PDA

View Full Version : Boxx Benchies

reForm
10-28-2005, 04:37 PM
A shout out to the Boxx'ers out there...

Is there anywhere that I can see like-for-like benchmarks showing the difference between :-

Dual core X2
Dual processor Opteron
Dual Dual-core Opteron

Boxx workstations in 3d Max/Vray rendering.

Given previous threads about the ability of software to make the most of these dual-core systems, it would be interesting to see what the actual performance differences are between the different boxx systems.

It would be good to make some kind of unit to measure performance for these things... eg (£ / Ghz) / (rendertime of big scene) = Cost benefit ratio ....

Anyway... back to my main question... is there any benchmarks comparing boxx systems as above?

Cheers.
Ed Caracappa
10-28-2005, 05:34 PM
I've got my guys working on a similar exercise right now. You have a big scene you'd like to see rendered?

Ed
meanadam
10-30-2005, 05:40 PM
Doesn't include boxx.... but here is something I put together in July to determine the purchase of x2. This was reinforced by a benchmark scene in the Vray forums. Most of the heavy users were switching from xeon to x2 at the time, us included.

http://www.studio2a.net/a2a/benchcost.jpg

Donno how Ed will justify the costs of the Boxx. I would like to hear.
[nitin design]
10-30-2005, 06:14 PM
Doesn't include boxx.... but here is something I put together in July to determine the purchase of x2. This was reinforced by a benchmark scene in the Vray forums. Most of the heavy users were switching from xeon to x2 at the time, us included.

http://www.studio2a.net/a2a/benchcost.jpg

Donno how Ed will justify the costs of the Boxx. I would like to hear.

Thanks for the data...very interesting.
I think it's not impossible to justify the cost of the dual core opterons if time is very important for the designer/artist. On big jobs if it saves you lot of time then the cost is justified. Plus just one descent sized project can pay for the machine.
reForm
10-30-2005, 07:02 PM
Wow... thats a pretty good comparison there. Thanks for that.

Looks like the X2 is the way to go then for sure. Especially if you take into account the overclockability of the X2.

Has anyone pushed the dual opteron's to see how far they overclock?
Ed Caracappa
10-31-2005, 01:13 PM
meanadam,

i'm not sure what you mean when you say, "justify the cost of the boxx"?

Ed
GregHess
10-31-2005, 01:39 PM
Ed,

Vray still seems to give a slight edge to HT enabled processors (Xeon's), but not from a price/performance standpoint.

EGZ has created a wonderful vray benchmark which many owners have contributed to in this thread.

http://www.chaosgroup.com/forum/phpBB2/viewtopic.php?t=12220&start=0

You might be able to email the admins of the forum to get unrestricted access to this section (in case you can't see it). Generally I think only vray owners can access certain sections to help eliminate those individuals with problems caused by various pirated copies.

Here's some quick snipets of performance data (in case you can't see the thread immediately).

Dual 3.2 Xeon HT (2Gb) 4 min 59 s
Dual 2.8 Xeon HT (1Gb) 6 min 29.3 s
Single AMD 4400x2 (1Gb) 5 min 47.2 s
Single AMD 3200+ (1Gb) 13 min 30.2 s
Dual Opt 242 (1Gb) 8 min 48.4 s
Dual Opt 275 (DC) (4GB) 3 min 8.1 s

There is alot more data (Thats only going through 3 of the 10 pages). The dual cores are rediculously fast. Based on the Vray data, I'd say a single dualcore processor is faster then two equivilantly spec'd single core opterons. (ie. single dualcore 2 ghz > dual single core 2.0 ghz).
CGTalk Moderation
10-31-2005, 01:39 PM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.