2000x GPU Performance Needed To Reach Anatomical Graphics Limits For Gaming? 331
Vigile writes "In a talk earlier this year at DICE, Epic Games' Tim Sweeney discussed the state of computing hardware as it relates to gaming. While there is a rising sentiment in the gaming world that the current generation consoles are 'good enough' and that the next generation of consoles might be the last, Sweeney thinks that is way off base. He debates the claim with some interesting numbers, including the amount of processing and triangle power required to match human anatomical peaks. While we are only a factor of 50x from the necessary level of triangle processing, there is 2000x increase required to meet the 5000 TFLOPS Sweeney thinks will be needed for the 8000x4000 resolution screens of the future. It would seem that the 'good enough' sentiment is still a long way off for developers."
Ha (Score:2, Informative)
Tim's explanations of first- and second- and third-order approximations are somewhat bizarro. Unreal doesn't use second-bounce in its lighting. All game engines are first-bounce only unless they contain some realtime radiosity simulation, and very very few do. This has been true since Wolf 3D and is true today.
And once you have a system for second-bounce, third- and fourth-bounce can be trivially computed (over multiple frames if need be), and the results are hardly different to second-bounce.
I wish I knew what he meant by these levels of approximation.
Re:Development costs? (Score:2, Informative)
I have bills to pay and much better things to do with my time and money than to spend half a C-note on hardware
Fifty dollars on hardware doesn't sound that bad, honestly. Presumably that will last for at least a couple years.
Not far fetched... or far off (Score:2, Informative)
Exponential improvement in technology is the historical norm, yet it can still be difficult to fathom.
2000X should be achievable by 2024, at 2x improvement per year; or by 2029 at 2X every 18 months.
Some of us should see 2 trillion-fold improvement in about 40+ years at 2X per year; or by 2075 at 2X every 18 months.
Barring the occurrence of any variety of manmade and natural disasters, of course.
Re:Development costs? (Score:2, Informative)
No.
Many SNES games were 75-80 dollars.
Re:Development costs? (Score:4, Informative)
A gaming rig that can more than handle medium settings of any modern game shouldn't cost you more than $1000 to build, and almost certainly wouldn't be getting used JUST for gaming by anyone who is budget conscious.
The rig I currently game (and do a lot of work and other personal computing use) on was just over $1000 when I set it up 2 years ago and I can play modern games at medium settings no problem, and have decent frame rates.
It came with:
- i7 860 @ 2.8GHz
- 8GB RAM
- ATI Radeon 5850
- Win 7 64
- Running 2 displays at 1920x1080
I added:
- 8GB more RAM (I do a lot of work via virtualization - I highly doubt any games I'm playing even use the base 8 that I had originally). $200 when I got it - and though I wrote it off, I'll add it to my "gaming" cost
- 64 GB Intel SSD that I install games I'm actively playing on; $90 on Newegg when I got it.
I'll use this system for 2 more years before replacing it and turning it into a server, but let's pretend I'll throw it in the garbage, so it comes out to $325/year invested.
I won't even try to pro-rate the cost due to work I do with it or personal, non-gaming use, so let's pretend that my gaming hardware costs me $325 a year with no other benefits.
If you're saying that a $325/year investment in multi-purpose hardware is too much, but dropping $60 at a shot to play modern games is cheap, you have very, very weird budgeting.
And I'm assuming I'm right on the $60 price because that's the price of the bleeding edge games that I'm assuming you imagine require insanely expensive hardware to run (but that I'm quite capable of playing on my rig)
Or maybe your machines cost sub $500 to put together, in which case, yeah, you're not going to be having a very fluid gaming experience.
Re:Rasterization (Score:4, Informative)
Because raytracing uses so much less computing resources? Or, because you don't really know what it is?
For very complex scenes, yes it does use less resources. Raytracing grows logarithmically while Rasterization grows linearly. Intel estimates the scene complexity (whats visible) where Raytracing overtakes Rasterization is ~1 million polygons. After that, Rasterization has no chance to compete in efficiency.
Re:We dont realy need 8000x4000 with eye-tracking (Score:4, Informative)
Re:Development costs? (Score:4, Informative)
But the PS2 game "ICO" taught me a few things. It's hard to explain the impact the graphics had when the game came out. Particularly the trees...they look absolutely amazing for a PS2 game which was actually developed for the PS1 (it fits on a CD, rather than DVD).
I tried to get a close FPV on the leaves, and I realized there weren't any leaves. Just simple shapes that shimmered, glittered and moved in mass like a tree. The PS1-era developers didn't have anywhere near enough polygons to actually generate leaves; they didn't have raytracing hardware to simulate light glittering off millions of leaves, and they didn't have subsurface scattering to model light going through the leaves. But it didn't matter, because they managed to hack something that looks just like a fucking tree from any reasonable distance. They didn't synthesize a TREE...they synthesized something that looked like a tree, using minimal primitive elements arranged to give a stunning impression of a tree--some real Bob Ross shit.
In other parts of the game, there are what appear to be very realistic dust effects and lighting effects (in the cathedral area). These effects were just amazing at the time...beautiful. A closer inspection shows that they just hand-placed luminescent polygons to construct every shaft of light in the cathedral, and the apparent dust effects are just moving texture on the polygons. Again...no ray tracing, no particle effects, but they made something that looked absolutely convincing, the way a good painter can give an impression of light paint and canvas--basically human visual cortex hacking.
There is no point to this post, except that there is more to creating good graphics than technology.