Scotch 7.0 published

We announce the release, as free/libre software, of version 7.0 (codename “Sankara”) of the Scotch + PT-Scotch software package. This is a major release, fruition of six years of development, which brings many innovative features.

Major new features

Thread management is now fully dynamic. Efficient multi-threaded versions of many time-consuming algorithms have been coded. These algorithms concern the Scotch (formerly) sequential library, as well as the PT-Scotch MPI-based library (which now provides hybrid parallel algorithms, mixing threads and MPI). Existing threads can be imported from the caller’s environment, to avoid creating unnecessary threads and benefit from the mapping of preexisting threads.

Execution options, such as selecting either deterministic algorithms or more efficient yet non-deterministic versions, are now also dynamic (while in previous versions, they had to be prescribed at compile time).

New objects, called contexts, allow users to launch Scotch tasks concurrently. They encapsulate a thread pool, execution options and a private pseudo-random generator, so that multiple Scotch tasks (mapping or partitioning, ordering, etc.) can be performed concurrently and reproducibly on the same graph, or on different graphs, at the same time.

More here:

Your help is needed

In 2022, the Scotch project will turn 30. In order to perpetuate software development around Scotch, a consortium is being created, to pool development goals and investments from its members so as to fund engineering staff to carry out strategic R&D.

We are currently identifying ambitious key parties interested in joining the consortium. This is a great opportunity for people and organizations having needs in fast and efficient graph partitioning to support and ensure the sustainability of the Scotch software for its users all around the world.

Would Scotch users  (or prospective users) be so kind as to fill the questionnaire below?

The preparatory document for the Scotch consortium is available here: