warsaw time
Dates |
Topics |
Time |
|
|
|
Day 1 (26/10) |
HPC today: : technologies for fast and reliable data processing, transmission, and storage.
|
|
HW: Intel XPU overview : CPU/GPU roadmap & product overview
|
15min |
|
Intel oneAPI initiatives: Software on GPUs, workloads & containers, Transitioning from Intel® Parallel Studio XE to oneAPI
|
45min |
|
ML hands-on (1/2) |
|
|
Intel® Distribution for Python* on Intel XPU
|
1h |
|
BREAK |
15min |
|
Intel® Scalable Dataframe Compiler – effective usage of data frames in Python
|
1h |
|
Intel® oneAPI Data Analytics Library – data analytics performance library
|
1h |
|
|
|
|
Day 2 (27/10) |
ML hands-on (2/2) |
|
Optimized frameworks for ML XGBoost*, TensorFlow*, PyTorch*, etc
|
1h |
|
Intel oneAPI Collective Communications Library – using HPC clusters for DNN training
|
1h |
|
BREAK |
15min |
|
Analytics Zoo – neural network infrastructure for Spark Hadoop
|
1h |
|
Intel® openVINO toolkit – visual inference and neural network optimization toolkit
|
1.5h |
|
|
|
|
Day 3 (28/10) |
Code optimization hands-on (1/2)
|
|
Integration of Intel tools into CI/CD
|
45min |
|
Intel® Math Kernel Library , help accelerate math processing routines, increase application performance, and reduce development time
|
1h |
|
BREAK |
15min |
|
Intel DCP ++
|
1h |
|
|
|
|
Day 4 (29/10) |
Code optimization hands-on (2/2)
|
|
Offload Advisor - code modernization, programming guidance, and performance estimation tool that supports the DPC++ language and OpenMP* on CPU and GPU
|
1.5h |
|
BREAK |
15min |
|
Profiling applications with Intel® VTune™ Profiler
|
1h |
|
Intel® VTune™ Profiler - performance analysis and microarchitecture optimizations
|
1h |
|
|
|
|
Seminarium pt. "SIMULIA Optimization Day" jest organizowane przez firmę Budsoft wraz z Poznańskim Centrum Super-komputerowo Sieciowym i Instytutem Konstrukcji Budowlanych Politechniki Poznańskiej.
TARGET AUDIENCE:
Master students, PhD students, postdoctoral researchers, other students and researchers
BRIEF CONTENTS:
Workshop gives a short introduction to the purposes of HPC computing. Overview on basic principles: main components of HPC system, working with queue system. Presentation of how one can exploit potential of a supercomputer with different kinds of parallelism.
ORGANISERS:
Radoslaw Januszewski, Maciej Brzezniak
TIME SLOT:
Wednesday May 15 2019,
Room 0.31AB, 14:30 – 18:30
PREREQUISITES
● Basic knowledge of Linux command line;
● Own laptop with SSH client installed.
TARGET AUDIENCE:
Developers of services and applications for HPC, security analysts BRIEF CONTENTS: Modern authentication and authorization mechanisms ‘CTF’ (Capture The Flag) workshop, where participants will learn and design auth(z) and auth(N) and in the end... code-in (into a prepared environment):
ORGANISERS:
PSNC Cybersecurity Dept. team: Pawel Berus, Gerard Frankowski, Mikołaj Dobski, Paweł Węgrzak
TIME SLOT
Wednesday May 15 2019, Room 0.26 C, 14:30 – 18:30
PREREQUISITES
● Own laptop
● Fluent English
● Basic knowledge of Java
● Running HyperV/VirtualBox
● WiFi
On invitation only.