John Carmack interview talks PS3 & X360

Thinks in next-next-gen (one coming up after this one, ie PS4/Xbox3) that it will be a lot harder to tell the difference between games and CG unless you closely examine it

agreed. but only to an extent. there is a massive range of complexity and quality within what is called CG. (obviously). I don't think Xbox3 or PS4 will be capable of rivaling current (2000-2005) movie/film CG let alone film CG that will be around during those console's lifetimes. I think Xbox3 and PS4 will be able to rival current videogame CG used in PS2/Xbox/Gamecube/PC games. the Xbox360 and PS3 will likely give us graphics that are around the level of PS1 CG, with more detail, but not better quality (sometimes even lower quality).
 
The four key owners of id Software were on G4tv last night...


The four key owners of id Software were on G4tv last night, and they had some interesting quotes that I thought I’d highlight in case you missed them:

John Carmack: “Xbox 360 has far and away the best development toolsâ€￾

John Carmack: “[talks about software development support]…and the hardware is comparableâ€￾

Geoff: “So you don’t think PS3 is going to more powerful [than Xbox 360]?â€￾
John Carmack: “PS3 is probably marginally more powerful, in terms of raw flops and graphic operations, but that’s not really the best way to look at things. When you look at these development cycles that stretch over years and years, being 20% easier to develop on is much more important than being 20% more powerful.â€￾

John Carmack: “I make little nitpicky decisions about say, well, I prefer the symmetric approach that MS has over the asymmetric Cell approach, but you can do great games on either one of them, and I make fundamental decisions based on development tools and depth of documentation, which Microsoft has been superior on.â€￾
 
xbdestroya said:
Well, Carmack's a PC guy, and 360 more closely resembles a PC developing environment than does so the PS3, so it's not surprising he would prefer it. Plus with his propensity to develop in the FPS genre, and 360's demographics lining up with that perfectly, there are any number of reasons for Carmack to prefer 360 to the PS3.
Ummm...no.

Doesn't the PS3 use OpenGL? Doesn't Carmack prefer OpenGL over DirectX? I think the "Xbox 360 is closer to a PC so PC developers like it" argument is a cop-out and so last gen. There's nothing in the Xbox 360 that you'll find in a PC or that you will find in a PC anytime soon. Hell, one can even argue that the PS3 is more PC like since it has the modified PC GPU.

PS3 is basically an OpenGL box with an Nvidia GPU. Carmack should be loving that shit. But he doesn't. In fact, he prefers a DirectX architecture over it. That says a whole lot considering who we're talking about.
 
Alpha_Spartan said:
I think the "Xbox 360 is closer to a PC so PC developers like it" argument is a cop-out and so last gen. There's nothing in the Xbox 360 that you'll find in a PC or that you will find in a PC anytime soon. Hell, one can even argue that the PS3 is more PC like since it has the modified PC GPU.

John Carmack: “I make little nitpicky decisions about say, well, I prefer the symmetric approach that MS has over the asymmetric Cell approach

It is not about the GPU. PC's already have dual core. a.k.a. a symmetric scheme, which is fundamentally closer to the X360's CPU scheme than the PS3's.
 
Last edited by a moderator:
Sony has made it quite clear they are using OpenGL.

I think John's comments are directly related to the CPU differences, and also having a less mature development system on the PS3 side. John background being a PC developer, and not a console developer is hardly surprising.

Console developers will certainly see the differences and jump on taking advantage of the respective differences, that a PC developer would be less than thrilled to participate in.

We have had 64-bit extensions in the PC world for quite awhile, and dual core for a decent while, and yet how many games take advantage of this. PC developers are very slow at embracing changes on the CPU side, so of course they are going to be less happy to get great performance out of CELL's state-of-the-art processors.
 
Last edited by a moderator:
Alpha_Spartan said:
Ummm...no.

Doesn't the PS3 use OpenGL? Doesn't Carmack prefer OpenGL over DirectX? I think the "Xbox 360 is closer to a PC so PC developers like it" argument is a cop-out and so last gen. There's nothing in the Xbox 360 that you'll find in a PC or that you will find in a PC anytime soon. Hell, one can even argue that the PS3 is more PC like since it has the modified PC GPU.

PS3 is basically an OpenGL box with an Nvidia GPU. Carmack should be loving that shit. But he doesn't. In fact, he prefers a DirectX architecture over it. That says a whole lot considering who we're talking about.

hmm, i wonder which console's alpha dev kit was a dual g4 mac with a x800..

of the two revealed next gen consoles the xbox360 definitely is the one that is closer to a dektop pc. the fact that 360 it is not a re-packed pc bolted around a uma which was the case with microsoft's previous "console" does not mean that the ps3 is more like a pc that the xbox360 is. add on top of it the software environment (dx api and nt-based os kernel) and you may directly go full subscription for msdn ; )
 
Edge said:
We have had 64-bit extensions in the PC world for quite awhile, and dual core for a decent while, and yet how many games take advantage of this. PC developers are very slow at embracing changes on the CPU side, so of course they are going to be less happy to get great performance out of CELL's state-of-the-art processors.

