Real-time ray tracing gets an airing

"... a small hardware accelerator and some very innovative software to be able to deliver real-time, complex, high-resolution raytraced images... Caustic's first-generation technology will deliver an average 20X increase in the speed used to create stunning, realistic 3D imagery for film and video, game development, as well as automotive and consumer product design. The second generation of Caustic's technology, due early next year, is expected to gain an additional order of magnitude in performance, offering 200X speed over today's state-of-the-art graphics products. This massive speed jump is due to Caustic's patent-pending raytracing algorithms implemented in a semiconductor design... "

'doesn't sound exactly like it works the same as a PPU.

This is for enabling designers and animators beyond the confines of the render farm.

Awk! There it were all along right before me eyes.

Trouble is. Now, they'll need to start all over for Duke Nukem with a ray gun.

"However the word on the web is that even if Caustic is able to do what it says, it might be a bit hard to find someone who wants to buy it."

*someone* like Intel or AMD? ;-)

200x?
Crysis at 4,000 fps????

More likely from 0,01fps to 2fps...

I don't think the technology here is really complex. Looks like a duel processor single board computer with dedicated RAM per CPU. And, I would bet that a notebook GPU lies under that blue heat sink. So everything revolutionary in probably in software. Which is not a bad idea. Nice to see people are still working on getting ray tracing out there.

honestly it seems great on paper, but what about when the game doesn't support its technology? what performance gains are to be had in actually good games?

sure, I buy some ridiculously expensive piece of hardware, install drivers for it and such, only to find out 1 in 20 games utilize its technology.

fantastic, another PhysX clone!, no thanks.

Saarcor

They did the same thing looooong time ago....

Altough I doubt they will sell more than a thousand cards, the step is in the right direction.

Yes, it's the same story as with PhysX, but if Ageia never pushed it, we wouldn have PhysX engines nowdays nor the ability of GPU calculations of physix in games.

This could be the same thing. One shouldn't bash the company for it. The community should support the effort so we can have raytracing on CPU (or GPU) in 2-3 years time. But someone has to do the hard work, and these guys are doing it...

This is NOT something for the gamer. It's for the visualisation industry, whether it's film FX studios, design agencies, or architectural agencies.
This type of raytracing accelerators have been out for quite a while - the ARTVPS Renderdrive & Raybox are the most popular.
Howeverer it seems very strange to me that at the same moment ARTVPS have stopped making hardware accelerators and focus on software, a new company arrives to take it's place, with a strangely similar product.
I guess some hardware engineers just made their own venture and started creating their own ridiculously expensive hardware.
I also remember some guy who modified a common Celeron processor into a raytrace accelerator. This could be based on it.

The press release states..
Caustic's first-generation technology will deliver an average 20X increase in the speed used to create stunning, realistic 3D imagery for film and video, game development....
20x increase in what exactly...time taken to calculate something that takes about 2 minutes? or some compromised algorythm that fakes GI so much that the enhanced lighting techniques used in DX10 engines would probably look better. Ray tracing is ray tracing...optimise and simulate all you like but brute force calculations are the only way to go - otherwise George Lucas would have jumped on this ages ago for use in real-time pre-vis... Best think i've seen is Cinema 2.0 and thats years away from being commercially viable...

Wow, old news. Yet, it makes me notice we still don't have any benchmarks.

So they have 15 patents? I wonder how many patents they _don’t_ have, that some patent troll will emerge and sue them over, the moment they show signs of achieving any market success.

Amazing that nobody seems excited about what this company is about to launch in the near term and at the end of 2009. Basically, nothing fresh or innovative has come about in the high-end AIB or discrete graphics market in ages, and yet these guys seem to have something special, and yet nobody gives them a chance. This is a billion dollar a year business with 70% gross profit margins, think it might deserve a few seconds of attention given the game changing possibilities. Gamers are a tough lot to crack, they're basicallu skeptics and conspricy theorist crumudgins whose lack of vitamin D makes them a day late a dollar short to most everything, but CG studio professionals and powerhouse 3d designers are waiting for something that enables the next big jump. I think we should br digging deeper, rather than be so quick to dismiss.

Here, here Chief...reading a lot about Caustic lately and can't help but think that the overwhelming dismissiveness is akin to the story of US Auto Industry. 'Nobody wants a hybrid, bigger is better' that got those boys and our country in a proper mess. Well maybe these guys have something, an approach worth looking at, an approach that is fresh, of course maybe not but Far Cry 2 running at death slow speeds ain't cutting it for me, one Ratatuille every 5 years is lame - we need to think bigger expect more and stop tripping over our feet. If not Caustic than hopefully someone will figure this out. Rasterization is an SUV, Tesla anyone?