You are invited to join our interactive event including the following highlights:
Register here. Registration is necessary to attend.
This event is jointly organized by HPC@HRZ, the HPC/A Lab and TRA Modelling.
Update March 7: Here are the seven finalists, as selected by the jury on March 6, in alphabetical order:
We congratulate the finalists and would like to thank all submitters for their contributions. You either have been or will soon be notified with further information.
Update March 5: Abstract submission is now closed. You can still register to attend however.
Want to enter the competition?
Submit an abstract about your most interesting result generated with Marvin!
Win prices such as a temporarily higher compute priority or premium support from the HPC Team!
Click on "Call for Abstracts" on the left for more information.
Tim Mattson is a parallel programmer obsessed with every variety of science (Ph.D. Chemistry, UCSC, 1985). In 2023 he retired after a 45-year career in HPC (30 of which were with Intel). He has had the privilege of working with people much smarter than himself on great projects including: (1) the first TFLOP computer (ASCI Red), (2) Parallel programming languages … Linda, Strand, MPI, OpenMP, OpenCL, OCR and PyOMP (3) two different research processors (Intel's TFLOP chip and the 48 core SCC), (4) Data management systems (Polystore systems and Array-based storage engines), and (5) the GraphBLAS API for expressing graph algorithms as sparse linear algebra. Tim has over 150 publications including six books on different aspects of parallel computing.
Hardware trends are clear. Driven by economics and the need to deliver increasing performance within a fixed power budget, computer systems are becoming increasingly complex. This complexity is directly managed by software, hence, the need for programmers with a detailed understanding of computer architecture.
Unfortunately, programmers today are trained with programming languages that hide the hardware. You can't specialize an algorithm to hardware features if there is a virtual machine between your code and the system or if you program in an interpreted language (such as python).
How are we going bridge this disconnect between our processors, the people who write our software, and the programming languages they use? We must fundamentally change how we construct software. We must automate key steps in software development using machine learning and Al technologies to map code onto the details of different systems.
In this talk, after describing the fundamentals of hardware evolution, we'll explore research to automate key aspects of software development. We will describe successes and reasons for hope, but also fundamental challenges that limit the applicability of Al to address this problem.