Carmack talks next gen consoles... and beyond Comments by Richard Leadbetter

Many interesting points! If he is right...

* We can forget MS or Sony doing a "Wii". To me it's great news, wouldn't like PS4 or X720 to be an overclock of old hardware.
* No more optical drives. The good: faster access, less power consumption, fewer breakdowns, no more Sony trying to win home-threater battles by means of it's users. The bad: say goodbye to second-hand and import market.
* Never thought of Sony trying to be the first to launch the next gen, always thought MS would try the same strategy. They will likety launch very close to each other. In any case we should be hearing about the next-gen by 2011 the latest.
* So the future is similar to Intel's Larrabee architecture, with is described as multiple programmable small cores. Will that be an advantage for PS3 programmers, used to fighting against the bunch of SPEs to get performance? Time will tell.

It's always interesting to read from JC.

I don't see Sony jumping the gun here, I mean I really don't. I think he's being a bit mischievous.

While Sony are probably desperate to replace the RSX more than anything, they can't just do it on a whim.

And I don't see a need to....it seems to me that judging by impending PS3 titles that the PS3 is finally going to be coming into its own and differentialting itself from the 360; Uncharted 2 looks simply spellbinding, and God Of War 3 looked pretty good as well considering it's still 8 months away.

If the technology in the PS3 is only just beginning to be realised now, why would Sony want to jump the gun on a technology that they spent so much money investing?

Sony would appear to have the clearer roadmapped future given the CELL used in the PS3 is first iteration part of what is likely to be an ongoing line of processors. Jumping to a 16 or 32 core variant is just a matter of yield rates, and IBM has already done a lot of work with their newer chips of improving PPU performance.

The RSX is a stock part, so upgrading to a newer NVidia GPU shouldn't be an issue, as would increasing the amount of system RAM.

Its unclear what MS are likely to do, the Xenon is a good chip but its uncertain how much R&D is going to be involved to shift to a greater number of symmetrical cores - clock speed is unlikely to be improved due to heat emission issues.

GPU performance can be improved, but again the Xenos GPU is significantly different from standard parts due to its embedded eDRAM frame buffer.

theoretically if everyone goes optic-drive less, they can sell the games for cheaper since they don't have to manufacture disks or boxes or pay stores for stocking space.

though of course this depends on the market: if EA puts out a digitally distributed 60$ motherfucker, and a good 100,000 people buy it on day one, you better believe everyone will keep charging 60$. everything depends on what John Consumer does.

The only certainty next gen is a large bump to the internal ram. Carmack probably called it correctly. 2 gig minimum.

@Greybeard
Xenon and Cell share the same fundamental PPE main processor, derived from the PowerPC. Larrabee doesn't have a main processor. Memory access is also much more like a traditional CPU than Cell which should make it easier to program, so arguably programming the Xenon is closer to programming for Larrabee, even though the Cell might be closer in hardware terms.

"It's always interesting to read from JC."

Yep, love these articles . . .

I don't see Sony replacing the PS3 anytime soon, but I don't see Microsoft doing that either.. I can definitely see them releasing a more durable slim line model, with more of a focus on Natal, but not a whole new console.

If i were a betting man, i'd say the chances of a each company releasing their next macine first are:

10:1 - sony (theyve lost too much on the ps3 to replace it any time soon)
5:1 - microsoft (new machine with natal built in)
evens - nintendo (wii plus - hidef wii with motionplus built in)

My predictions

The next new console will be from Nintendo;
Sony's next console will be download only;
MS's next console's games will be available on disc and via download;
Cloud computing won't take off;
All these crazy graphics techniques will only take off once some one writes a decent coding interface for them (probably MS with a new DirectX, expensive engines from id and Epic will follow);
Proper ray tracing won't be in the next gen consoles but the ones after.

I'm not sure about sony. I'm wondering if they'll do a saturn here, and just give up on the PS3 as a failure. People might say they'll make a huge loss, but they're not exactly setting the world on fire with the PS3 at the moment.

With Sony, i think it could go either way. It has a strong case for both staying and ditching.

MS is the more likely candidate though.

i reckon for the next 5 to 10 years we wil see more digitalize games and movies ( netflix,Sony pictures and even Nintendo doing some streaming movies or shows), cloud computing wil take some time on the console market, while for pc`s it might be within a year or three to get some serious pull in the gamingworld.

