Strange output from Zappa X64

[M ANSARI]

In this position Zappa seems to get a strange output of +222.06 for a long long time and eventually after a few minutes will show -1.63 ... anyone else have this behaviour?

[d]r1b1kb1r/1p3p2/p3pP1p/1Bn5/3NN3/q6p/P1PQ2PP/1R3RK1 b kq - 0 17

[PauloSoare]

Could not reproduce the problem.

Paulo Soares

[Chris Taylor]

[quote="M ANSARI"]In this position Zappa seems to get a strange output of +222.06 for a long long time and eventually after a few minutes will show -1.63 ... anyone else have this behaviour?

[d]r1b1kb1r/1p3p2/p3pP1p/1Bn5/3NN3/q6p/P1PQ2PP/1R3RK1 b kq - 0 17[/quote]

Hi,

What gui and os are you using?

Chris

[M ANSARI]

Using DF10 GUI with Winxp 64bit. Hiarcs, Rybka and DF10 have no problems. It takes Zappa Zan 64 about 3 minutes to get out of the +222.06 lock and give a -1.63 output.

[Chris Taylor]

[quote="M ANSARI"]Using DF10 GUI with Winxp 64bit. Hiarcs, Rybka and DF10 have no problems. It takes Zappa Zan 64 about 3 minutes to get out of the +222.06 lock and give a -1.63 output.[/quote]


Hi

Can't get the fault with Fritz 10, suspect it might be a fault in the deep version?!
I have tried on 32 bit and 64 bit Windows..


Chris

[Anabolic Karpov]

No, in XP 64 bit in GUI10 service pack 6 evaluation momentarily goes to +2.64 with 1 .. Nd7 at 10 ply and then +2.63 1 .. axb5 at 10 ply before changing to -1.16 1..axb5 at 10 ply all in first 5 secs.

Only time I've seen these odd high scores has been when Zappa's move has been forced so evaluation isn't completed properly. e.g. a rook check move when king must take rook. May be an issue peculiar to mullti-core in these types of positions but could not reproduce here on my dual AMD 64 4800
AK