Dreadnought Demos Released
John Callaham writes to tell us that Gamecloud is heralding the latest release from Torc Interactive and AMD. The latest demos for the upcoming FPS, Dreadnought, have been released. The first is strictly a gameplay movie while the other gives a comparison between the game running on a 64 bit processor (which it was ultimately designed for) and a 32 bit processor.
It's interesting that AMD is pushing their 64 bit technology with this game. If it weren't for video games then what other reason would we have to continue to build faster computers?
Bradley Holt
If it weren't for video games then what other reason would we have to continue to build faster computers?
one word: porn
drunk chemists
******************
# uncompressed normal maps allowing for higher texture quality and greater detail
# significantly higher number of particle effects (e.g. more flames, more steam, more smoke, etc.)
# persistent decals (e.g. bullet holes stay on walls and don't fade away over time as in 32-bit)
# post-processing effects (e.g. screen glows)
# more pixel shader instructions (the adrenaline vision mode is built upon and replaces the base lighting shader to produce the effect)
*****************
there's the rundown on the 32bit vs 64 bit changes.
now, call me an idiot if you will, but none of those really smell like something they couldn't have done in 32 bit - which makes the 64bit vs. 32bit comparision TOTALLY FRIGGIN USELESS unless you're a phb or something. mostly it just seems like they assume 64 bit system to have more memory and a faster graphics card tied to it.
world was created 5 seconds before this post as it is.
This isn't an upcoming FPS. This is a tech demo for Torc's game engine and AMD's AWESOME64(TM) processors.
I'm a little curious to see what the inside of a Kirov class cruiser looks like. I doubt even the CIA or the navy knows exactly. Based on what I've been allowed to see of our own conventionally powered Aegis cruisers and destroyers in the tours I've been on, and the fact that the layout of nuclear-powered ships is even more protected in both the US navies, there has to be a huge amount of conjecture going on here.
Still, I've always been interested in playing a game that effectively encompasses all of a relatively small area, rather than meandering through small parts of a huge area. Quite a few pictures of the Kirovs are publicly available, and if they study other ships (retired destroyers, battleships, etc), they should be able to create a very believable environment, with all sorts of mundane curiousities to explore. If the entire game really takes place on this ship, notwithstanding the fact that it's dimensionally nearly as big as an Iowa class battleship, it should lend itself well to that sort of game design. Being on a ship also offers some fun. For example, they could change the weather and sea states throughout the game (how about throwing a grenade in a 40 knot wind?). I thought the varying weather was one of the nice subtleties of Enigma: Rising Tide (although that's not an FPS).
Of course, it could end up simply sucking like most other games out there.
All of what they say pretty much pertains only to the graphics card's capabilities.
No compression for textures? Cards with 512MB are the hot things now, and there's not really a good use for it. With 32 bit addressing, we can address 4GB of memory. That's what? 8 times what's currently available on anything less than a ginormous SGI simulation center? Yeah. 64 bit doesn't help us there, not even in the long term.
There's not a graphics card alive that's going to need 64 bit addressing to render literally billions of particles, and there won't be for at least 10 years, barring some extreme advances, or the use of alien technology (teehee). Same with decals, even if you "only" had memory to store the location of 512 million of 'em, there's no way the system will handle displaying even a few thousand all at once.
Glows? Unless they need 64 precision math done on the CPU (which they don't), yeah, non-issue. Consumer GPUs are limited to what? 24 bit plus alpha? Same for pixel shaders, this has nothing to do with the CPU in almost all instances.
So yeah, for games, as with most general purpose computing, this is pretty much useless. What's really sad is that they've rallied around arguments for their 64 bit push that are essentially limited by the decidedly non-64 bit GPU. Brilliant.
Constitutional rights may be respected, repealed, or modified; but they must never be ignored.
Glows actually can be helped by 64+ bits, but not CPU bits. All of the latest generation of GPUs, and some previous (basically all DirectX 9 GPUs) support colour depths greater than 32-bit. A 32-bit colour depth is 8 bits per colour, and 8 bits for transparancy. Given that 8-bits per channel is the normal output, it was generally thought to be enough. However anyone who's worked with shaders will tell you that small errors start to add up, and you can get nasty results. So new GPUs support 64-bit and 128-bit floating point colour.
Now because of the extra bits, and because it's a FP representation, you get much better colour, highlights, and shadows. Though the output is ultimately the same resolution, you don't have the clipping problems you used to, and you don't have errors that add up to incorrect colours.
Both nVidia and ATi have a bunch of demos that'll show you this, if you like. Any Radeon 9 or X series or GeForce 6 or 7 series will handle it.
However none of this is relivant to a 64-bit CPU. This can, and does work great on a 32-bit CPU. It's all internal to the GPU. Even if the CPU needed to do some work as 64-bit CPU would be irrelivant as only the integer unit is increased in size. FP units have been larger than 32-bits for a long time, current CPUs generally can handle up to 128-bit FP numbers, depending on what you are doing.
So you are correct, all fluff as far as the CPU is concerned.
doesn't seem to have removed this. Time to update my filter list!
No, I didn't think so. Does anyone even use Windows around here? Right, so why post links to .exe movies?