That's not always true:

Quake III had SMP support and Quake 1 supported the Pentium Pro. Unreal was one of the first to support MMX, etc.

Most of these and like 64-bit extentions and multi-core are not switches you turn for for more performance. If you only enable them you'd most likely get a performance decrease instead. For instance, Things have to mature and new groundwork has to be laid to properly take advantage of more hardware features.

If on top of this you also have less than optimal documentation and/or dev tools you're going to take longer to make use of those features properly.
 
deep voice from outter space: "whuhaha, you cant kill my thread muhhahhah...."

ohwell : at least he says that the ps3 is more powerfull then xbox360. that should make the fans happy.


and if we didnt, we would blame it on other things like moneyhats from MS
 
Ya he says it's more powerful on paper, which mean approximately nothing! PS2 had more flops on paper, the 7800 has more flops on paper than the 520...so flops = meaningless.

As for which is closer to a PC isn't it sort of split? With the CPU for X360 being much closer to PC than CELL, however the RSX will be much closer to traditional GPU's than Xenos?

So would it be fair to say all dev's would have a greater learning curve with CELL than Xenon, and alternatively they will have a great learning curve with Xenos than the RSX. Both consoles have tons of potential, in different areas.
 
Last edited by a moderator:
Megadrive1988 said:
the Xbox360 and PS3 will likely give us graphics that are around the level of PS1 CG, with more detail, but not better quality (sometimes even lower quality).

Wrong, wrong, wrong. We have already seen games that look better than PS1 CG. Your standards are set too low.
 
scooby_dooby said:
So would it be fair to say PC dev's would have a greater learning curve with CELL than Xenon, and alternatively they will have a great learning curve with Xenos than the RSX.

Probably, but Xenos is using D3D. So if both console's gfx parts have PC API support it all boils down to CPU (and general APIs).
 
Forgive me I don't sound like I know what I'm talking about, cause I don't, but I thought stuff like hardware tesselation, MEMEXPORT, unified shaders and EDRAM where all things offer alot of extra power/flexibility over a traditional PC GPU, yet dev's will have to master these techniques as time goes on.

For example, doesn't the MEMEXPORT capabilities of Xenos enable it to *possibly* perform physics calculations in a more feasable way than typical PC GPU's? aren't there many other cool new things dev's will be able to use MEMEXPORT and EDRAM for? I'm presuming these aren't going to happen overnight...

Won't the unified shaders allow dev's to break be more flexible in their game engines, not having to rely on hardwired shaders(i.e. 20% vertex 80% pixel shading)

Isn't the hardware tesselator something that will take Dev's a while to really master/utilize?

These are just off the top of my head, but they all seem like features that, in the hands of creative masterful console programmers offer a ton of exciting flexibility and power, yet will take some time to get over the learning curve.

Compared to the RSX which is basically a straight PC GPU (yes I'm assuming) which would not have any real learning curve. I don't think it's strictly about the CPU's.

Would I be correct in saying the Xenos GPU will take longer to reach it's potential than the RSX will (lets assume RSX is overclocked G70) and alternatively the CELL will take longer to reach it's potnetial than th XeCPU? That seems like a fair assessment.
 
Last edited by a moderator:
Edge said:
I think John's comments are directly related to the CPU differences, and also having a less mature development system on the PS3 side. John background being a PC developer, and not a console developer is hardly surprising.

Console developers will certainly see the differences and jump on taking advantage of the respective differences, that a PC developer would be less than thrilled to participate in

Are you saying his opinion on development environemnts is not valid becuase he has primarily developed on the PC?

I think we should give this carmack guy the benefit of the doubt, seeing as he created a whole gaming genre for us and ushered in 3d acceleration and all..

:)
 
randycat99 said:
...for PC's. (this reasoning keeps going in circles)

Hmm, as for the future, I don't see a lot of differences between console and PC (other than currently it's closed environment vs open). However, when the PC was in its infancy it was pretty much a closed system. Developers had to be very familiar with the hardware and dig deep down to get every extra cycle out of it. As the plethora of devices and configurations started to set in, developers started to abstracted out the interfaces.

No doubt about it. The future gaming console will move toward that direction. Given the complexity of current gen, a lot of game developers are started to use middlewares to shorten their development cycle. On top of that, both MS and Sony wanted people going through their API instead of directly accessing the hardware. This allows them to provide a migration path for the future console and maintain a pool of highly skilled high-level programmers that can quickly develop quality software for the new platform instead of spending time trying to learn the new hardware.

Because of this, I don't see any reason to discredit JC.

edit:

Actually, probably because of his PC background, he's probably more justified in his assessment.
 
Last edited by a moderator:
The comment wasn't to "discredit". It was context.

Nevertheless, it is still a true statement. It is a relevant statement.
 
Back
Top