I was actually quite amazed how Half-Life 2 could run beautifully even on really crappy PCs. So I can understand why he feels this way. There is no need for multicore processors.

I wouldn't think this will rub off on me, a consumer, when I get a PS3 - is it cost based or something I presume, which would matter for me.

It WILL matter to you if your games are more expensive because the costs to make the games has risen. It will make a difference if the quality of games suffers because the complexity is higher. It will matter if the games become shorter because it takes longer to make games. It will make a difference if the cost of the console goes up to cover the needlessly advanced technology.

Overall, making these systems incredibly complex is not good for anyone except the engineers who develop more complex game consoles.

I'm a game programmer, and I suppose scared is a decent word to describe my attitude towards programming on PS3.

I mean, coders will adapt, and the people who don't adapt will end up doing PSP games, but make no mistake, it is a HUGE shift to running in-game processing tasks as distinct and parallel entities. I can vaguely imagine how I'd code AI or some such for multiple processors, but I can't even concieve of how to lay the groundwork for managing those processors properly.

I mean, this is all very speculative - nobody really knows what's going on with the PS3 or Xbox 360. It certainly is true that these claims of huge power are essentially founed on unproven or undeveloped technology, however.

Actually you're right Eran Hawke, I was just having a stupid moment.

I more of a bargain bin gamer on the 'Station, only the odd GTA/MGS makes me want a game straight away, and on that note let's hope Revolution games stay the same price. Most 1st party N games are bought the second they hit.

Hopefully quality isn't a problem, just depends on cost / developer commitment.

And length, well I don't mind short games providing they have quality and replay value - Yoshi's Touch & Go is a recent testament to that.

He should know about making games that run like crap. HL2 is not really the poster boy of good performance. #1 here I would like to know your specs.

I don't see what the big deal is. AMD and Intel have started to push multicore processing on the PC... so the multicore era is coming, like it or not. Whatever developers learn from a multicore PC environment can likely be applied to the next-gen consoles.

And once more middleware vendors wrap their heads around multicore architecture, that will help Joe Developer make the transition.

The real reason Gabe isn't psyched about programming for next gen consoles is because Source is not a next gen engine! Do any of you realize how absolutely retarded a mutated version of the Quake 2 engine (that's really all Source is) will look in comparison to UE3 (Gears of War, etc). If you're still wondering, it will look pathetic. I'm not even trolling here, it's just the sad truth. I really wanted to like HL2, but Valve have f***ed up way too many times at this point. We'll see how well their Steam platform works on Xbox 360, when every other week people get popups about a retarded level that's been "in-development" for 6 months. Nobody gives a s**t about how you all just figured out HDR lighting, so why would they care about your view on next gen consoles?

This is pretty much the consensus for most game developers regarding multi-core processors. The additional programming to utilize the additional cores is staggering.

What is going to happen is most developers will create half-assed games that only use one core because it's easier that way. The end result is graphics quality that is slightly higher than the Xbox for a lot of future games.

This is why the Rotary engine (for cars) never really took off. It's a great idea and gives you tremendous power and efficiency, but it's so complex and difficult to manufacture that nobody really wants to make cars that use it.

I agree with Peters post on the 25th of July,and if Gabe Newell can't deal with the next gen consoles maby its time for him to retire.It's a whole new world baby,with exciting possibilities!If Edison would have listend to all the negative s**t people were spewing at him we would all be sitting in the dark right now.

With XNA programming on Multiple cores will not be a problem :)

And yes, a lot of new PC's are using multiple cores.. What the Problem?

The Developers are just unhappy because they wont be able to get away with sloppy programming anymore (No more MOH:BackDoorassault games).

I read the article, it was more that he was worried about the costs and difficulty. If anything he seemed positive that we may see some really neat ideas in coding. I think we will just see new ways of doing things and new types of thinking. If there is money to be made, someone will figure it out and be profitable.

I would argue that the Rotary's problem is its lack of gas milage and poor emmisions. Anyway.

Who is going to FUND the development of innovative next-gen titles if they are all huge time hogs and require expensive programmers?

The more effort that goes into making next-gen games; with their higher poly counts, higher resolution textures and multi-threaded programming means only studios with deep pockets will be able to survive. They sure as hell aren't going to be taking any risks.

Instead of new ideas, we are going to see the same old genres. Say hello to countless Halo / GTA ripoffs or drek licensed from movies, comics or movies based on comics.

I don't like where the industry is heading and the majority of shelf space in the game store is already dedicated to games with bigger marketing budgets than they have development budgets.

Gabe is right.

And for all you people who like the needless complexity of these new consoles, I hope you enjoy paying $60-$70 per game.

He sounds like a whiny little bitch. How much 'core' development is actually done on consoles these days anyway? The majority of console games these days are just built around a small set of 3d engines (Renderware anybody?) and a large set of custom 'assets' (models, textures, shaders, sounds, levels). So what's the problem? Some game engine company will do the dirty work for the next-gen consoles and developers go through the usual grind of tweaking the engines to produce their games. Who gives a sh1t if the physics engine runs on a different processor from the AI?

