Manta Interactive Ray Tracer Development Mailing List

Text archives Help


Re: [Manta] Re: manta


Chronological Thread 
  • From: "Brian Budge" <brian.budge@gmail.com>
  • To: "Ingo Wald" <ingowald@gmail.com>
  • Cc: "manta@sci.utah.edu" <manta@sci.utah.edu>
  • Subject: Re: [Manta] Re: manta
  • Date: Sun, 23 Dec 2007 21:07:41 -0800
  • Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=message-id:date:from:to:subject:cc:in-reply-to:mime-version:content-type:references; b=Jd99KDacph18rc75Ai+4Q3YCu+fVjUuA9UYCZM7WSeACk0qSbAwxz6xVS2gVpHjALOi8m1DK5FkuhUKBaNDnAsNP6nJMBHAxhW15vqaTFShdc2Ka2+RP3qLJjTFAndn+vZFviEPUYa2b2Bw7jK0nL0QIlK9HUW4OEVTV3BazZYk=
Hi Ingo -

I just ran a couple of experiments through again.  It looks as though the DynBVH is definitely faster than both kd-tree implementations.  Running the 1.5 million triangle buddha scene, I get the following timings (at 1280x720):

DynBVH: 1.1 seconds/frame
Manta KD: 1.7 seconds/frame
My KD: 1.4 seconds/frame

I'm a total Manta newbie, so it may be possible that I don't have packet tracing turned on somehow?  I think it's on by default?

I guess there is an untold story here too.  If I run smaller scenes (say 7,500 triangles), the timings are quite a bit different:

DynBVH: 0.33 seconds/frame
Manta KD: 0.3 seconds/frame
My KD: 1.2 seconds/frame

I can't say I understand this too well, but I would venture a guess that the packets stay coherent much longer with the smaller scene.  Either that or I am doing something stupid with small scenes that doesn't show up with larger scenes.

Any insights?

BTW, I am using a simple traversal scheme very similar to the one in your thesis.

Thanks,
  Brian

On Dec 23, 2007 1:12 AM, Ingo Wald < ingowald@gmail.com > wrote:

>
>     > for the viewer?  My own kd-tree ray tracer without sse is about the
>     > same
>     > speed as tte DynBVH that you guys have, but I would guess that your
>     > kd-tree
>     > is a good deal faster.
>
Hi Brian,

I actually doubt it is ;-).
I've written the SSE'd kd-tree traverser and kd-SAH builder (assuming
the  one in manta is the one I wrote for Thiago's project...), but it
certainly isn't faster than DynBVH (how could it !?).
I'm also a bit surprised that your _single-ray_ kd-tree is as fast as
DynBVH, which would be _very_ strange indeed (in theory, "perf(SSE
packet/frustum kd-tree) > perf(single-ray kd-tree)" and "perf(dynbvh) >
perf(SSE kd-tree)", so your message confuses me a bit.

Can you elaborate a bit on
a) what traverser you actually use, and
b) what performance you're getting with dynbvh ? anything close to the
numbers we reported in the paper ?

Ingo



Archive powered by MHonArc 2.6.16.

Top of page