From Concurrent to Parallel

[@SourabhBhat]

This talk is relevant not only for Java but CS in general. Thanks.

[@usun_current5786]

it's even relevant for management and planning, where people or teams are "cores" of sorts and functional scope goes into "streams"

[@gsit80]

This man is the future of Java. Valhalla

[@ShuaibShubby]

This talk make value Elixir and Erlang more. Thanks for everyone that created those tool

[@stoneliu5575]

Thanks for the course.

[@nO_d3N1AL]

Excellent talk, very useful

[@weiliang8653]

Fantastic!!!! Where could i get this PPT?

[@DavidSaintloth]

When he talked about the incompetence of us software engineers to feeding cores with data to process in parallel I immediately see machine learning models being used here to learn a continuous way to optimize distributing instructions to an arbitrary number of cores on a chip to prevent performance fall out from software choices taken at the high level abstraction of the developer. I wonder if any one is starting to look at this as a solution?

[@ITech2005]

David Saintloth For too many reasons to list here, this would be pointless.

[@tibordigana2551]

@@ITech2005 David is really right and the architect mentioned that as well. The problem is with *unparking* the threads. And it is also in the Fork-Join as well. This unparking may take several microseconds which means that without it you could execute thousands of native operations of computing in some algorithm. Therefore they said in another kino that Oracle is preparing a new Thread API and they want to use advantages of moredern Kernel and h/w.

[@atttpww]

Excellent

[@morthim]

D: my watch later playlist deleted my comment.

[@abdelghaniainouss]

This due is great !

[@mukeshbarman]

He is actually a Legend!