backtop


Print E-mail del.icio.us 96 comment(s) - last by artpearson.. on Feb 6 at 2:42 AM

Industry authority and rocket scientist John Carmack shares his views on the latest software and technology

When John Carmack speaks, the industry tends to listen. While it can be argued that his influence today on the gaming industry isn’t as big as it was when nearly every 3D shooter was using one of his Quake engines, he is still regarded as part of the heart of that keeps PC gaming alive. He continues to influence gaming hardware too, especially in the area of graphics. In fact, NVIDIA and ATI consult with John Carmack on design decisions when engineering new GPUs.

Carmack and id Software were recognized last week with two Technology Emmy Awards from the National Academy of Television Arts & Sciences for the areas “pioneering development work in 3D game engines” and “technological leadership in rendering breakthroughs with the Quake technology.”

At CES, Game Informer magazine sat down with John Carmack and Todd Hollenshead of id Software to discuss many facets of the game industry as it applies to both PCs and consoles. Right away, Carmack confirms that he is working on a new engine for a completely new franchise not based on any of the company’s currently existing intellectual properties. Carmack said that Quake Wars, which is based on an upgraded Doom 3 engine, will not be a DX10 game.

On the topic of DX10, Carmack said that there’s nothing at the moment motivating him to move to the new API just yet for Quake Wars, citing that he’s quite satisfied with DX9 and the Xbox 360. “DX9 is really quite a good API level … Microsoft has done a very, very good job of sensibly evolving it at each step--they’re not worried about breaking backwards compatibility--and it’s a pretty clean API,” he said. “I especially like the work I’m doing on the 360, and it’s probably the best graphics API as far as a sensibly designed thing that I’ve worked with.”

Gamers often look to Carmack to tell the fortunes of PC gaming hardware. His opinions on hardware can sway hardcore gamers to purchase one hardware choice over another. Those in awe of the potential offered by DX10 may want to hold off on that shiny graphics card purchase, as Carmack says that there isn’t a huge need for new hardware just yet, as current hardware is more than adequate. “All the high-end video cards right now -- video cards across the board --are great nowadays,” he said. “Personally, I wouldn’t jump at something like DX10 right now. I would let things settle out a little bit and wait until there’s a really strong need for it.”

Those wishing to take the plunge into DX10 will also have to do so while upgrading to Windows Vista. Carmack, however, isn’t all that excited about upgrading to the new OS: “We only have a couple of people running Vista at our company. It’s again, one of those things that there is no strong pull for us to go there. If anything, it’s going to be reluctantly like, ‘Well, a lot of the market is there, so we’ll move to Vista.’”

Carmack then said that he’s quite satisfied with Windows XP, going as far to say that Microsoft is ‘artificially’ forcing gamers to move to Windows Vista for DX10. “Nothing is going to help a new game by going to a new operating system. There were some clear wins going from Windows 95 to Windows XP for games, but there really aren’t any for Vista. They’re artificially doing that by tying DX10 so close it, which is really nothing about the OS ... They’re really grasping at straws for reasons to upgrade the operating system. I suspect I could run XP for a great many more years without having a problem with it,” he said.

Then on to the topic of multi-core gaming systems. Carmack has expressed his dislike for multi-cores, but with the two high-powered new generation consoles both making use of multiple cores, it may be something he just has to deal with. He says of the Xbox 360: “Microsoft has made some pretty nice tools that show you what you can make on the Xbox 360 [with the multi-cores] … but the fundamental problem is that it’s still hard to do. If you want to utilize all of that unused performance, it’s going to become more of a risk to you and bring pain and suffering to the programming side,” he laments. “So we’re dealing with it, but it’s an aspect of the landscape that obviously would have been better if we would have been able to get more gigahertz in a processor core. But life didn’t turn out like that, and we have to just take the best advantage with it.”

As far as the PlayStation 3 goes, Carmack isn’t thrilled at the lack of developer support in comparison to what he’s received from Microsoft. Nevertheless, he plans to support Sony’s console with his next generation engine and games. “We’ve got our PlayStation 3 dev kits, and we’ve got our code compiling on it. I do intend to do a simultaneous release on it. But the honest truth is that Microsoft dev tools are so much better than Sony’s,” he comments. “I think the decision to use an asymmetric CPU by Sony was a wrong one. There are aspects that could make it a winning decision, but they’re not helpful to the developers … It’s not like the PlayStation 3 is a piece of junk or anything. I was not a fan of the PlayStation 2 and the way its architecture was set up. With the PlayStation 3, it’s not even that it’s ugly--they just took a design decision that wasn’t the best from a development standpoint.”

