PDA

View Full Version : Lightwave v9.0 and intel optimization...



SaturnX
10-22-2005, 05:36 AM
A simple question.
Will amd/amd64 users of lightwave v9... suffer a slow down in rendering speed... becasue of intel bias ?

And will amd64's that have onboard SSE3 help ?

hrgiger
10-22-2005, 06:41 AM
I hope so. Of course, I'm biased too since I use an Intel processor.

SaturnX
10-22-2005, 06:57 AM
Yeah ok. Cool.
Point taken.

hrgiger
10-22-2005, 07:02 AM
Actually Saturn, I wouldn't worry too much about it. Lightwave 9 is supposed to have speed increases for scenes that involve raytracing by 2-3X. The times may vary according to the processor but it will probably be a matter of a few seconds. I'm sure everyone will be happy. Somebody was also saying by the time 9 rolls out, they think it could be more then 2-3X but we shouldn't speculate on that yet until we see the release.

Lewis
10-22-2005, 07:06 AM
LW have AMD optimizations long time ago (i think it was in 6.5b when the added AMDNow+ instructions side by side wiht AltiVec, MMX and SSE) so i don't see reason why would AMD based machines suffer from any slowdown.

Cageman
10-22-2005, 12:03 PM
LW have AMD optimizations long time ago (i think it was in 6.5b when the added AMDNow+ instructions side by side wiht AltiVec, MMX and SSE) so i don't see reason why would AMD based machines suffer from any slowdown.

Well, as an example. I had a scene with alot of hypervoxel smoke. I rendered a frame on my dual AMD Athlon 1.8 GHz, using 2 threads. It took about 16 minutes. I took the scene over to my laptop, a Ferrari 3400, 2GHz AMD64. and the exact same frame rendered in about 9 minutes. The difference between the processors is not only 200MHz, but the fact that the 64bit AMD has SSE2, which the 32-bit amd doesn't have.

EDIT: The speed-difference in this case is alot more, in favor for the dual athlon, which, when rendering with both processors, is 3.6GHz, in theory at least. Still, it got beaten by a single 2GHz chip, and I suspect that it has to do with the lack of SSE2 in the 32-bit athlons :)

CAClark
10-22-2005, 12:09 PM
Actually Saturn, I wouldn't worry too much about it. Lightwave 9 is supposed to have speed increases for scenes that involve raytracing by 2-3X. The times may vary according to the processor but it will probably be a matter of a few seconds. I'm sure everyone will be happy. Somebody was also saying by the time 9 rolls out, they think it could be more then 2-3X but we shouldn't speculate on that yet until we see the release.

When it comes to NT and claims of speed, I'll believe it when I see it. There seems to be a habit of a degree of hype being built up that just isn't fulfilled on delivery. 9 has the makings to change that, but the doubt is always there.

lots
10-22-2005, 08:58 PM
AMD CPUs already have SSE3, so there should be less of a "optimization deficit" should NT decide to impliment SSE3 (the main reason the "optimized for Intel" thing came about was SSE2 was on the P4 for quite some time.. Until the K8 came out, AMD did not support SSE2. Hence the big difference between P4 and Athlon in LW...)

Also, AMD is planning on releasing an instruction set geared towards speeding up 3D rendering, sometime in the future... Though what I read was not too clear on if this was in reference to realtime 3D in games, or if this was meant as the kind of rendering we're used to... Or both.. Anyway In the future, since AMD CPUs are gradually getting more and more market share, NT will have more reason to have Optimizations for these CPUs, since more and more customers will be using them. Though if this translates into actual fact remains to be seen :)

Steve McRae
10-23-2005, 06:38 PM
anything new on the release date of 9?