It's only a problem if you're dumb enough not to license a decent engine and try and write your own...

The market is there, the hardware is there, developers will either sink or swim...

The real reason Newell doesn't like the next gen consoles is because Steam doesn't mean crap on a console. Valve has spent a ton of cash setting up this grand infrastructure, and it's completely useless for a vast majority of gamers. I think Gabe would prefer it if PC gaming were dominant, and he could charge subscription fees for Steam without mass defections.

It doesn't help that Valve has never created a 3D engine of their own, either. I wonder which engine they will license then "rewrite" next?

Gabe has a point. The more difficult it is to write a game for an architecture, the fewer games will be written for it.

What this means for the 2006 consoles is that the biggest development houses will still make games for the 360 and the PS3 -- but all the other developers, the ones who can't afford to double the budget it takes to produce every title, will move their focus elsewhere. To the Revolution, perhaps. Or to the handheld market. Or they may even take advantage of the next-gen consoles' reverse compatibility and stick to writing Xbox and PS2 titles.

In general it's a bad idea to reduce the number of developers that are producing games for your console. That's what ultimately killed every console Sega and Atari made, and it's what people say is killing Nintendo now. I hope Sony and Microsoft aren't shooting themselves in the foot with their lust for bleeding-edge tech.

Why are all these devs bitching about multicore CPUs?? I sure didn't hear none of these whining when moving from 2d to 3d. If it's going to take sooo much more effort to code an next gen engine, why don't you just to buy UE3 or something??

As a consumer, why should I give a crap about all these anyway? Increasing game price? That's a BS argument. What did you guys do during SNES days? How about this thing called inflation?? I paid $60US for Street Fighter 2 in 1992, which is $81 in today's money.

All these "don't go there cuz I suck!" talks from devs are pathetically laughable.

since intel and amd already have multicore chips on the x86 platform, perhaps the actual unspoken dread is working with multi-core powerpc based processors in all three of the next-gen consoles. maybe these coders should've tried porting some games over to the macs eons ago so they'd have some experience with the chips, instead of reacting with "oh sit" lamentations today with mere months away from the platform debuts in the marketplace.

heh, the really funny thing is everyone is the bitching of it being harder will be fixed by hiring more programmers. it's been done before. compare the credits of nes games(some as few as like 8 people within my memory), then now you get games that all have 50-70 people working on them. so what if it jumps to 100 to match up the times. or even 150. other companies in other fields have managed. it's evolution. i think the only difference is now there is a tool for voicing everyone's opinion(the internet), as opposed to these people just whining, crying, bitching, and quitting without us hearing about it.

so yes, it will be harder. yes, smaller companies won't have the budget to make console games anymore. oh well. i wouldn't mind if there were 3 releases for my system a month if they were solid games. it's not like i buy more than one a month to begin with.

besides, if anyone wants to cold old-school games, the pc has a market where you can make games that are low-end requirements just fine, and they spread well. look at soldat or fate. if not pc, still got the gba and ds and psp. like someone else said, the ps2 and xbox as well.

fun times lie ahead.

lol #6 this pc is running with a 128mb vid card and 512mb ram. im running it with an amd athlon xp 2600+ with 1.92ghz. and it runs beautifully on here.

first off, nintendo hasn't announced anything about its cpu, it could be multi-core.

second, who "frets" about programming for a controller? honestly?

nintendo's cpu is multicore. (2 cores)

wow. i can't believe all these people who are calling him a whiner. you don't get it - if the engines are 10x harder to program for, much less time gets spent on actually making the game "fun", the games will cost more to make, the games will be more expensive, and it will take years before most developers will fully learn how to write efficient multi-threaded game engines. bitchy developers = crappy games.

i don't know why anyone would say the the hl2 engine (source) is crappy whilst comparing it to ue3. no currently available games use that engine yet. as far as i'm concerned, the source engine is the king of the hill at the moment. disregarding it as a second-rate engine is, well...retarded.

i read somewhere that the xbox 360 will employ some "helper threads" which may actually make development easier. that may be a slight relief. ps3 development sounds like it could be a real nightmare.

btw - i have worked with multiple threads before and chased down deadlocks and obscure seemingly random program crashes caused by syncing issues with the threads. it's feasible to write a complex multi-threaded engine - it's just not a lot of fun.

i don't know what this is all about. much ado about nothing, i am sure.

operating systems exist to hide the hardware from you.

multiprocessor machines have existed for quite a while, now. just create your threads and the os will schedule them using available processors.

hopefully a core will be used to eliminate loading... screens.

dralt, (and alot of others) please don't spew crap that you have no idea about.

games totally bypass oses. direct to hardware.

as usual, the public shows up to try to convince everyone that it is they who know all about the subject at hand - not the freakin expert in the industry.

just typing the words "he doesn't know what he's talking about - programming is easy" doesn't make it true, and so basically its a waste of bandwidth.

here's hoping xna will be something that micro$oft makes successful - and not just vaporware.

#23: Nintendo has also emphasized that developing for the revolution will be very familiar and easy, and not to fret. :-)

#23: Nintendo has also emphasized that developing for the revolution will be very familiar and easy, and not to fret. :-)