If you are less familiar with computer hardware terminology, this glossary might help.
As we need computers to run our calculations, it is appropriate to have a look at the computer landscape. In this video, you’ll learn what it takes to make a fast computer, and how you can position your resources against what is available at the international level.
What’s the fastest computer on Earth? That’s a moving target. Here is some information about the current fastest one (2022).
What is the fastest computer you can have access to? That will probably be your university compute cluster, or your regional or national computing center. Students from Flanders can have a look at the website of the Flemish Supercomputing Centre, in particular at the specifications for the TIER1 (national level) and TIER2 (university level) computing resources. If you are based elsewhere, search similar information for your relevant computing resources.
Having done this, you’re ready for these two tasks:
Describe the computing system you personally have access to for your ab initio calculations later in this course (type of cpu, clockspeed, RAM, number of cores). This may be your own laptop, or a computing cluster at your site. Put the result in this forum (and peek to the resources your fellow students have access to):
Started by: aldairpaz1809
Started by: DavideP
Started by: basnetsanket
Started by: sara_shahbazi
Started by: FANYANG
Started by: StijnVanVooren
Started by: nkechioffiakalu
Started by: George-Andreev
Started by: ianlesnik
Started by: Yilak Alemu Abbo
Started by: hrijckae
Started by: y.mohammadi@cfu.ac.ir
Started by: Menna ghannam
Started by: k.ozmert
Started by: Nan
For those who have access to a university, institute or government computing cluster: try to find out its performance specifications (in TFlop/s). At www.top500.org you find a historical archive with the Top500 lists going back to June 2005. Find out in which year – if any – your computing cluster would have been the fastest one on Earth.
Started by: Tristan Boen
Started by: Ruben Quiroz
Started by: cm
Started by: Dietger Van den Eynden
Started by: Adedeji Akindadelo
Started by: Cedric
Started by: Arthur Adriaens
expected time: 20m
report time spent (page code AW01G)