Finally, the console wheel spins to the company from Kyoto, which Carmack says that id Software has never “been that tight with.” He does express his respect of Nintendo’s courage to take a different direction with input methods in controlling games, but his current and next generation of game technology is not targeted at the Wii.



Comments     Threshold


This article is over a month old, voting and posting comments is disabled

Carmack speaks sense...again.
By Viditor on 1/15/2007 8:34:42 PM , Rating: 4
Considering both the cost and most especially the power requirements of DX10 cards (300w each???), Carmack is spot-on when he says to avoid them for awhile...
Besides, there are so many changes happening in the graphics field before DX10 is mainstream (Next-gen Physx, Fusion, etc...) that early adoption seems like a big mistake...




RE: Carmack speaks sense...again.
By cochy on 1/15/2007 8:51:43 PM , Rating: 2
Well unfortunately for those building brand new PCs or upgrading vid cards and looking for the best performance, why wouldn't this person go with a brand new DX10 card when it is clearly where the industry will be moving. Personally I like making purchase that last me at least 3-4 years.

It would have been nice if he were a bit more enthusiastic about multi-core gaming. Is it really that difficult from a programming point of view to design games that take great advantage of multi-core?

Sony should listen to what he has to say and start supporting developers better or all that shiny hardware will go to waste.


RE: Carmack speaks sense...again.
By Viditor on 1/15/2007 9:04:42 PM , Rating: 2
quote:
Is it really that difficult from a programming point of view to design games that take great advantage of multi-core?


In a word, yes...
Writing parallel code is VERY hard and much more time consuming.


RE: Carmack speaks sense...again.
By cochy on 1/15/2007 9:28:10 PM , Rating: 2
Well then I think it's a good time to start developing game engines and programming languages that introduce a layer of abstraction here to make this job easier for the programmer. Soon everyone will have this multi-cores might as well take advantage of them.


RE: Carmack speaks sense...again.
By UsernameX on 1/15/2007 11:23:27 PM , Rating: 2
quote:
In a word, yes... Writing parallel code is VERY hard and much more time consuming.


This is exactly why programmers need to start chipping the ice berg. Multi core processing is MUCH more powerful then our current single core processors today. Once we find an efficient means of utilizing this free resource... the games we play today will pale in comparison to the games of tomorrow.


RE: Carmack speaks sense...again.
By FITCamaro on 1/16/2007 6:57:45 AM , Rating: 2
Exactly. I don't know why a guy who's supposed to be an industry leader would make the comment that he's unhappy with multi-cores. Just because its easier doesn't mean its better. Seems to me like he's only think about making his job easier.

He's spot on with Vista though. Other than DX10 and the enhanced security, there's really nothing pulling people to Vista. Microsoft could have easily made XP have 3D Windows since 3D Desktop already achieves this.


RE: Carmack speaks sense...again.
By masher2 (blog) on 1/16/2007 7:47:04 AM , Rating: 2
> "Exactly. I don't know why a guy who's supposed to be an industry leader would make the comment that he's unhappy with multi-cores..."

Because he's human, and he can't see the forest for the trees. His entire career has been predicated upon parallel processing...the invisible kind, done transparently by the GPU. Now that parallel processing is moving into the CPU, he becomes dogmatic. A bit sad...but very human.

In any case, his opinion certainly won't slow the introduction of highly parallel CPUs. And a couple decades from now, when games are running on hundreds of cores, no one will even remember he said it.


RE: Carmack speaks sense...again.
By EODetroit on 1/24/2007 4:36:46 PM , Rating: 2
You mean just like no one remembers when Einstein said "God doesn't play dice with the universe."? And yeah I think Carmack is just as brilliant as Einstein was, just in different fields. People will certainly remember this.

But I certainly agree with the rest of what you said... Carmack has come up against something that's hard for him, and this time he doesn't want to figure it out himself. In a way though I think Young Carmack in this situation would figure out the "Theory of Everything" for mutiple CPUs... Old Carmack doesn't have to, he's already got his fame and fortune.


RE: Carmack speaks sense...again.
By somata on 1/16/2007 2:43:17 PM , Rating: 2
quote:
Multi core processing is MUCH more powerful then our current single core processors today.


Well obviously that's the case because two processors will always have a potential advantage over one (assuming they're the same speed and architecture). Having said that, I can't believe how passionate some people are about the apparent "magical abilities" of multi-core processors. Yes more cores are nice, but they should be looked at as an addition to increased single-core performance (a la Core 2 Duo) rather than a desired replacement for it. We should all hope that single-core performance does not stagnate with the industry's recent tendency towards multi-core madness.

