[Beowulf] Re: Beowulf Digest, Vol 33, Issue 23

[Håkon Bugge]

At 21:00 22.11.2006, beowulf-request at beowulf.org wrote:
> > A common confusion ... x86_64 changes nothing 
> about the precision of floats
> > or doubles in
> > C or Fortran.
>
>well, sort of.  it was pretty common to find at least some computations
>in ia32 using 80b FP, intentionally or not.  but iirc in long mode
>(colloquially x86_64), you no longer get x87 access.

I agree with Robert. Its not the architecture 
which changed the precision, its the change of 
compilers to use the SSE instruction set. And the 
SSE instruction was available on ia32 before x86_64 .

You can still perform FP arithmetic using native 
precision on ia32, given it has SSE, or you use 
doubles and set the FP-precision to 64-bit. 
Similar, you can get the 80-bit arithmetic on 
x86_64, by disabling the compiler from using the SSE instruction set.

Håkon



--
Håkon Bugge
CTO
dir. +47 22 62 89 72
mob. +47 92 48 45 14
fax. +47 22 62 89 51
Hakon.Bugge at scali.com
Skype: hakon_bugge

Scali - http://www.scali.com
Scaling the Linux Datacenter