Jump to content

Better platform for rendering?


Recommended Posts

Hello.

I work in a small architectural firm that uses a mix of Mac and Windows systems. Recently, a number of clients have asked whether we can provide animations (walk-throughs, flyovers, etc.), so we are planning to purchase a computer system that will be dedicated exclusively to producing high quality renderings and video animations (walk-throughs, flyovers, solar animations, etc.)

We have narrowed our choices of hardware to a 2 Ghz PowerMac G5 Dual or a Sony VAIO PCV-RZ54G Pentium 4 3.2GHz. Both computers come with the ATI Radeon 9600 XT graphics card with 128 Mb VRAM, and will be upgraded to 2 Gb of system RAM.

I am wondering which system will be better for its intended purpose. I recognize that Macs have a reputation for being better for graphic-intensive tasks than PCs, however, I'm wondering whether both lab testing and real world experience justify that reputation. Can anyone point me to test results comparing performance on the two platforms or provide anecdotal feedback?

TIA.

Link to comment

i'm not sure why this info is not posted by NNA on the vectorworks site. there was an interview with one of lead programmers of Vectorworks in which there was discussion of speed on various platforms. although is almost a year old. it does have a serise of in Vectorworks Benchmarks, for Mac's including the Dual G5 and x86 both Intel and AMD. You may best guess scale the numbers to compair your two options

www.architosh.com

Link to comment

Both of these computers would be very nice to work with. There is no inherent advantage to using a mac with vectorworks and for RenderWorks tasks a 3.2 GHz PC will defininitely be faster than a 2 GHz mac.

One other thing worth noting is that RW11 now supports dual processors and the P4 w/HT enabled will be recognized as 2 processors and will give a speed increase of about 20% over a non HT processor.

[ 08-04-2004, 09:40 AM: Message edited by: AndyM ]

Link to comment

> Both of these computers would be very nice to work with. There is no inherent advantage to using a mac with vectorworks and for RenderWorks tasks a 3.2 GHz PC will defininitely be faster than a 2 GHz mac.

Andy, are you comparing apples to apples? The Sony has a single processor, while the Mac has dual processors. Are you saying that a 3.2 GHz single processor system will render faster than a 2 GHz dual processor system?

Link to comment

The problems on the pc don't radicate only on the hardware, the problem it's the software. The stability of the os. I used both platforms with very similar specifications and the memory management it's better on panther. Now i experienced a dual 2.0 g5 and the velocity on renders (renderworks 11 and cinema 4d) it's realy fast, more than any pc i ever seen.

Link to comment

I didn't catch that the G5 was a duallie, but since the P4 would be at least a $1000 cheaper that's not really apples to apples is it? Take a look at this site:

http://www.3dfluff.com/mash/cbdual.php

It lists benchmarks for Cinebench which uses another multi-threaded raytrace renderer and is the most relevent benchmark I am aware of. The dual G5 2.0 scores about 520 if the app were optimised for the G5 which neither C4D or VW/RW are. The single p4 3.2 scores about 385. But if you compare a comperably priced dual xeon the Intel machines win every time. OS preference aside, the P4 w/HT is the best bang for the buck with multi-threaded raytrace rendering.

I also did some speed tests way back when:

http://techboard.nemetschek.net/cgi/ubb/ultimatebb.cgi?ubb=get_topic;f=17;t=000219

The gist of it was that VW/RW is not optimised for the G4 or G5 and processor clock speed rules.

Link to comment

> I also did some speed tests way back when:

Andy, not trying to be argumentative, but July 2002 was a long time ago. From the results posted in the link by iboymatt above, it appears that NNA's testing found that dual processor G5s significantly outperform dual processor Xeons. Do you have results from real world tests on current generation equipment that contradict their findings?

Also, it appears that the G5 dual results posted on the 3dfluff website are for beta software, so it's hard to judge how realistic the numbers are. Can you point me to performance comparisons using full release versions of the real, working (non-benchmarking) software? (Having written benchmarking software in the past, I tend to doubt the value of benchmarking as an indicator of real world results, not only because it is relatively trivial to tweak a testing setup to produce the outcome one wants, but also because something as simple as changing the compiler used to compile the benchmark program can skew the results by several orders of magnitude.)

Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...