New engine releases 2019

[Gabor Szots]

MORA 1.1.0 is now available.
https://github.com/gonzaloarro/MORA-CHESS-ENGINE

I added support to set the size of the transposition table with the UCI setoption command. I also fixed a bug related to the pawn hash table.

Fine. BTW, what does MORA stand for?

[Guenther]

MORA 1.1.0 is now available.
https://github.com/gonzaloarro/MORA-CHESS-ENGINE

I added support to set the size of the transposition table with the UCI setoption command. I also fixed a bug related to the pawn hash table.

Fine. BTW, what does MORA stand for? :wink:

I have a deja vu...

[gonzaloarro]

MORA 1.1.0 is now available.
https://github.com/gonzaloarro/MORA-CHESS-ENGINE

I added support to set the size of the transposition table with the UCI setoption command. I also fixed a bug related to the pawn hash table.

Fine. BTW, what does MORA stand for?

Haha that was a funny misunderstanding...

I didn't clarify this in the other post. Mora is the name of my cat, but I didn't actually choose it myself because I adopted her from a shelter and didn't want to change the name they had already gave her.
It is the spanish word for blackberry.

[Gabor Szots]

MORA 1.1.0 is now available.
https://github.com/gonzaloarro/MORA-CHESS-ENGINE

I added support to set the size of the transposition table with the UCI setoption command. I also fixed a bug related to the pawn hash table.

Fine. BTW, what does MORA stand for?

Haha that was a funny misunderstanding...

I didn't clarify this in the other post. Mora is the name of my cat, but I didn't actually choose it myself because I adopted her from a shelter and didn't want to change the name they had already gave her.
It is the spanish word for blackberry.

So in English it would be named BLACKBERRY.

[voffka]

Igel 2.1.0 at https://github.com/vshcherbyna/igel/releases/tag/2.1.0 (binaries for Raspberry Pi, Linux, macOS and Windows). This is a feature release that implements ponder support as well as level uci option to allow human players play (and win) against Igel. Despite this, the evaluation parameters have been re-tuned using lichess-quite set (thanks to Fabien, Vivien and to all people who made it available) and this added ~ 40 elo strength.

Strength reduction for levels:

- 1 to 10 is a level suitable for amateurs
- 11 to 19 is a level suitable for professional players
- level 20 is suitable for chess engines

What's new:

- Implement ponder option
- Implement level uci option: where 1 to 10 is a level suitable for amateurs, level 11 to 19 suitable for professional players and level 20 is suitable for chess engines
- Re-tuning evaluation using licess-quite set and texel tunning
- Increase threads limit to 256 to allow participation in TCEC

Relative strength against Igel 2.0.0

Code: Select all

tc=all/10+0.1
Score of Igel 2.1.0 64 POPCNT vs Igel 2.0.0 64 POPCNT: 1043 - 734 - 1088  [0.554] 2865
Elo difference: 37.62 +/- 10.03

[voffka]

Igel 2.1.1 at https://github.com/vshcherbyna/igel/releases/tag/2.1.1 (binaries for Raspberry Pi, Linux, macOS and Windows). This is a bugfix release that fixes issues discovered in TCEC when running Igel on 256 cpus. Most of the fixes only benefit in extreme hardware configuration with many cpus, so most likely it will not affect a playing strength in CCRL tests.

What's new:

- Use reliable api to determine passed time in unix OS
- Avoid cache coherency problems when calculating nodes and tbhits
- Clear hash during resize operation

[fabianVDW]

I have a new version of Fabchess at: https://github.com/fabianvdW/FabChess/r ... tag/v.1.13

This version increases the playing strength of Fabchess, preparing its appearance in TCEC.

Changes done:
- Decrease LMR
- Restructure the Transposition Table. New replacement schema and introduce buckets. Fab should be stronger in low cache size scenarios now.
- Increase the king zone at the edges of the board (A- and H-File)
- Changed the way pruning is done. In endgames, forward pruning besides NMP is not done anymore. The NMP conditions were simplified and allow pruning when in check
- Add new evaluation features: Namely Bishops, Rooks, and Queens XRAYing the enemy king and rooks on semi-open and queens on semi-and open file.
- SEE pruning of captures moves at low depths (quiet moves SEE pruning still missing)
And miscellaneous:
- Fix the zobrist keys
- Started using openbench
- The evaluation wiki was updated accordingly

Single threaded regression tests at 60+0.6s TC:

4-Cpu regression tests at 30+0.3sTC:

Note that here, RuyDos 1.1.9 was only playing with 1 Thread (because it doesn't support SMP), in that sense it provides an anchor.

Both tests obviously have SSS, but I would say strength has increased maybe 40-60 Elo.

[voffka]

Igel 2.1.1 at https://github.com/vshcherbyna/igel/releases/tag/2.1.1 (binaries for Raspberry Pi, Linux, macOS and Windows). This is a bugfix release that fixes issues discovered in TCEC when running Igel on 256 cpus. Most of the fixes only benefit in extreme hardware configuration with many cpus, so most likely it will not affect a playing strength in CCRL tests.

What's new:

- Use reliable api to determine passed time in unix OS
- Avoid cache coherency problems when calculating nodes and tbhits
- Clear hash during resize operation

Further testing of Igel 2.1.1 has indicated that the performance tweaks did not brought much benefit and affected performance especially in Linux OS. I will remove Igel 2.1.1 from github and advise testers to use Igel 2.1.0 (https://github.com/vshcherbyna/igel/releases/tag/2.1.0) in tournaments instead.

[Guenther]

not reported yet:

- Scorpio 3.05
- TomThumb 3.0
- RapChess 20191107
- Demolito (currently daily dev versions via appveyor)

Other info:

- PyTuroChamp github and site is gone!
(wayback only has a very old github snapshot saved)
- Natwarlal download site now only via wayback available

- compilation of Sapeli 1.64 added for older hardware (my compile)

[Edsel Apostol]

So the last 2 months I've been caught up with the chess programming virus again. Nothing too serious as I don't have much of a test environment (just a quad core laptop). It came to a point where I couldn't do much improvement anymore without doing intensive testing. So here goes the latest version.

Invictus revision 305 now available:

https://github.com/ed-apostol/InvictusC ... s/tag/r305

Last CCRL 40/4 rating is:
Invictus r228 64-bit 4CPU 2376 (+120 vs 1 CPU)!
Invictus r228 64-bit 2256

Latest version should be much stronger!