I predict we'll all have robot monkeys who will take over the world while we're watching Transformers 5: Megatron Rides Again.

Talk of cloud computing is premature, the infrastructure needed to support it for real-time applications is a long way off even in more developed countries.
Even disc-less consoles next generation are unlikely in my view, we'll probably get a gradual shift with simultaneous disc/digital releases for a while.

ray tracing my ass!!!! where is 1080p in this gen? aren't we in the next gen already ... lying dicks. 2GB of ram... for crying out loud! this shit doesn't cost anything anymore. especially if you are selling 20 million consoles a year or more. standardize the fucking hardware and just create a beast of a console and wrap that up in a small watertank or fridge. Sell it for £500.

or... continue as usual. spend the least amount of money possible and still make a loss for some bizarre reason.

@ GreyBeard

"The RSX is a stock part, so upgrading to a newer NVidia GPU shouldn't be an issue, as would increasing the amount of system RAM."

Another scenario is going back to the original PS3 plans and don't include a dedicated GPU. With GPUs shifting more towards CPUs anyway that might not be a bad decision.

"10:1 - sony (theyve lost too much on the ps3 to replace it any time soon) "

That's only an issue insofar they don't have the resources to actually fund the development and production of a replacement. If they believe a PS4 would have more success (in the sense of generating more cash for them) in the current or short term future market they will release it as fast as they can.

My predictions:

* PS4 will not have Cell
* MS will have an all digital download option and will have save games online.
* MS will launch after Sony
* Next gen consoles will vary wildly in processing power once again.
* All of them will have motion options
* Halo 4 will be there day one.
* None of them will have cloud.

*I too am glad that Xbox 720 (I will call it XB3) and PS4 will most likely be a full generational leap from 360/PS3, and not merely a modest upgrade of the current 360/PS3 architectures.

*I expect XB3/PS4 to launch sometime between fall 2011 and fall 2013. Whether Sony actually launches first is another matter. Good chance that they will, given the fact that they are in last place this gen.

* I expect Intel's Larrabee GPU / GPGPU to be the main part of at least one of the next-gen consoles.

*The graphics engines/GPUs of next-gen consoles will probably be a GPU-generation *beyond* the upcoming DX11 GPUs from AMD (R8xx) and Nvidia (GT300).

*I do not believe in so-called 'diminishing returns' as far as graphics. Real-time graphics have several entire console generations yet to go (5-6 years each) and many GPU-generations to go (2-3 years each) before graphics are "good enough", and before developers can't go any further. Next-gen console graphics should be more 'CGI-like' than current-gen, although not on par with high-end CGI used in high-budget movies.

*Hopefully next-gen XB3/PS4 graphics will be on par with typical PS2/Xbox1 prerendered game intros & cutscenes. That would be a NICE leap from current HD graphics that 360/PS3 pump out. It won't be about increasing resolution, 720p and 1080p will be used for XB3/PS4 games, but with dramatically increased complexity (more triangles), and image quality and post processing effects.

* I also never believed in Sony's 'ten year lifecycle' for PS3. They never said PS3 would be their only console for 10 years, until 2016.

*By the time 360/PS3 are ten years old in 2015-2016, the next gen XB3/PS4 consoles will have been out for several years and the next-next gen (XB4,PS5) will be well into R&D.

*hopefully the PS4 GPU (assuming its another Nvidia part and not Larrabee) will be far more impressive for its time than RSX. The RSX was a downgraded version of the NV47 / G70/G71 / GeForce 7800/7900. The PC versions had 16 pixel rendering pipelines/ROPs (not to be confused with the 24 pixel shader pipes), 6800 Mpixels/sec fillrate, 256-Bit external memory bus interface, 54 GB/sec bandwidth.
RSX had 8 pixel rendering pipelines/ROPs, 4000 Mpixels/sec fillrate, 128-Bit bus, 22.4 GB/sec bandwidth, and PS3 launched in late 2006. That's a huge downgrade from the first PC version which launched in mid 2005! The original Xbox1 GPU (NV2A) was much more impressive for its time, in fall 2001. It was comparable to the highest-end PC GPUs of the day (GeForce 3 Ti 500, Radeon 8500) and actually NV2A was better because it had 2 Vertex Shaders, the PC GPUs only had 1 VS. The RSX was so much weaker than the highest-end PC GPUs (GeForce 8800). So for such a high-end, high-end console, the PS3 was really lacking in graphics department, and no, the CELL did not and does not really make up for it.