Aside from being more difficult on developers, Amdahl's Law sets hard limits on how much any application can be parallelized, even the so-called embarassingly parallel ones like rendering. Going from one to two cores can significantly improve most applications; four to eight less so, and so on. In short, a 3GHz P4 is always desirable over two 1.5GHz P4s, so doubling CPU frequency is always preferrable to doubling cores... if you have a choice.


RE: Carmack speaks sense...again.
By lplatypus on 1/16/2007 5:58:54 PM , Rating: 2
quote:
In short, a 3GHz P4 is always desirable over two 1.5GHz P4s, so doubling CPU frequency is always preferrable to doubling cores

Not always... if what you are doing is inherently dual threaded, then the single core CPU has worse throughput because of the overhead of switching between threads. Also, a single core CPU can have worse latency for dealing with incoming events (eg network interrupts) because you're less likely to have an idle core when the event arrives. These are some of the marketing arguments that Sun uses for its Niagara CPUs.


RE: Carmack speaks sense...again.
By therealnickdanger on 1/17/2007 4:16:22 PM , Rating: 2
My 1.6GHz Core Duo rips MP3s and encodes video faster than my old P4 3.0GHz... Granted, it's got faster RAM too, but still, I can't think of anything my P4 does better than my CD.


RE: Carmack speaks sense...again.
By AnnihilatorX on 1/18/2007 2:05:25 PM , Rating: 2
That's more to do with the doomed P4 and NetBurst architecture


By therealnickdanger on 1/25/2007 1:01:15 PM , Rating: 2
Right, just proving that GHz don't matter...


RE: Carmack speaks sense...again.
By mindless1 on 1/25/2007 11:47:22 AM , Rating: 2
It is seldom the case (if ever?) that someone dual threaded has equal but non-dependant threads. You can't just arbitrarily split the workload in half, thus a single 3GHz P4 would virtually always be MUCH faster than two 1.5GHz P4 even at a dual threaded app.


By artpearson on 2/6/2007 2:38:50 AM , Rating: 2
The idea that games are inherently serial is based on an extremely limited concept of what a game is.
The computer game buyer market today is dominated by the pinheads who think that "computer game" means fps (first person shooter) game.
There is really only one such game, in many guises. It consists of roaming around and shooting at things.
The people in this market buy the same game over and over, with different wallpaper.
Gaming taps into our primordial hunter mindset. The thing is that early hunters were smaller, weaker, and slower than their prey.
They won the hunting game by teamwork, with each other, and the hunting dogs that co-evolved with us.
The technology aspect: clubs, spears, bows and arrows was secondary to this.
War came about as a byproduct of team hunting. It was team against team.
The AI required for challenging teamwork on the part of computer player characters is where the need for parallel processing comes in.
The computer science for this is a little beyond its infancy, but is currently neglected. Read some older books like "Communicating Sequential Processes" and "Fairness" and take it from there.


By artpearson on 2/6/2007 2:40:19 AM , Rating: 2
The idea that games are inherently serial is based on an extremely limited concept of what a game is.
The computer game buyer market today is dominated by the pinheads who think that "computer game" means fps (first person shooter) game.
There is really only one such game, in many guises. It consists of roaming around and shooting at things.
The people in this market buy the same game over and over, with different wallpaper.
Gaming taps into our primordial hunter mindset. The thing is that early hunters were smaller, weaker, and slower than their prey.
They won the hunting game by teamwork, with each other, and the hunting dogs that co-evolved with us.
The technology aspect: clubs, spears, bows and arrows was secondary to this.
War came about as a byproduct of team hunting. It was team against team.
The AI required for challenging teamwork on the part of computer player characters is where the need for parallel processing comes in.
The computer science for this is a little beyond its infancy, but is currently neglected. Read some older books like "Communicating Sequential Processes" and "Fairness" and take it from there.


By artpearson on 2/6/2007 2:42:13 AM , Rating: 2
The idea that games are inherently serial is based on an extremely limited concept of what a game is.
The computer game buyer market today is dominated by the pinheads who think that "computer game" means fps (first person shooter) game.
There is really only one such game, in many guises. It consists of roaming around and shooting at things.
The people in this market buy the same game over and over, with different wallpaper.
Gaming taps into our primordial hunter mindset. The thing is that early hunters were smaller, weaker, and slower than their prey.
They won the hunting game by teamwork, with each other, and the hunting dogs that co-evolved with us.
The technology aspect: clubs, spears, bows and arrows was secondary to this.
War came about as a byproduct of team hunting. It was team against team.
The AI required for challenging teamwork on the part of computer player characters is where the need for parallel processing comes in.
The computer science for this is a little beyond its infancy, but is currently neglected. Read some older books like "Communicating Sequential Processes" and "Fairness" and take it from there.