Release: Nemo 1.0 Beta

Discussion of anything and everything relating to chess playing software and machines.

Moderator: Ras

noctiferus
Posts: 364
Joined: Sun Oct 04, 2009 1:27 pm
Location: Italy

Re: Release: Nemo 1.0 Beta

Post by noctiferus »

Forget my previous post: the problem is due to Komodo3, that has the same behaviour against other engines too.
User avatar
geots
Posts: 4790
Joined: Sat Mar 11, 2006 12:42 am

Re: Release: Nemo 1.0 Beta

Post by geots »

noctiferus wrote:Forget my previous post: the problem is due to Komodo3, that has the same behaviour against other engines too.

Look, it shows up the same as he described with ALL ENGINES. So all engines can't be the problem with Nemo ok. Maybe not all, but the 18 engines I tried Nemo against- same as he mentioned.


george
tmokonen
Posts: 1355
Joined: Sun Mar 12, 2006 6:46 pm
Location: Kelowna
Full name: Tony Mokonen

Re: Release: Nemo 1.0 Beta

Post by tmokonen »

When I pit Nemo against another engine on my dual core laptop (32 or 64 bit version), it always takes up 50% CPU (ie all of one of the CPUs), and I never use ponder. Even after the game is finished and both engines are idle, Nemo still takes 50% of the CPU. It's not a Chessbase specific problem; I use Arena 3.0. I think what I'm experiencing is similar to what's happening to George.
User avatar
Dr.Wael Deeb
Posts: 9773
Joined: Wed Mar 08, 2006 8:44 pm
Location: Amman,Jordan

Re: Release: Nemo 1.0 Beta

Post by Dr.Wael Deeb »

tmokonen wrote:When I pit Nemo against another engine on my dual core laptop (32 or 64 bit version), it always takes up 50% CPU (ie all of one of the CPUs), and I never use ponder. Even after the game is finished and both engines are idle, Nemo still takes 50% of the CPU. It's not a Chessbase specific problem; I use Arena 3.0. I think what I'm experiencing is similar to what's happening to George.
This reminds me of The King chess engine although the author stated that it actually doesn't use the allocated CPU when idle.....
Dr.D
_No one can hit as hard as life.But it ain’t about how hard you can hit.It’s about how hard you can get hit and keep moving forward.How much you can take and keep moving forward….
perejaslav
Posts: 240
Joined: Sat Mar 18, 2006 4:01 am
Location: Cold

Re: Release: Nemo 1.0 Beta

Post by perejaslav »

tmokonen wrote:When I pit Nemo against another engine on my dual core laptop (32 or 64 bit version), it always takes up 50% CPU (ie all of one of the CPUs), and I never use ponder. Even after the game is finished and both engines are idle, Nemo still takes 50% of the CPU. It's not a Chessbase specific problem; I use Arena 3.0. I think what I'm experiencing is similar to what's happening to George.
Yes, when I tried to test Nemo in LB 2.72 and run 3 instances of each engine, my CPU was loaded at 90-100% instead of 75%.
noctiferus
Posts: 364
Joined: Sun Oct 04, 2009 1:27 pm
Location: Italy

Re: Release: Nemo 1.0 Beta

Post by noctiferus »

Another little piece of info:
I am running Swami's STS on nemo: while all other engines release the CPU between one epd analysis and the next one, Nemo holds the CPU.
noctiferus
Posts: 364
Joined: Sun Oct 04, 2009 1:27 pm
Location: Italy

Re: Release: Nemo 1.0 Beta

Post by noctiferus »

Finished my Sts test.
Very good for a starting engine: the result is a bit better than Greko 8.0 32 bit (Swami's result, roughly equivalent to my settings).
I'll run nemo32 for precise comparison...
User avatar
Sylwy
Posts: 4812
Joined: Fri Apr 21, 2006 4:19 pm
Location: IAȘI - the historical capital of MOLDOVA
Full name: Silvian Rucsandescu

Re: An important post for Maestro Hoffmann !!!!!!!

Post by Sylwy »

Hi Maestro Hoffmann !

My final conclusions about Nemo 1.0 beta chess engine:

1.it's a promising engine;
2.it has a big problem:

something in the communication protocol between GUI (Fritzy 11) - engine (rectae UCI protocol) is broken. Loading/unloading Nemo engine is a very - very slowly process ( something like an idle process ). This behaviour generates the process shows in the below images.
Nemo uses a % from the CPU usage of Loop 10.32 f engine - causing to this engine a low kN/s and depth. So , Nemo bug affects the final result.



1.-Nemo thinking (99% CPU usage):

Image

2.-Loop 10.32f thinking (Loop-48% CPU usage , Nemo [on the 3-rd rank of Task Manager] -52% CPU usage:

Image

3.Waiting to load Nemo ( a long process - this is in fact the main problem: loading/unloading the Nemo engines generates someting like an idle process):

Image

4.-Near the final result versus Loop 10.32f (irrelevant because of the above problem):

Image

Well Maestro Hoffmann , the seniors of CCRL team (Maestro George and Maestro Werner) were right !

Best to you & Nemo,
SilvianR :wink:
User avatar
Desperado
Posts: 879
Joined: Mon Dec 15, 2008 11:45 am

Re: An important post for Maestro Hoffmann !!!!!!!

Post by Desperado »

Hi Ruxy :) ,

1:
The "cpu-load" issue has been fully accepted, and is already fixed.
From the buffix Version on, there shouldnt be a problem any longer
with that issue.

2:
We will have now a closer look on the "loading" issue.

3:
Because George used the term "memory", we additionally are checking
Memory-Allocation (context Hash) problems. But, for now this second
point looks like a misunderstanding because of using the term memory,
that has nothing to do with the "cpu-load". But anyway, we are checking
that out, and so far we will have some tiny changes
for the memory-allocation-code (Transtable,Hash), just to be on the
safe (safer) side.

Thx for your report.

Michael
User avatar
Sylwy
Posts: 4812
Joined: Fri Apr 21, 2006 4:19 pm
Location: IAȘI - the historical capital of MOLDOVA
Full name: Silvian Rucsandescu

Re: An important post for Maestro Hoffmann !!!!!!!

Post by Sylwy »

Maestro Hoffmann ,

I love this engine. Just I want to see the perfection ( that's because Ruxy is an imperfect being - a psychological matter :lol: ).
I can't wait for the new release !

Have a nice day !
SilvianR :wink: