ICGA WCCC and WCSC in Santiago de Compostela

[Gian-Carlo Pascutto]

Software tournament finished. Hardware provided was Intel i5-14400 with 16GB RAM. There was no GPU in the machines so in this tournament I played with Deep Sjeng, not Stoofvlees as the tournament site erroneously claims.

Rofchade is the World Chess Software Champion.

Code: Select all

Competition name - University of Santiago de Compostela - Software World Chess Championship

Ranking after Round 9 - 19-10-2024 - Normal ranking

Pos   Name                  Gms   W   D   L   Score   SBgr    Rtg   TPR   SRtg   Playoff
----------------------------------------------------------------------------------------
1     Rofchade              8     3   5   0   5.5     17.75   0     0     0      WINNER 1.5-0.5
2     Fritz                 8     3   5   0   5.5     17.75   0     0     0      
3     Raptor                8     2   6   0   5.0     17.25   0     0     0      
4     Deep Sjeng            8     2   6   0   5.0     16.25   0     0     0      
5     Sashchess/Stockfish   8     2   6   0   5.0     16.25   0     0     0      
6     Jonny                 8     2   6   0   5.0     16.25   0     0     0      
7     Tornado               8     2   3   3   3.5      9.00   0     0     0      
8     Ares                  8     1   1   6   1.5      2.50   0     0     0      
9     Tech 4                8     0   0   8   0.0      0.00   0     0     0    

[Viz]

Stockfish clone not winning, amchess at it best.

[mclane]

Software tournament finished. Hardware provided was Intel i5-14400 with 16GB RAM. There was no GPU in the machines so in this tournament I played with Deep Sjeng, not Stoofvlees as the tournament site erroneously claims.

Rofchade is the World Chess Software Champion.

Congratulations to the new software champion.
What a great tournament. And the presentation of it in the net was very good.
Thanks to alle the other participants. You made all a cool tournament.

[hgm]

Then tell me what your 8-game tournament is supposed to measure? CPU noise?

How reliably t he engine performs?

I will bet my $1000, put SF in the pool of those engines in 8 game sample. If SF did not win the tournment , take my money.

Somebody already claimed the $1000?

[Rebel]

A world championship without a GPU?

Where the games single threaded also?

[hgm]

You can use the entire machine. Which appears to have 6 'performance cores' (= 12 hyper threads), plus 4 'efficiency cores'. So I assume the participants either use 6 or 12 search threads.

[Gian-Carlo Pascutto]

A world championship without a GPU?

As already explained, this was the Software tournament. I don't think they've ever had decent enough GPUs in the provided machines to be worth using. We'll have to wait until the sponsor is a gaming PC company instead of someone supplying business machines, or - in this case - the desktops provided to the students on campus.

The WCCC starts tomorrow.

[Gian-Carlo Pascutto]

You can use the entire machine. Which appears to have 6 'performance cores' (= 12 hyper threads), plus 4 'efficiency cores'. So I assume the participants either use 6 or 12 search threads.

4, 6, 10 or 16 were common. The "efficiency cores" on the Raptor Lake aren't that bad so I was surprised some people stayed on less than 10 cores (Fritz played with 6). But the machines being business PCs in small cases, clockspeed did drop very quickly to ~3Ghz as the machines essentially overheated.

I can't imagine many situations where you'd prefer 12 over 10 or 16?

[Rebel]

A world championship without a GPU?

As already explained, this was the Software tournament. I don't think they've ever had decent enough GPUs in the provided machines to be worth using. We'll have to wait until the sponsor is a gaming PC company instead of someone supplying business machines, or - in this case - the desktops provided to the students on campus.

The WCCC starts tomorrow.

Aha, good luck.

[hgm]

I expected the efficiency cores to be pretty slow, but fast enough to handle the GUI and OS background tasks.