What if Atari sold consoles again?

[Kalvan]

Tilera Tile G series ... roughly eight times as powerful as a PS3 that would be twice as easy to program for (They already have Linux for Tile architecture,for example, which uses a slightly modified version of the MIPS instruction set)

 

Hm, the Tilera Gx isn't a game console chip. It only clears 50 gigaflops theoretical peak, around 25% of the peak performance of the Cell in the PS3, or about 2% of what a modern GPU can do. It's also not entering mass production for 12-18 months. By 2011, compared to a next-gen GPU, it will be under 1% of the performance and probably no easier to program for standard "game logic" tasks.

 

The Tilera is definitely good for the markets it aims at -- massively multithreaded server processes. But it's no good at the kind of AI/physics calculations needed by modern games. GPUs will keep getting better at multi-core game logic acceleration, but it will be a while before you can ditch a traditional CPU.

 

- KS

 

Which is why I'm using it as the CPU! I'd be using a Radeon HD 5850 (or maby two 6770s if the pricing of Hecatonchires stays reasonably sane) for the GPU.

[kskunk]

Which is why I'm using it as the CPU! I'd be using a Radeon HD 5850 (or maby two 6770s if the pricing of Hecatonchires stays reasonably sane) for the GPU.

I was just pointing out that the Tilera Gx is inadequate for modern games, due to the lack of floating point oomph. Number of cores doesn't matter when a single Cell processor has 4 times the raw power of all 100 Tilera cores put together.

 

Game logic heavily relies on floating point processing. Integer logic is a small detail. Tilera excels at integer workloads which is important in webservers and databases. That's what the chip is for, so they neglected features that would help gaming.

 

The only reason I compared the Tilera to a GPU was to point out that the <i>style</i> of code that fits well on a Tilera will fit even <i>better</i> on a next-gen GPU. They are as easy to program as the Tilera in the respects that matter for game inner loops. This is an exciting development because it means more game logic (such as physics or possibly even AI) will move to more powerful multi-core GPUs.

 

It may be that future game consoles have only a 2-4x faster CPU, relying more heavily on GPU to make up the 10x performance difference we all expect.

 

Some other ideas for your hypothetical machine: ARM is investing heavily in low-power multicore technologies that are much more compatible with the needs of game developers -- see Cortex A9 (or maybe A10 depending on the year you're thinking). ARM NEON floating point acceleration is also much more suitable for game logic than a MIPS single-issue FPU.

 

Still, for a multi-core CPU, PowerPC may be hard to beat for a modern console. Except possibly if AMD decided to supply a stripped-down multicore Bobcat?

 

Sorry for going off-topic. I'm really interested in console design obviously.

 

- KS

Edited January 19, 2010 by kskunk

[carmel_andrews]

Only sort of games system Atari could be interested in doing is some sort of re-badged wii/xbox or ps3 type scenario but with some built in h/w mode for runnig old school atari games (i.e 26/52/78 and lynx)

 

That is ofcourse if you can convince the clowns at infoatari to invest in using atari's name on hardware again (which as we all know won't happen with the present owners)

[atari181]

At This Point Atari is not much more than a name. It is owned by infogames. Atari did not release the popular Plug N Play games a few years ago, A toy company called Jakks Pacific did. They DID however release the Flashback and Flashback 2. However Atari has nowhere near the capital to release any real new gaming system. In 2008 they were to be pulled from the NASDAQ and struggled to show that the company had assets of 15 million to stay listed.

Any new system today would cost considerably more, as Microsoft saw with around 2 Billion in losses from the introduction of the first Xbox in the first year. Unfortunately this will never see the light of day. Wish it could/would but not anytime soon.

[+wood_jl]

At This Point Atari is not much more than a name. It is owned by infogames. Atari did not release the popular Plug N Play games a few years ago, A toy company called Jakks Pacific did. They DID however release the Flashback and Flashback 2. However Atari has nowhere near the capital to release any real new gaming system. In 2008 they were to be pulled from the NASDAQ and struggled to show that the company had assets of 15 million to stay listed.

Any new system today would cost considerably more, as Microsoft saw with around 2 Billion in losses from the introduction of the first Xbox in the first year. Unfortunately this will never see the light of day. Wish it could/would but not anytime soon.

 

It was not $2 billion, it was around $4 billion in losses, according to Forbes. That's why I used that figure earlier in the thread.

