Maybe because I like using Arena GUI for Testing ?Nordlandia wrote:Anil V Dharan: tell me the reason for not using Fritz 15.
Komodo 11.2
Moderators: hgm, Rebel, chrisw
-
- Posts: 1339
- Joined: Fri Nov 02, 2012 9:43 am
- Location: New Delhi, India
Re: Dynamism in Komodo 11.2
i7 5960X @ 4.1 Ghz, 64 GB G.Skill RipJaws RAM, Twin Asus ROG Strix OC 11 GB Geforce 2080 Tis
-
- Posts: 2821
- Joined: Fri Sep 25, 2015 9:38 pm
- Location: Sortland, Norway
Re: Komodo 11.2
Graham: do you intentionally avoid using 6-men EGTB, or is it because your current SSD is not fit enough?Graham Banks wrote:http://talkchess.com/forum/viewtopic.php?t=64666lkaufman wrote:.......I'd bet a lot of money on any match between 11.2 and Houdini 5 at any reasonable time control.
Upgrading to 6-men syzygy should be prioritized.
-
- Posts: 41575
- Joined: Sun Feb 26, 2006 10:52 am
- Location: Auckland, NZ
Re: Komodo 11.2
Space.Nordlandia wrote:Graham: do you intentionally avoid using 6-men EGTB, or is it because your current SSD is not fit enough?Graham Banks wrote:http://talkchess.com/forum/viewtopic.php?t=64666lkaufman wrote:.......I'd bet a lot of money on any match between 11.2 and Houdini 5 at any reasonable time control.
Upgrading to 6-men syzygy should be prioritized.
gbanksnz at gmail.com
-
- Posts: 2821
- Joined: Fri Sep 25, 2015 9:38 pm
- Location: Sortland, Norway
Re: Komodo 11.2
Do you plan purchase another SSD soon?
Samsung 960 EVO 256 GB edition is affordable and good enough for most engine testers.
As long your motherboard support M.2 SSDs
Samsung 960 EVO 256 GB edition is affordable and good enough for most engine testers.
As long your motherboard support M.2 SSDs
-
- Posts: 41575
- Joined: Sun Feb 26, 2006 10:52 am
- Location: Auckland, NZ
Re: Komodo 11.2
Unfortunately, I'm on a tight budget, as we're a sole income family.Nordlandia wrote:Do you plan purchase another SSD soon?
Samsung 960 EVO 256 GB edition is affordable and good enough for most engine testers.
As long your motherboard support M.2 SSDs
No big plans at present.
Cheers,
Graham.
gbanksnz at gmail.com
-
- Posts: 3557
- Joined: Thu Jun 07, 2012 11:02 pm
Re: Komodo 11.2
Space is not the only issue. On long time control the O/S keeps on caching 6-men access until the RAM on your machine is full and it starts swapping, and then you can throw your testing out the window. Would not happen at blitz, it may not happen at 40/40 either but I don't know.
-
- Posts: 2821
- Joined: Fri Sep 25, 2015 9:38 pm
- Location: Sortland, Norway
Re: Komodo 11.2
So probing 6-men will eventually backfire unless sufficient enough RAM installed. Once main RAM get full, doesn't windows keep filling up storage and retrieves data from secondary storage for use in main memory?Modern Times wrote:Space is not the only issue. On long time control the O/S keeps on caching 6-men access until the RAM on your machine is full and it starts swapping, and then you can throw your testing out the window. Would not happen at blitz, it may not happen at 40/40 either but I don't know.
Swapping - Paging ->
When your computer runs out of memory, your computer will start to use the hard drive space for "Virtual Memory"
I got about 50 GB left on my M.2 SSD. Is it possible to make windows select that one once memory get filled up?
-
- Posts: 3557
- Joined: Thu Jun 07, 2012 11:02 pm
Re: Komodo 11.2
It isn't the same as normal paging or swapping. It is quite technical. There are some posts on here somewhere about it.
-
- Posts: 5961
- Joined: Sun Jan 10, 2010 6:15 am
- Location: Maryland USA
Re: Dynamism in Komodo 11.2
OK, I did some tests at 3' + 1" of Dynamism 130 (the value that is currently testing as best) vs 80 (your value). On one thread Dyn 130 won by 80 elo (!!) (that is not a typo!) after 754 games. On four threads Dyn 130 is currently ahead by 140 Elo (!!!) after 67 games. I agree that Dyn 80 seems to give more reasonable evals, but it's hard to argue with statistics like these. It seems this Dyn parameter is very important.shrapnel wrote:Granted. My bad.lkaufman wrote: You said "left at 80", which implies that the default is 80.
I usually play at 4'+2'' or 5'+1" on Infinity Chess or while Testing on Arena GUI. No time nowadays for my favorite LTC games.
Yes, my opinion is mostly based on results, but yes, sometimes I've observed that Komodo shows a over-optimistic evaluation which changes later.
Number of Threads is 8 because I use 8-Core Processors.
I've tried HT ON also, but using 12-14 Threads doesn't really seem to make a positive difference.
Komodo rules!
-
- Posts: 1339
- Joined: Fri Nov 02, 2012 9:43 am
- Location: New Delhi, India
Re: Dynamism in Komodo 11.2
Interesting. I'll give 120-130 Dynamism a shot again and see.lkaufman wrote:OK, I did some tests at 3' + 1" of Dynamism 130 (the value that is currently testing as best) vs 80 (your value). On one thread Dyn 130 won by 80 elo (!!) (that is not a typo!) after 754 games. On four threads Dyn 130 is currently ahead by 140 Elo (!!!) after 67 games. I agree that Dyn 80 seems to give more reasonable evals, but it's hard to argue with statistics like these. It seems this Dyn parameter is very important.
I was getting too many Draws with Dynamism 80, anyway.
i7 5960X @ 4.1 Ghz, 64 GB G.Skill RipJaws RAM, Twin Asus ROG Strix OC 11 GB Geforce 2080 Tis