Xbox 360 Selected as the Primary Development Platform for Next-Gen Wolfenstein

Nite_Hawk said:
I get the feeling that he laments the move to multicore processors instead of looking forward to it. You get the same feeling from Gabe Newell. Then you get people like those from Ninja Theory and Epic who while acknowleding the problems, actually sound excited about solving them.

He may be looking beyond id's and his own business here, you know. In the past decade he has been usually trying to act for the good of the whole industry, and I'm quite sure that his objections against this multicore stuff is based on how hard it will make the life of the majority of game developers... not everyone is a Team Ninja or Epic...
 
randycat99 said:
It would actually more plausible to expect the development in X360 environment as a gateway to PS3 environment where the ppc kinship, multiple CPU core, fast/wide bus topology will map across fairly well.

Actually Crytek's CEO noted in an interview awhile back that before getting the devkits he expected PS3-360 to be more alike than 360-PC but in reality PS3-360 had the biggest difference between the two. Likely because you're dealing with so many threads with Cell and its asymmetric nature.

Of course porting from PC to PS3 would be an even bigger gap so your point still stands.
 
Last edited by a moderator:
Laa-Yosh said:
He may be looking beyond id's and his own business here, you know. In the past decade he has been usually trying to act for the good of the whole industry, and I'm quite sure that his objections against this multicore stuff is based on how hard it will make the life of the majority of game developers... not everyone is a Team Ninja or Epic...
I could not have said it better myself. Literally I could not find a way to put it. Thank you. I also get this feeling that many of the people who are saying that coding for these system is easy also have a lot of money backing them which most studios don't. In Epics case it more of a motive trying to get people to buy their engine. That is why I think middleware will rule at least the first half of this generation. Valve, Epic and ID could stand to make a killing off of it.
 
Laa-Yosh said:
He may be looking beyond id's and his own business here, you know. In the past decade he has been usually trying to act for the good of the whole industry, and I'm quite sure that his objections against this multicore stuff is based on how hard it will make the life of the majority of game developers... not everyone is a Team Ninja or Epic...

<shrug> That could certainly be the case. What good does it do though? Should we lament quantum physics because it's so much more difficult to understand than newtonian, and far fewer people understand it?

Nite_Hawk
 
Well, one possible good outcome could be to influence the console vendors to produce a next generation architecture for PS4/Xbox3 that has a smaller margin between the theoretical (marketing) performance, and the practical performance that can be attained by an average game developer studio.
Some of the stuff that Carmack has been asking for has indeed found its way into actual hardware and software before, so there's a chance that if he stands up and talks about problems, then the vendors will listen.
 
Carmack is still the man because he's on the cutting edge of technology all the time. Being old doesn't mean beans if you're staying on top of things. The guy has never quit coding or quit learning. The years of experience have also given him a great feel for what it takes to stay on top and how you gotta think in order to stay one step ahead of where everyone is now. Sure, someone will come up with a better engine than you, but Carmack's engines are significant because they're on the cusps new paradigms. Each engine--wolf3D, Doom, Quake, Quake3, and Doom3--was revolutionary in some way (yes, Descent came before Quake, but it wasn't a very flexible engine compared to the raycasting engines of the time). Epic's engines always strike me as extremely prettified syntheses of things that are already being done, with the exception of the first Unreal engine. You can bet Carmack's already thinking in the back of his mind about the Next Big Thing.
 
Laa-Yosh said:
Well, one possible good outcome could be to influence the console vendors to produce a next generation architecture for PS4/Xbox3 that has a smaller margin between the theoretical (marketing) performance, and the practical performance that can be attained by an average game developer studio.
Some of the stuff that Carmack has been asking for has indeed found its way into actual hardware and software before, so there's a chance that if he stands up and talks about problems, then the vendors will listen.

Perhaps I'm just out of the loop, but has John commented on how he'd rather things go?

Nite_Hawk
 
Laa-Yosh said:
Well, one possible good outcome could be to influence the console vendors to produce a next generation architecture for PS4/Xbox3 that has a smaller margin between the theoretical (marketing) performance, and the practical performance that can be attained by an average game developer studio.
Some of the stuff that Carmack has been asking for has indeed found its way into actual hardware and software before, so there's a chance that if he stands up and talks about problems, then the vendors will listen.
It's because the market share of Doom/Quake engines was huge and taking a good benchmark score in those games was crucial to sell video cards. Now these conditions may not be true for id software engines in the PC space, let alone the console space.
 
