| SPECIII? [message #34372] |
Wed, 12 March 2003 11:51  |
dj_connor
Messages: 1
Registered: March 2003
|
Junior Member |
|
|
I am in the market to buy a new computer to be used in part for ct 3D
reconstruction and was wondering if there has been any progress made
towards IDLSPECIII. Additionally, if there is anyone who is currently
doing 3D reconstruction, could you advise as to which platform would
be best?
Thanks,
DJ Connor
NCSU
|
|
|
|
| Re: SPECIII? [message #34466 is a reply to message #34372] |
Wed, 12 March 2003 15:42  |
JD Smith
Messages: 850
Registered: December 1999
|
Senior Member |
|
|
On Wed, 12 Mar 2003 12:51:47 -0700, DJ Connor wrote:
> I am in the market to buy a new computer to be used in part for ct 3D
> reconstruction and was wondering if there has been any progress made
> towards IDLSPECIII. Additionally, if there is anyone who is currently
> doing 3D reconstruction, could you advise as to which platform would be
> best?
Sadly not. Despite the best intentions, the hurdle of creating a new
benchmark from scratch was too much. IDL's default benchmark,
time_test, is rather long in the tooth, and it shows. For instance,
you can't get a meaningful I/O benchmark, since it writes at most 1
MB, which fits in all hard drive caches these days.
There has been some renewed interest in this lately, so it probably
should be done. Unfortunately, IDLWAVE takes up enough of my time as
it is ;). If anyone is interested in reviving IDLSpec, I'm happy to
help them on their way. I did it as an ugly bunch of shell scripts,
etc., but probably a bit of Perl would do the trick much more cleanly
(and you could CGI it, no less). I even collected a list of names
who'd be interested in helping design a new benchmark, potentially
with OpenGL performance included.
As for 3D modeling, the real question is which graphics card and
driver works best. I'm sure there are a few opinions on that here.
JD
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
