Upgraded Graphics Card, but what's the difference?

Hello,

Our company just paid $2,000 for a Workstation from Dell, with a Quadro
FX 3400 PCIe card, 3.2 P4, and 1gig Ram in the hopes it would speed up
my work. Well, Right now I am using a Precision Laptop M60 with a 1.5
P(M), 512 Ram, and a Quadro FX Go700. I am doing routing work and I
have one route with lots of pipe segments and fittings that takes
forever to edit (changing a dimension or adding a fitting takes about
15 seconds of "hang time"). I only have one large route because as
soon as I figured out that it's much faster to have multiple routes
than one big one, I made smaller routes. But besides that, I thought
getting a faster computer with a better graphics card would speed up my
editing times on this one large route, well, both graphics cards or
computers perform at the exact same speed when editing the large route.
I set them side by side & compared them. I was surprised, but maybe I
am not understanding what would speed up editing and changing this
large route. Is it processor speed that would increase the speed at
which SolidWorks rebuilds and makes changes to a route? I am sure this
would apply to all SolidWorks assemblies with lots of parts- rebuild
times, changing a dim and waiting for it to update. Thanks for any
advice.

 

Processor speeds between Pentium 4 and Pentium M (Centrino) cannot be
compared as directly equivalent. Here's a benchmark comparing a 1.7
GHz Centrino configuration to a 2.6 GHz Pentium 4.
http://www.intel.com/performance/mobile/notebook_configc.htm
It kind of looks like the Centrino wins. The effective processor
speeds may not actually be as different as they look between your M60
and the new Dell.

Brenda

 

How fast is your disk? Does it thrash?
Try a faster one?

 

Joe,

The video card won't help selecting, editing, or rebuilding. That's the
realm of the CPU. As was already mentioned, The Pentium M, and the P4 aren't
a 1:1 deal with regards to clock speed. The M is either more efficient, or
faster than advertized, or both. I'd read about this some months ago,
although I had no idea the discrepency was 2X.

If you really want to boost your desktop performance, you need to get away
from Intel. The AMD Athlon64 FX-53 is currently at the top of the
performance heap. At 2.8ghz, it's performance would be equal to a 5.04ghz
P4, (which deosn't exist).

Of course, if your stuck with Dell, your stuck with Intel.

Regards

Mark

 

There were quite a string of benchmark comparisons recently on the SW
Performance forum. Guess what, Intel is just ho-hum when it comes to
performance and the setup you bought would be considered just adequate.


When buying hardware and making these decisions it helps to have some
kind of benchmark. Since you do routing, take your assembly to the Dell
store, temporarily install SW and try it on several of their systems
till you find one that is sufficient for your needs.

After you buy a system use benchmarks like SpecAPC and Ship in a Bottle
to determine that what you bought is performing correctly. Yours might
not be the first time that a Dell came in through receiving with
something wrong with it. One user I know got three supposedly identical
units from Dell with widely varying benchmarks and real world
performance.

Benchmarks are extremely helpful in wringing out performance
bottlenecks, bad graphics drivers and flakey hardware. Make sure your
system is on par with other similar systems. Microsoft has some
procedures for finding performance bottlenecks and there are a bunch of
tips on the SW website known as best practices and such.

 

It is interesting that you are having speed problems with routing, I have
been using it since it was first released and it has always been a pain -
but I thought it was only me!! I have only noticed the usual incremental
performance increases as I have upgraaded my computers over the years -
there was no-one else here in Western Australia using piping (routing) that
I could compare notes with.

To help speed it up I try to add as much of the detail to a route
(instruments fittings, etc.) by opening the route assembly in it's own
window (i.e. not in the main assembly).

I also try to work with sub-assemblies as much as possible - e.g. in a large
steam installation I will add the piping to the individual equipment, then
bring that piece of equipment, along with its piping systems, into the main
assembly. This way I only have to do the main steam and condensate piping
routes in the main assembly.

Merry