--Signature_Fri__30_Sep_2005_11_47_16_-0400_DlQN1d64pCdLcCnu
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: quoted-printable

Hello,

> > So should x86 be faster (almost twice as fast) in this instance?
> > Does python suck on SPARC64 still due to threading issues?
>=20
> Of course a 600 MHz processor should outperform a 400 MHz one ;-)

Well, I have a 195MHz one that outperforms an 800MHz one :p
( MIPS R10K vs. G4 in a 64bit floatpoint-heavy number crunching )

> But what you are realy testing here, is the adaption of the python
> interpreter to your host system. I have never looked at python, but I
> suppose there are no sparc64 specifics in it, while there actually
> may be i386 assembler stuff or what have you. Maybe python plays
> better with little endian machines in general?

I'm not sure Python uses any 64bit arithmetics at all. The SPARC could
have an edge when you're manipulating Really Big(tm) data structures
though.

> Any 64 bit advantages (if your application might have had) will
> probably be lost when using an interpreted language anyway.

And the bigger address space doesn't help at all in a box that maxes out
at 1GB RAM.

have fun
Michael

--Signature_Fri__30_Sep_2005_11_47_16_-0400_DlQN1d64pCdLcCnu
Content-Type: application/pgp-signature

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (NetBSD)

iQEVAwUBQz1eBcpnzkX8Yg2nAQL30wf/ZmaFfPcwYP4j+P8z9WsXJ1tL1UFwoWgs
z4jA6OaCrMwyINDfkdhoMUvvBIKyWUllp7wJox2Lv/iGTk6QSMyXP/atZVNSfC46
1F/CLnaO1tVtpHRef6BNE81EY/KkwkfRx05Z70XfgzG8UM1tLvmg1cIdk6DqU3ua
3LPCjbBSkPBgeIGfgKl8TGBJ8uFt2d2YYoh4YJMDsWXqFvgA0gLgXOJvmRq16D47
VqfhF6u+5ixMq0Uy73pyu//JTWPUiJ3u8iikUwsWInEmICkkWFk52y+hSWK79Mf2
MdtLb9sBlt247tGpLyYe9Vs8t2oUutHgAigKKk5fKo0ZOgM1QpD+LA==
=A8hx
-----END PGP SIGNATURE-----

--Signature_Fri__30_Sep_2005_11_47_16_-0400_DlQN1d64pCdLcCnu--