PS4 will need an impressive GPU if developers are to get excited about making exclusive games for Sony again.

"for such a high-end, high-end console, the PS3 was really lacking in graphics department, and no, the CELL did not and does not really make up for it."

Uncharted and Killzone 2 disagree. True, it's more difficult to unlock but not impossible.

"i do not think next gen will be a big an increase in power as the jump from ps2 to ps3 etc..costs have to taken into account as its all about the money at the end of the day"

Exactly. Till Wii, common knowledge (which is a pretty accurate predictor for falsehood) said a console should consist of cutting edge technology and initially be sold at a loss if it was to have any chance in the market.

I think for Sony, PS3 will turn out to be a blessing is disguise (as not being able to keep up the tech arms race turned out to be for Nintendo) in that it hurt them a lot but not enough to turn them into the next Sega.

So the only one capable to release a next gen machine in the antiquated sense is MS but with Nintendo and Sony very likely following a different route and the PC platform there as the alternative for gamers that are actually willing to pay for what they get, there won't be much pressure for them to hurry.

how about a rampak for both for £100? 2011 for a new xbox or ps? too bloody soon!

My only prediction is that MS and Sony will come up with wildly different architectures while graphically being pretty much the same.
Much like how it is this current gen.

I really think the Wii has proven that there's a bigger barrier to better graphics than technology and that is the customer. The customer just doesn't care enough about all these technical things and considers the Wii pretty enough. Maybe in some years customers will demand more than Wii graphics and by then can probably be satisfied with a 360-level console which would be dirt cheap to make by then. There's just no point in making stuff vastly better than the customer needs, especially if that ends up raising the price.

Carmack should look at the business situation. The Wii has a majority share of the market and his company is NOT there to supply customers (whether they are gamers or companies) with products. That's a massive missed opportunity and he already recognized that years ago but they're not bothering to adjust themselves to the reality of the market. Considering how little id released lately and how unsuccessful their last products were (including their engines, the Doom 3 engine was used by maybe 1-2 games outside of their own ones while the Quake 3 engine was everywhere) I wonder how long this company can hold out before hitting the brick wall of bankrupcy.

"but they're not bothering to adjust themselves to the reality of the market"

To be fair, that goes for most developers and the publishers that fund them. It's hard to get rid of superseded dogmas. The current breed of developer is addicted to ever increasing numbers of pixels, stimulated in this behaviour by publishers who thought that graphical fidelity was a sure-fire test for market potential.

>where is 1080p in this gen?

Using up too much memory and frame rate to be of any use to anyone?

@KDR_11k : Agree

Carmack is going on the assumption that better graphics automatically means better games.

Seems to me most consumers dont care - and just want FUN games.

Uncharted and Killzone 2 disagree. True, it's more difficult to unlock but not impossible.

Well, as you say, it's more difficult. Very difficult, as it turns out.

Both Uncharted and Killzone 2 cost squillions to develop, using custom engines designed specifically for the PS3, and ONLY the PS3. Each game is tailored for the abilities and deficiencies of the PS3, confining themselves to those areas that the PS3 can handle well. (You will notice that neither game features massive amounts of alpha overdraw.)

Not coincidentally, both of these companies are owned outright by SONY...

Third-party PS3 developers -- which is most of them -- cannot do this. They don't have $50m being provided by the mothership. So they have to use and extend existing engines, and they're never going to get the same level of performance out of the hardware that an in-house developer will.

@ smelly "Seems to me most consumers dont care - and just want FUN games."

This is more a question of Nintendo tapping the entire mass-market, which has actually enabled MS to step in and realise their dreams of putting their digital media center under the TV. Not yet - but over the next few years they will totally gear up to take over every single living room much like PS2.

Really what we've seen this gen is that games for games sake are going down the pan, and everything is shifting towards the casual pick up and play market. I think this bubble is going to burst, but I don't know how long it will take.

