News

Category: info

From: 5 November 2019

To: 7 November 2019

Talks by Ulrich Drepper from Red Hat on machine learning and systems engineering

Red Hat, in cooperation with FIT, prepared two talks by Ulrich Drepper and his colleagues on the following topics:

1. "Introduction to Machine Learning and its Application in Systems Engineering".
Date and time: 5.11.2019 at 15:00 (approx. until 17:00)
Registration: https://forms.gle/pPEDJwGXhr4FdGMH8
Abstract: "With the advent of cheap and fast computational power, large data-sets, as well as algorithmic advances, machine learning is being adopted by both industry and scientific fields at an accelerating pace. Red Hat has invested resources in using machine learning for applications in finance and accounting, sales, better customer service but also in systems engineering. A modern computing stack is replete with sophisticated hardware and software that interact in complex ways and rely on heuristics or rules-of-thumb for making many decisions. Can we replace these heuristics by learned models? Can these models result in efficiencies in both performance as well as energy consumption? This talk introduces some basics of machine learning followed by a tour through some interesting problems at the interface of systems engineering and machine learning."

The talk is primarily designated for students.

2. "Software-Configured Compute Environments"
Date and time: 7.11.2019 at 15:00 (approx. until 17:00)
Registration: https://forms.gle/r9oRUseBZVhz7HZE6
Abstract: "Hardware and software environments are designed as a compromise between many different requirements. This sacrifices performance, among other aspects, while at the same time the need for compute increases.
Specialists can certainly create more optimized systems. The challenge is to automate this.
To research these new systems we need hardware specialists to create re-configurable processors, compiler writers to deduce the best architecture from source code and generate configurations for hardware and OS, improved OSes to efficiently run that code. All that while preserving API and ideally ABI compatibility. First steps toward this are already on the way in the Boston University/Red Hat collaboration with the OpenShell project to utilize FPGAs as a suitable platform and the Linux-based unikernel project to optimize the runtime environment."

The talk is designated both for researchers, as well as for students.

Both talks will be held in the Q304 room. Since the room capacity is limited, please use the attached registration forms. The talks will be in English.

Author: Malík Viktor, Ing.

Last modified: 2019-10-23 07:38:48

Back to top