Engine release: weiss

[Terje]

I allocate 256 MB hash for the engine but by the end of the game its memory allocation reaches almost 2 GB. Probably because of syzygy bases but I am afraid my 8 GB RAM won't like it.

Is that playing a lot of moves after reaching a tablebase position or something? Locally I test with 1024MB hash, and have only ever seen it sit around 1053MB total in task manager.

Yes, although I do not know if it was a TB position because there were still many pieces (I think 10) on the board.

That's really weird - weiss definitely does not do this on my machine. What OS are you on?

[Gabor Szots]

I allocate 256 MB hash for the engine but by the end of the game its memory allocation reaches almost 2 GB. Probably because of syzygy bases but I am afraid my 8 GB RAM won't like it.

Is that playing a lot of moves after reaching a tablebase position or something? Locally I test with 1024MB hash, and have only ever seen it sit around 1053MB total in task manager.

Yes, although I do not know if it was a TB position because there were still many pieces (I think 10) on the board.

That's really weird - weiss definitely does not do this on my machine. What OS are you on?

64-bit W10.

[Terje]

I allocate 256 MB hash for the engine but by the end of the game its memory allocation reaches almost 2 GB. Probably because of syzygy bases but I am afraid my 8 GB RAM won't like it.

Is that playing a lot of moves after reaching a tablebase position or something? Locally I test with 1024MB hash, and have only ever seen it sit around 1053MB total in task manager.

Yes, although I do not know if it was a TB position because there were still many pieces (I think 10) on the board.

That's really weird - weiss definitely does not do this on my machine. What OS are you on?

64-bit W10.

Hmm, what time control?

Tried a game without pext, 256MB hash, 2 seconds per move, even after reaching a position with 4 pieces and getting 200k tbhits the memory usage was barely higher than at the start (264 -> 268).

[Gabor Szots]

Hmm, what time control?

Tried a game without pext, 256MB hash, 2 seconds per move, even after reaching a position with 4 pieces and getting 200k tbhits the memory usage was barely higher than at the start (264 -> 268).

40 moves in 2 minutes, repeating.

[Terje]

Hmm, what time control?

Tried a game without pext, 256MB hash, 2 seconds per move, even after reaching a position with 4 pieces and getting 200k tbhits the memory usage was barely higher than at the start (264 -> 268).

40 moves in 2 minutes, repeating.

Even with the compile you provided I can't seem to replicate the issue. At most it rose to 270MB here, while getting 500k+ tbhits for many moves. Could it matter what the tablebases are stored on? Could you try running again without tablebases to see if it happens without them?

[Gabor Szots]

Hmm, what time control?

Tried a game without pext, 256MB hash, 2 seconds per move, even after reaching a position with 4 pieces and getting 200k tbhits the memory usage was barely higher than at the start (264 -> 268).

40 moves in 2 minutes, repeating.

Even with the compile you provided I can't seem to replicate the issue. At most it rose to 270MB here, while getting 500k+ tbhits for many moves. Could it matter what the tablebases are stored on? Could you try running again without tablebases to see if it happens without them?

I did not compile anything, must have been someone else.

I am in the middle of a tournament and can't change engine parameters permanently without corrupting the result. However, for a couple of games maybe it won't hurt so I am going to do it.
I am sure it's the syzygy, I have seen this with other engines too.

[Terje]

I did not compile anything, must have been someone else.

I am in the middle of a tournament and can't change engine parameters permanently without corrupting the result. However, for a couple of games maybe it won't hurt so I am going to do it.
I am sure it's the syzygy, I have seen this with other engines too.

Oh oops. Ok, I see. Thanks for helping!

[Gabor Szots]

I have just followed one 11-piece ending until the conclusion without syzygy in use. The memory allocation remained at 259 MB throughout. Q.E.D.

[Terje]

I have just followed one 11-piece ending until the conclusion without syzygy in use. The memory allocation remained at 259 MB throughout. Q.E.D.

Ok, thanks for confirming.