Also does anyone think that PC will continue to be the platform that pushes tech onwards as it (just about) still was this-gen (the consoles launched with parity or better but the PC market soon made up the difference).

Isn't it more likely that consoles will become the tech breeding ground and gaming on PC will slim down along with the rate of the technological growth. If there isn't mass gaming being done on PC then there's no reason for PCs to get any faster than they are now.

You basically might see PCs becoming niche 'consoles' once again, which if it happens I hope will lead to a rennaisance for PC games.

"But then theres a lot more than just graphics that a decent & powerful chip can do."

True. But TBH, I don't think that developers have the skills to properly use that processing power for other means than pushing more pixels.

Whilst it's possible Sony and MS are going to tone down the leap in performance between this and the next gen, we shouldn't forget the "prestige" factor, running specs up: neither company really don't want their competitor to be vastly superior, especially not in the eyes of the "core" market (which would be foolish to ignore).

Both will (should) probably have two controllers/devices right out of the box, though: the "classic" one and some sort of "waggle" one.

It's nice to see a developer talking candidly about technical things rather than mud slinging other devs or claiming how much better they are than anyone else.

Good read.

carmack has always been interresting, I liked reading that.

Though I'm concerned about game's lenght. I mean they already spend 1 or2 years on a product, which you just end in a day.
or two. IE : uncharted. Sure you can replay it, but it's damn short, and take a huge amount of time to make. What will happen if graphics are going like real time, ray tracing and all.

3/4 years of production and 2 hours of gameplay? Just like the 3D movies? I think they should alreayd devellop awesome tools if such technologies come on the market.

Carmack is going that way, trying to unify everything under one engine, I think it's the good way.




Oh and on a personnal note, please stop stupid ports on PC in the futur, when you'll have more the 512 ram :D

"Whilst it's possible Sony and MS are going to tone down the leap in performance between this and the next gen, we shouldn't forget the "prestige" factor, running specs up"

True. Idiocy like that has run plenty of companies into the ground. That's why I said it might be a blessing in disguise for Sony if they're hurt so much by the failed PS3 that they just don't have the money to develop and manufacture a super powerful PS4. Because if they have the chance, there's little doubt that would be the road they would pursue and it would yet again lead to a huge financial disaster.

"especially not in the eyes of the "core" market (which would be foolish to ignore)."

No, it would be very smart to ignore consumers that don't make you money.

No, it would be very smart to ignore consumers that don't make you money.

I think MS is doing pretty well this gen, specificaly addressing the so-called "core" market, and would be doing even better if they didn't fuck up with the whole RROD debacle. The Big three should aim for the "sweet spot" and make happy/lure in both "sides" of the market. Easier said then done, of course.

I'd love it if there was a consortium of MS, Sony & Nintendo and say the 10 biggest publishers who got together and standardised a console spec in much the same way DVD was designed. Any manufacturer who wishes to make a console has to pay a small fee to the consortium and comply with the spec. This would mean you could get a variety of consoles which all have the same game-playing capability, and we get a nice price due to competition. The other advantage would be that you could get a basic model a few years on from ALDI whichwas cheap as hell and looked like arse and still played all the games just as well as the high end Sony version which could have a built in 7.1 system and high-end DVD upscaling etc. A few years down the line they would be able to build it into TVs as an option. It would possibly lower the entry barrier to developing games due to standardisation, and lower overall costs due to having one platform only to develop for. It would be like a modern take on the 3DO.

It wouldn't be perfect but it would be interesting.

Ray tracing and Ray Casting are the "designer's dream" (what a designer would love to have) but I doubt that they will be used, the cost / benefit is just not there. :/

colostomyexplosion: You mean like the 3DO? Thanks but no thanks, that experiment ended in a failure years ago. Besides, variety is the spice of all things good.

"I think MS is doing pretty well this gen"

MS is rather successful in marketing the 360 as a success to core gamers and the gaming press despite the fact that it pretty much failed to achieve sales targets from the moment it was released. Their luck is that the PS3 fares even worse (compared to expectations). But 360 isn't making MS money. The initial R&D costs were too high to recoup with how their installed base has developed so far.

I agree that it would be great if a platform holder was able to create a machine with appeal to both core and casual gamers. But given the obsession of the former with graphics coupled with the unwillingness to pay for it, that seems like an economic impossibility.

