Changes

based on the [[Shared Hash Table|shared hash table]] approach of a parallel search to profit from probing hash entries written by other instances of the search, as already used by [[Vincent David|Vincent David's]] αβ* <ref> [[Vincent David]] ('''1993'''). ''[http://cat.inist.fr/?aModele=afficheN&cpsidt=161774 Algorithmique parallèle sur les arbres de décision et raisonnement en temps contraint. Etude et application au Minimax]'' = Parallel algorithm for heuristic tree searching and real-time reasoning. Study and application to the Minimax, Ph.D. Thesis, [https://en.wikipedia.org/wiki/%C3%89cole_nationale_sup%C3%A9rieure_de_l%27a%C3%A9ronautique_et_de_l%27espace École nationale supérieure de l'aéronautique et de l'espace], [https://en.wikipedia.org/wiki/Toulouse Toulouse], [https://en.wikipedia.org/wiki/France France]</ref>. Multiple [[Process|processes]] or [[Thread|threads]] search the same [[Root|root position]], but are launched with different [[Depth|depths]], and/or varying [[Move Ordering|move ordering]] at the root node, to statistically improve the gains from the effect of nondeterminism, otherwise depending on random timing fluctuations <ref>[http://www.talkchess.com/forum/viewtopic.php?t=59389&start=5 Re: Lazy SMP - how it works] by [[Evert Glebbeek]], [[CCC]], October 19, 2016</ref>. The term '''Lazy SMP''' was coined by [[Julien Marcel]] end of 2012 <ref>[http://www.talkchess.com/forum/viewtopic.php?t=46597 Lazy SMP] by [[Julien Marcel]], [[CCC]], December 27, 2012</ref> with further elaborations by [[Dan Homan|Daniel Homan]] <ref>[http://www.talkchess.com/forum/viewtopic.php?t=46858 Lazy SMP, part 2] by [[Dan Homan|Daniel Homan]], [[CCC]], January 12, 2013</ref> <ref>[http://www.talkchess.com/forum/viewtopic.php?t=48536 Lazy SMP and Work Sharing] by [[Dan Homan|Daniel Homan]], [[CCC]], July 03, 2013</ref>, [[Martin Sedlak]] and others. Today, many chess programs use this easy to implement parallel search approach, which [https://en.wikipedia.org/wiki/Scalability scales] surprisingly well up to 8 cores and beyond <ref>[http://www.talkchess.com/forum/viewtopic.php?t=55170&start=11 Re: A new chess engine : m8 (comming not so soon)] by [[Peter Österlund]], [[CCC]], February 01, 2015</ref>, not only in [[Nodes per secondSecond|nodes per second]] (as expected), but in [[Playing Strength|playing strength]], while it seems worse than [[Young Brothers Wait Concept|YBW]] in [https://en.wikipedia.org/wiki/Speedup speedup] concerning time to [[Depth|depth]] <ref>[http://www.talkchess.com/forum/viewtopic.php?t=61131 Time to depth concerns] by [[Carl Bicknell]], [[CCC]], August 15, 2016</ref>. Notably [[Stockfish|Stockfish 7]], released in January 2016, switched from YBW to lazy SMP <ref>[http://www.talkchess.com/forum/viewtopic.php?t=58779 Stockfish 7] by [[Joona Kiiski]], [[CCC]], January 02, 2016</ref> <ref>[http://www.talkchess.com/forum/viewtopic.php?t=58887 Threads test incl. Stockfish 7] by [[Andreas Strangmüller]], [[CCC]], January 11, 2016</ref> <ref>[http://www.talkchess.com/forum/viewtopic.php?t=58935 Stockfish 7 progress] by Carl Lumma, [[CCC]], January 16, 2016</ref>.

: [http://www.talkchess.com/forum/viewtopic.php?t=64824&start=2 SMP NPS measurements] by [[Peter Österlund]], [[CCC]], August 06, 2017 » [[Nodes per secondSecond]]