Engines with nps limitation and FRC support

[Alexander Schmidt]

Hi,

I am looking for chess engines with a combination of this features:

- limit nps, or ELO settings with a limit of nps
- FRC
- Linux/Windows compatible

So far I found Galjoen, Minic, Wasp. Any more ideas?

[xr_a_y]

Minic "level" implementation can
- limit the number of nodes searched, but not "nodes per second"
- select not the best move (using multipv)

[j.t.]

Hi,

I am looking for chess engines with a combination of this features:

- limit nps, or ELO settings with a limit of nps
- FRC
- Linux/Windows compatible

So far I found Galjoen, Minic, Wasp. Any more ideas?

Limited NPS can be simulated by allowing the engine to use only a fraction of the actual time.

[Uri Blass]

Hi,

I am looking for chess engines with a combination of this features:

- limit nps, or ELO settings with a limit of nps
- FRC
- Linux/Windows compatible

So far I found Galjoen, Minic, Wasp. Any more ideas?

Movei supports FRC and supports fixed number of nodes per move(by supporting the uci protocol).
I do not know if it is good enough for you because fixed number of nodes per move is not the same as fixed number of nodes per second and the level of fixed number of nodes per move is practically the same at all time controls.

[Alexander Schmidt]

Minic "level" implementation can
- limit the number of nodes searched, but not "nodes per second"
- select not the best move (using multipv)

Yes, I missunderstood this.

My goal is to have some engines to play against, that play the same moves on different plattforms, e.g. on my Linux Laptop, on my Windows tablet, and on future computers. So a limitation of the nps is the best for this purpose, limitation of nodes is OK as well as long as the engine don't play immediately. I don't like reducing strength with randomness or reducing knowledge. This leads to an ugly style.

[AdminX]

Minic "level" implementation can
- limit the number of nodes searched, but not "nodes per second"
- select not the best move (using multipv)

Yes, I missunderstood this.

My goal is to have some engines to play against, that play the same moves on different plattforms, e.g. on my Linux Laptop, on my Windows tablet, and on future computers. So a limitation of the nps is the best for this purpose, limitation of nodes is OK as well as long as the engine don't play immediately. I don't like reducing strength with randomness or reducing knowledge. This leads to an ugly style.

Have you tried Move Delay adapter by Odd Gunnar Malin

https://github.com/OGMalin/Binaries

[Alexander Schmidt]

Didn't know this, many thanks for the hint!