[Rd] RE: 64-bit R on Opteron [was Re: [R] Windows R 1.8.0 hangs when M em Usage >1.8GB]

[Liaw, Andy]

> From: Douglas Bates
>
> How does the Opteron perform on floating point?  Can you try something
> like
> 
> > mm = matrix(rnorm(1e6), nc = 1e3)
> > system.time(crossprod(mm))
> [1] 0.51 0.02 0.53 0.00 0.00
> > system.time(crossprod(mm))
> [1] 0.37 0.03 0.40 0.00 0.00
> > system.time(crossprod(mm))
> [1] 0.38 0.02 0.40 0.00 0.00
> > system.time(crossprod(mm))
> [1] 0.38 0.02 0.40 0.00 0.00
> 
> (That was with R compiled to use Goto's BLAS on a 2.0 GHz P4.)
> 
> Are you using Goto's BLAS or Atlas?

I managed to link R against Goto's BLAS for Opteron (see below).  The result
looks like:

> mm = matrix(rnorm(1e6), nc = 1e3)
> system.time(crossprod(mm))
[1] 0.5 0.0 0.5 0.0 0.0
> system.time(crossprod(mm))
[1] 0.39 0.00 0.39 0.00 0.00
> system.time(crossprod(mm))
[1] 0.39 0.01 0.39 0.00 0.00

On the dual 2.4GHz Xeon (using libgoto_p3_512-r0.6.so), I get:

> mm = matrix(rnorm(1e6), nc=1e3)
> system.time(crossprod(mm))
[1] 0.91 0.01 0.93 0.00 0.00
> system.time(crossprod(mm))
[1] 0.82 0.02 0.84 0.00 0.00
> system.time(crossprod(mm))
[1] 0.83 0.02 0.85 0.00 0.00

We have not found the Opteron to be faster than the Xeon, but the Xeon does
have faster RAM...

[What I did to link R against Goto's BLAS (mostly following Prof. Bates'
instruction): 
- Install libgoto*.so and symlink to libgoto.so.
- Download xerbla.f and compile to xerbla.o.
- Run the R configure script with --with-blas="-lgoto /path/to/xerbla.o".
- Edit Makeconf and delete the path to xerbla.o in BLAS_LIB.
- make; make check
]

A bit of gripe about the SUSE ES8 that was pre-loaded on the Opteron:
- VNC segfault out-of-box.  Needed to purchase maintanence contract to get
update that fixed it.
- No teTeX, so can't build R manuals in pdf.
- No XEmacs (but does have GNU Emacs).

Best,
Andy