"I'd love it if there was a consortium of MS, Sony & Nintendo and say the 10 biggest publishers who got together and standardised a console spec in much the same way DVD was designed."

That too would mean an end to bleeding edge console technology sold at a loss.

@Les

I agree it would end the sold at a loss era, but like DVD and Blu-Ray players, you would be able to get them at the £400-500 mark early, and cheaper models would come along pretty quickly afterwards. I'd expect the prices to fall quicker than they have done for the PS3 or Wii (think how much a standalone Blu-ray player was 2 years ago, now they sell them at ALDI for less than 100 quid).

There would still be competition as there is with DVD players, for example a cheap DVD player will still play all your DVDs but a more expensive one may have a faster seek time, whilst still conforming to the minimum spec. This could happen with a standardised console, with premium systems offering faster RAM, flash memory than the base model, in addition to built in accessories which do not necessarily form part of the spec (such as a display panel, surround system etc).

A couple of years in and you could get a cheaper basic model for a bedroom or it could be built into the back or car seats as an option etc.

The main issue I see would be with the system software and online infrastructure, which would have little in the way of competition to drive the improvements we have seen to XBL and PSN.

@colostomyexplosion

it's called a PC

"I agree it would end the sold at a loss era, but like DVD and Blu-Ray players, you would be able to get them at the £400-500 mark early"

Maybe, but at such a price point it would have to consist primarily of stock parts as a console is a more complex device than a media player (e.g. it requires a much more sophisticated CPU).

IIRC, PS3's manufacturing costs at launch were more in the 800 - 900 GBP range. With a more limited market that doesn't care that much for fancy graphics in the first place, it would be a hard / impossible sell even at cost price, let alone with a profit margin on top. Especially not with the PC as an alternative and with an audience that is actually willing to pay for what they get.

As a leyman with no technical knowledge my unfounded predications are:

Download distribution will rely heavily on the success of the PSPGO i personally don't feel that our infrastructure is yet or will be in the forseeable future capable to cope with large dowloadss therefore i believe that it will be supported, but not the sole means of buying games.

Ninty won't release a new console anytime soon, bar a redesign and minor upgrades. The wii is selling like hotcakes desite it being in terms of hardware the weakest of the bunch.

MS will put too much focus on natal and it'll backfire, causing them to chance stratergies mid generation and try to claw back this generations success. The MS console will be the first of the bunch to launch.

My Guesstimation

"PS4 will need an impressive GPU if developers are to get excited about making exclusive games for Sony again. "

I don't understand this. The PS2 had a pretty crappy graphics card compared to the Xbox, but it didn't stop studios making exclusive games for the PS2, because even though the PS2 architecture was proving a headache for developers, they saw a reason to fight through it because they saw that the PS2 was clearly going on to dominate.

If devs saw that the PS3 was going on to dominate, they'd have seen a reason to fight through the difficult development cycle, and may have made games exclusive for Sony again...although this isn't really a given considering the cost of games these days, and the precarious position of some studios even after putting all their games multiplatform.

But in any case, the PS3 isn't dominating and will never dominate outright, which is certainly a contributing factor in why Sony has lost exclusives.

It's nothing to do with the graphics card.

Please be wrong. A lot of developers/publishers are struggling to make money as it is

I'm not sure the term 'next-gen' will be appropriate, as it seems that incremental model improvements (and software refreshes) are providing enough interest and revenue to keep them all happy. I wouldn't be surprised if we don't see a completely new platform for years.

On Sony, I don't see them believeing their own Cell hype again and dropping Nvidia, no matter how improved the Cell is or will be. Besides, given how their fortunes have changed, Nvidia will probably be very accomodating in supplying new parts.

"I wouldn't be surprised if we don't see a completely new platform for years."

I personally would be happy to see that scenario play out. But in the current market, there is one winner (Nintendo) and two losers. If they think that launching a new platform will make them more (or lose them less) money than the current iteration they will launch it sooner rather than later.

"No more optical drives. The good: faster access, less power consumption, fewer breakdowns, no more Sony trying to win home-threater battles by means of it's users."

Also save the environment by not having to distribute physical media; disc, box and all.

Although that's just me.