Ok, but the version without SSE is a little bit slower then with SSE, because the SSE version is using the faster PopCount function.
however if it stops again or you found something other bug, let me know.
Tornado 6.0 released
Moderator: Ras
-
Engin
- Posts: 1001
- Joined: Mon Jan 05, 2009 7:40 pm
- Location: Germany
- Full name: Engin Üstün
-
Damir
- Posts: 2889
- Joined: Mon Feb 11, 2008 3:53 pm
- Location: Denmark
- Full name: Damir Desevac
Re: Tornado 6.0 released
You were right Engin. My hardware does not support SSE. The other link works fine even though it is a little bit slower. There is no more crashes.
-
Dr.Wael Deeb
- Posts: 9773
- Joined: Wed Mar 08, 2006 8:44 pm
- Location: Amman,Jordan
Re: Tornado 6.0 released
Thanks Engin
What is the estimated Elo gain over the 4.88 version
Dr.D
What is the estimated Elo gain over the 4.88 version
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….
-
Engin
- Posts: 1001
- Joined: Mon Jan 05, 2009 7:40 pm
- Location: Germany
- Full name: Engin Üstün
Re: Tornado 6.0 released
well, that's difficult to say, because its depends on wich version, wich time control, how much cores are using, how much games playing, pondering or not, and of course on wich hardware you running it.
a quick test match Tornado 6.0 x64 SSE vs Tornado 4.88 x64 SSE on my own hardware (AMD Phenom II x6 1090) with shredder classic GUI, 3+2 sec. time control with ponder, i noticed after 100 games a improvement around +80 elo.
on different time controls, games and hardwares that can be change. i don't have the time to testing intensive.
a quick test match Tornado 6.0 x64 SSE vs Tornado 4.88 x64 SSE on my own hardware (AMD Phenom II x6 1090) with shredder classic GUI, 3+2 sec. time control with ponder, i noticed after 100 games a improvement around +80 elo.
on different time controls, games and hardwares that can be change. i don't have the time to testing intensive.
-
Dr.Wael Deeb
- Posts: 9773
- Joined: Wed Mar 08, 2006 8:44 pm
- Location: Amman,Jordan
Re: Tornado 6.0 released
Wow,even at this time control that's an amazing improvementEngin wrote:well, that's difficult to say, because its depends on wich version, wich time control, how much cores are using, how much games playing, pondering or not, and of course on wich hardware you running it.
a quick test match Tornado 6.0 x64 SSE vs Tornado 4.88 x64 SSE on my own hardware (AMD Phenom II x6 1090) with shredder classic GUI, 3+2 sec. time control with ponder, i noticed after 100 games a improvement around +80 elo.
on different time controls, games and hardwares that can be change. i don't have the time to testing intensive.
Thanks again Engin and good luck with further releases....
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….
-
lucasart
- Posts: 3242
- Joined: Mon May 31, 2010 1:29 pm
- Full name: lucasart
Re: Tornado 6.0 released
Yes, you are right. More precisely, SSE instruction sets and the POPCNT hardware instruction are separate things. The only thing we can say in general is that SSE4.2 implies POPCNT. But the reverse is not true: there are some AMD machines with SSE3 only but with POPCNT support.SzG wrote: I think the distinction is not between SSE and non-SSE but rather between SSE and SSE4.2 (or something like that).
That is why, SF "modern" compiles use SSE3+POPCNT, but not SSE4.2
Theory and practice sometimes clash. And when that happens, theory loses. Every single time.
-
Modern Times
- Posts: 3758
- Joined: Thu Jun 07, 2012 11:02 pm
Re: Tornado 6.0 released
Indeed, the Phenom II X6 is one examplelucasart wrote:there are some AMD machines with SSE3 only but with POPCNT support.
Given that I have one of the above CPUs, thanks to the SF team for understanding this situation. Some authors just provide SSE4,2 as a high-end compile, and of course they don't work on the X6.lucasart wrote: That is why, SF "modern" compiles use SSE3+POPCNT, but not SSE4.2
-
Sedat Canbaz
- Posts: 3018
- Joined: Thu Mar 09, 2006 11:58 am
- Location: Antalya/Turkey
Re: Tornado 6.0 released
Great news!Engin wrote:Hello,
Tornado 6.0 version is available for download on my website
URL: http://tornadochess.bplaced.net/
Thank you dear Engin!
Best,
Sedat
-
Wolfgang
- Posts: 989
- Joined: Sat May 13, 2006 1:08 am
Re: Tornado 6.0 released
?? ok, one match only vs. one opponent only and a "family"-match...Engin wrote:.... i noticed after 100 games a improvement around +80 elo.
...
http://cegt.forumieren.com/t238-testing-tornado-60
-44 to 5.0beta after 600 games and still -14 to 4.88
5 matches played on:
- AMD-FX 6300 @ 3,5GHZ (5 cores used, 1 core free for OS etc.)
- Classic GUI (started 5 times on the X6)
- 40 moves/3 minutes repeated
1 match played on:
- Intel Core i5-540 (notebook) @ 2,53GHZ
- Arena 3.0 GUI
- 40 moves/3 minutes repeated
-
carldaman
- Posts: 2287
- Joined: Sat Jun 02, 2012 2:13 am
Re: Tornado 6.0 released
Thanks ++, Engin. A complete rewrite sounds like a lot of work...Engin wrote:Hello,
Tornado 6.0 version is available for download on my website
URL: http://tornadochess.bplaced.net/
Much appreciated,
CL