Nite_Hawk said:
Perhaps I'm just out of the loop, but has John commented on how he'd rather things go?

The last time he talked about this he only wanted two things: virtualised video memory and faster hardware. Otherwise he's generally very happy with both IHVs (and both consoles' graphics hardware).
 
Alstrong said:
And really, he was done with the Doom 3 engine for half the overall development time. It was the art/content creation that took so long.

I guess some have forgotten that Carmack has dealt with multiprocessors with Quake 3. Even if it's been awhile, it's still experience.
Very true. The Quake 3 engine was capable of using a dual CPU setup.
 
Platon said:
No doubt I don't think that Cramack in this day and age is really as "unique" as he once was, time has cought up with him and passed him. Still, I would really like to see him take advantage of the xbox360 architecture and make an engine that will for sure give UE3 a run for its money. As far as I know UE3 still does not have predicated tilling, which I do not understand, wouldn't that be very advantageous down the line for their next xbox360 project, not to mention that they might get even more to licence thier engine for their xbox360 games?...

After the Gears of War development process i think that the UE3 engine will come to include predicated tiling support... I'm sure it will be included in the 360 SDK.
 
Mordenkainen said:
The last time he talked about this he only wanted two things: virtualised video memory and faster hardware. Otherwise he's generally very happy with both IHVs (and both consoles' graphics hardware).

I guess it's not so much his reaction to either console that I'm curious about as much as his reaction to the the way the industry is going in general. This seems to be the angle Laa-Yosh is thinking about this from as well.

Nite_Hawk
 
fearsomepirate said:
Carmack is still the man because he's on the cutting edge of technology all the time. Being old doesn't mean beans if you're staying on top of things.
John is old :?:
 
nintenho said:
.......buh, ah forget it. The guy above me answered it perfectly. Nobody is going to have a quad-core cpu.
Very few will. And...? Since when are the cores comparable between the dual-core X86 CPU's and the 360's?
Well, look at Call of Duty 2 on XBOX360 and how that has a locked 60 fps framerate,
It doesn't have a "locked" 60fps by any stretch, there are many instances of slowdown.
how many current PC's can do that?
Not that many...so? If we pick Quake4, then does the 360 automatically suck?
 
RavenFox said:
Very true. The Quake 3 engine was capable of using a dual CPU setup.
...poory, AKAIK actually.

However, the large gains made with Quake4's dual core patch show that he's no slouch in this area. In fact, I don't think there's any other PC game out there right now which shows as significant gains with dual CPU's.
 
Dave Glue said:
However, the large gains made with Quake4's dual core patch show that he's no slouch in this area. In fact, I don't think there's any other PC game out there right now which shows as significant gains with dual CPU's.

Actually, there's been no mention of who exactly is doing the multithreaded patch specifically, although I think it's a Ravensoft thing more than JC, since he's busy working on the next internal project.

Performance gains seem to be due to offloading some of the stencil shadow work.
 
Dave Glue said:
...poory, AKAIK actually.
I assume this is based on "SMP timedemos". At that time, I'll add.

What makes it poor? What aspect of dual-CPU did John take advantage of in his "Quake3 SMP" codes?

Did you know there is support for HDR in John's Doom3 engine? Did you know this is very raw and unoptimized? It's kinda "poory" too, actually.

However, the large gains made with Quake4's dual core patch show that he's no slouch in this area. In fact, I don't think there's any other PC game out there right now which shows as significant gains with dual CPU's.
Interesting (and perplexing at the same time) you comment on Q3's "poor SMP performance" (and that is what you were saying, right?) and then go on to a different engine of John's when it comes to this.

I'm almost sorry for replying to your post based solely on your "poory" comment. I'm not defending John (yeah, like he needs it) but I think some elaboration or explanation by you should follow that "poory" description. Of course, there was that "AFAIK" insurance of yours...
 
Last edited:
Back
Top