TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Alphazero crushes Stockfish in 1,000 game match

48 pointsby stillsutover 6 years ago

13 comments

AndyNemmityover 6 years ago
&quot;CCC fans will be pleased to see that some of the new AlphaZero games include &quot;fawn pawns,&quot; the CCC-chat nickname for lone advanced pawns that cramp an opponent&#x27;s position. &quot;<p>The &quot;fawn pawn&quot; naming comes from fans of kingscrusher on youtube who analyzes Leela Chess Zero games. <a href="https:&#x2F;&#x2F;www.youtube.com&#x2F;user&#x2F;kingscrusher" rel="nofollow">https:&#x2F;&#x2F;www.youtube.com&#x2F;user&#x2F;kingscrusher</a><p>His accent makes him saying &quot;thorn pawn&quot; sound like &quot;fawn pawn&quot;, and thus the name has been given.<p>Here is a link to shirts he sells with &quot;fawn pawn&quot; on them. <a href="https:&#x2F;&#x2F;teespring.com&#x2F;fawn-pawn?tsmac=store&amp;tsmic=kingscrusher#pid=211&amp;cid=5291&amp;sid=front" rel="nofollow">https:&#x2F;&#x2F;teespring.com&#x2F;fawn-pawn?tsmac=store&amp;tsmic=kingscrush...</a>
sinuhe69over 6 years ago
I feel compelled to repeat the previous criticism on the games: - Stockfisch was never designed nor tested on so many cores. Running on 44 cores may degrade the performance of Stockfish. - Stockfish was designed to start a game with an opening book. In games with opening book, Stockfish has won significantly more games, too. - Stockfish 8 was not a particularly good implementation. Stockfish 9 or 10 should be a better choice, though.<p>Nevertheless, the performance of AlphaZero was impressive, especially the positional knowledge it has acquired is second none. In all existing chess engines, positional knowledge is under-represented through simple heuristics. Acquiring positional knowledge was a longtime dream of chess programmers many generations. The dream was to create an engine which plays a more human-like style of chess. AlphaZero has realized this dream and even goes beyond that: extending the humans knowledge of chess.<p>I believe the most intriguing question right now is why AlphaZero stopped to improve after 9 hours of training? It’s due to the inherent problem of chess or due to the limits of ANN? If it’s the latter, how we can breakthrough and create a new generation of engines that can even surpass AlphaZero?
评论 #18629176 未加载
dragontamerover 6 years ago
I&#x27;m glad that they took a lot of criticisms to heart this time.<p>The opening book and Syzygy Tablebases were enabled, so we&#x27;re seeing Stockfish go at full power here. The only last bit of problem is that Stockfish&#x27;s scaling isn&#x27;t very good. But there&#x27;s not much that the admins of the test can do about that.<p>This test seems fair IMO.
maxanderover 6 years ago
Do we know what hardware each was using? Aside from time, s criticism of the previous AlphaZero&#x2F;Stockfish match was that AlphaZero was using a tremendous amount of TPU power while Stockfish was running on, essentially, an average laptop.
评论 #18624481 未加载
foobawover 6 years ago
How did AlphaZero lose 6 games? Do we have an analysis of those? I&#x27;d love to see what happened.
评论 #18624498 未加载
mindgam3over 6 years ago
Previous discussion from a submission by a DeepMind engineer here: <a href="https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=18620978" rel="nofollow">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=18620978</a>
sytelusover 6 years ago
More detailed article: <a href="https:&#x2F;&#x2F;deepmind.com&#x2F;blog&#x2F;alphazero-shedding-new-light-grand-games-chess-shogi-and-go&#x2F;" rel="nofollow">https:&#x2F;&#x2F;deepmind.com&#x2F;blog&#x2F;alphazero-shedding-new-light-grand...</a><p>Paper: <a href="http:&#x2F;&#x2F;science.sciencemag.org&#x2F;content&#x2F;362&#x2F;6419&#x2F;1140&#x2F;tab-pdf" rel="nofollow">http:&#x2F;&#x2F;science.sciencemag.org&#x2F;content&#x2F;362&#x2F;6419&#x2F;1140&#x2F;tab-pdf</a>
Mtinieover 6 years ago
&gt; According to DeepMind, AlphaZero uses a Monte Carlo tree search, and examines about 60,000 positions per second, compared to 60 million for Stockfish.<p>The previous statement was talking about how much faster and more efficient AlphaZero is, but the interpretation I pickup from that sentence is the opposite. Is this a “golf score” situation where lower is better?
评论 #18625170 未加载
评论 #18626074 未加载
评论 #18626610 未加载
评论 #18631729 未加载
评论 #18624985 未加载
clksover 6 years ago
&quot;Crushes <i>an old version</i> of Stockfish.&quot; The current Stockfish 10 is said to have a &gt;100 Elo advantage over Stockfish 8.
评论 #18627129 未加载
评论 #18629000 未加载
rurbanover 6 years ago
Still, Houdini is again leading on the CCCC live championship table, with Stockfish at the #2, and the open-source AlphaZero clone lc0 at #3<p><a href="https:&#x2F;&#x2F;www.chess.com&#x2F;computer-chess-championship" rel="nofollow">https:&#x2F;&#x2F;www.chess.com&#x2F;computer-chess-championship</a>
wslhover 6 years ago
Will AlphaZero be available to more chess players to play? It would be interesting to find a blind spot in this engine in a format where humans could use their brains and more tools trying to beat it. Or is it really unbeatable?
评论 #18627097 未加载
评论 #18627105 未加载
评论 #18633971 未加载
评论 #18626973 未加载
chrisMyzelover 6 years ago
What would happen if you let AlphaZero compete against AlphaZero?
评论 #18627513 未加载
评论 #18627836 未加载
评论 #18627670 未加载
nisuniover 6 years ago
Shouldn’t we take into account the computing power used to train Alphazero?<p>I feel the comparison is a bit unfair...
评论 #18626766 未加载