Release Candidate for Torque3D
by TheMartian · in Torque 3D Professional · 07/31/2009 (1:14 pm) · 86 replies
So we are at Beta 4 for Torque3D which is really cool. Definitely building up some excitement.
Now the meat of the question, is there a time frame you are shooting for to get to a Release Candidate for Torque3D?
(so in other words you have a ballpark ship date yet?)
Thanks!
Now the meat of the question, is there a time frame you are shooting for to get to a Release Candidate for Torque3D?
(so in other words you have a ballpark ship date yet?)
Thanks!
#82
08/07/2009 (12:23 pm)
The last thing GG should do is release early to compete in the marketplace. Bad press and unhappy users will smash any success T3D could have. I would assume that Unity is really making things difficult as they just doubled their team size. I personally get horrible performance with Unity (not that my Torque 3D performance is that great either). I get better performance with high settings on Crysis than I do with Unity (and T3D). Hopefully this changes once out of beta. 
#83
08/08/2009 (4:52 pm)
Who's Unity? :) 
#84
08/08/2009 (6:46 pm)
@Black Tengu: If I had to guess, I would say your card has poor memory bandwidth. If you were to attempt to run Crysis 3, it would perform similarly to T3D. Crysis 1 and 2 both use forward renderers, T3D Advanced Lighting (and Crysis 3) use pre-pass lighting.
#85
08/08/2009 (8:18 pm)
@Pat: Quite possible. I have an 8800GT 512MB. Not the best card.
#86
How/why does this matter? Well, in all types of deferred shading, there is a good deal of full screen buffer transfer going on. As your screen-resolution increases, the copy-bandwidth requirements increase exponentially. It has to push, and read, a huge amount of pixel data, especially if you are using 64-bit buffers (which Torque3D does by default so that everything looks good always. If your game is tuned properly, you can use a 32-bit Gbuffer.)
08/08/2009 (9:33 pm)
It sounds crazy, but the memory bandwidth difference between every version of the 8800, ranges from 38.4GB/s - 103.7GB/s. The top end 8800 has ~3x the memory bandwidth of the low end. A GeForce 8300 GS has only 6.4 GB/s.How/why does this matter? Well, in all types of deferred shading, there is a good deal of full screen buffer transfer going on. As your screen-resolution increases, the copy-bandwidth requirements increase exponentially. It has to push, and read, a huge amount of pixel data, especially if you are using 64-bit buffers (which Torque3D does by default so that everything looks good always. If your game is tuned properly, you can use a 32-bit Gbuffer.)
Associate Steve Acaster
[YorkshireRifles.com]
edit
It's a light-hearted way of saying, "save for a few current and obvious issues (all documented by the testers), the beta has worked out very well."
double edit for bad grammar