It seems they finally found a market for their product; Euclideon Geoverse.
Irf-HJ4fBls
An impressive looking tool; I can’t wait for it to be licensed to replace Google Earth, so we can all take a spin around a virtual representation of the real world 
This again, lol. Oh so thats how it works! Yeah ok cant wait for a demo.
“Here at Euclideon, we’ve found a way to make this work! Now let me continue not telling you how we did it!”
That’s a little unfair. They do say it’s a ray cast per pixel with clever search through a heavily precomputed index.
What was most impressive to me was the evidently modest bandwidth requirements (and downward scalability), permitting the data to be entirely centralised.
If the ‘heavy lifting’ (doing the ray casting & searches) is done clientside, the serverside load should be quite light & scalable to many simultaneous users.
I do wonder if this new target audience has been motivated solely by money, or if they finally conceded that completely static terrain (however highly detailed) isn’t that useful for game developers.
*Though I imagine it’d be a great fit for flight simulators.
I hate how he keeps saying “unlimited data”, which is simply impossible and silly to even mention. However, I’d like to try out a demo.
By “unlimited data” he means that their system places no practical limits on the amount of data you have, which it would appear is an entirely plausible claim. I’m not surprised they’re not telling anyone how it’s done; I can guess but the actual specifics must be extremely clever and novel.
Cas
I’m curious about the “3.5 billion points per hour” statement. (@7:30 in the video)
Giving specific numbers for something that is obviously a product of the applied computational capacity makes me think that they’re keeping a tight hold of the indexing algorithm, and selling the conversion process as a service.
IMO that kind of a approach is going to really stifle adoption outside the enterprise market.
just watch the video?
i don’t understand the hate for this and previous videos.
of course they don’t tell you in detail how it’s done because if they would the algorithm is copied in no time and they want to make millions with the unique technology.
if anyone is interested these youtube videos try to explain and implement something similar in 2d (inkl. source code):
Actually, that’s not quite right.
The best method for anything software related is usually available to everyone.
If the source is hidden, it probably means there are some flaws.
Chances are, with a project like this, there will be many flaws.
If they release the technique too early, someone will quickly create a more efficient method than them, and they will lose part of the credit.
If they try and make money off it (patent it?), they will fail.
There is simply not enough reasons for someone to pay for this.
HeroesGraveDev, you are so very wrong. They’re making a mint out of this tech because no-one else knows how to do it.
Cas
I lol’d yesterday xD
I’m not surprised with their change of costumer base they are aiming for. Perhaps they started wanting to develop some cool game tech, but realised that especially with the tech they have they can make more and more easily money in other industries.
Vastly more money.
Cas
Can this even be applied to games? Wouldn’t storing billions of tiny dots take a lot of disk space, even if you can render it quickly (i.e. you need super small maps to be able to actually use this on an average PC)?
Looking at the video, I don’t see how it could be applied to games since games have a huge tendency to rely on RAM. (Though it could help a lot in the background and CG effects.) I don’t want to go into too much detail, but I can see the potential of this to help with more than just Google Maps if the technology is as good as it looks.