 

http://www.joystiq.com/2005/09/26/forbes-xbox-lost-microsoft-4-billion-and-counting/

[atari181]

At This Point Atari is not much more than a name. It is owned by infogames. Atari did not release the popular Plug N Play games a few years ago, A toy company called Jakks Pacific did. They DID however release the Flashback and Flashback 2. However Atari has nowhere near the capital to release any real new gaming system. In 2008 they were to be pulled from the NASDAQ and struggled to show that the company had assets of 15 million to stay listed.

Any new system today would cost considerably more, as Microsoft saw with around 2 Billion in losses from the introduction of the first Xbox in the first year. Unfortunately this will never see the light of day. Wish it could/would but not anytime soon.

 

It was not $2 billion, it was around $4 billion in losses, according to Forbes. That's why I used that figure earlier in the thread.

 

http://www.joystiq.com/2005/09/26/forbes-xbox-lost-microsoft-4-billion-and-counting/

 

Yes they did loose 4 billion, over the life cycle of the Xbox, NOT the first year. The first year as I stated they lost around 2 billion.

[oky2000]

Little Johhny wants a PS3/360/Wii to play games on AND NOTHING ELSE, companies with the kind of financial strength to do a project like you are proposing know this and small companies could never build it without going bankrupt so ipso facto...nice dream

 

To misquote Men in Black:

In 1986 everybody knew that console video games were dead for good. In 1993 everybody knew that 3D0, Phillips CDi, and the Commodore Immagination Machine were the wave of the future and that Microsoft Bob would make conventional GUIs obsolete. In 1995, everybody knew that the Playstation would have way too slow loading times and the Virtual Boy would be a smash hit. In 2007, everbody knew that Pandora would kick the Nintendo DS and Playstation Portable to the curb. Who knows what people might know tomorrow?

 

Right now, If I could hack a new BIOS and fit a Tilera Tile G series (and if newegg.com stocked it and the Lucid Hydra chip by itself) to the slot of a major PC board maker, I could put together something roughly eight times as powerful as a PS3 that would be twice as easy to program for (They already have Linux for Tile architecture,for example, which uses a slightly modified version of the MIPS instruction set), have eight times the memory, and be far more quiet and reliable. All for about $800-900. Now if that sounds steep, remember that this is from off the shelf retail parts. Many of thses things have incredbly high margins that could be reduced through negotiation and bulk sales (I know for a fact that Radeons are selling with huge margins now that TSMC 40 vm yields are up to 80%, and that would be the single most expensive part). Okay, to truly enjoy it would in theory require a 22" computer monitor, but still, it's the principle of the matter.

 

Erm, where is your hugely powerful 3D GPU in that factoring.

 

The simple fact is, as much as I hate to admit it, EVERY company who wants to build a console at least as powerful as an xbox 360 graphically will be paying MORE for the components for that badly soldered together pile of crap. Look how the PS3 sales struggled for years due to the price point. Microsoft are now probably selling a 360 without hard drive for about $150-200 for a complete 1080p 3D lush gaming console.

 

Can you compete with that? Without the branding of the big three you need to do it better than them and for less.

 

If SEGA have no way back into the console market you can bet your ass nobody else will pull off gaming at this level of price/performance.

[Kernal]

so say i was to make a proposal to the atari corporation as it stands today..

i was thinking they could get with the open source crew of today..

make something that is competitive with the xbox...

 

prehaps make a gaming system that plays all the games of ALL systems (of course that would never happen..

at least make the new console backwards compatible where it will run all old atari software and NEW atari software.. prehaps thru emulation etc..

 

thats why im leaning towards and open source like linux based console...

 

licence free emulation... etc.. plus net working cabapabilites, but also include legacy joystick and xio ports so u can also play your original games.. mebbe a cart port too (in 2010!!@!)

 

prehaps we could have a tempest 2010..

or a 3d journey to the planets etc.. etc..

[mckafka99]

I think my dream retro system from Atari would be a 'vectrex like' machine (same or at least similar form factor) only in color which would include (or have the ability to play via some memory card) their classic coin-op games. I know this couldnt be a <$100 system, but I wonder if something reasonable couldnt be put together.

 

As I was using the Vectrex the other night I was thinking how awesome it would be to fire up a classic Atari coin-ops on such a device. Not sure what it is but there is something about the Vectrex, walking up to it, turning it on and having gaming available immediately all in one compact unit - no tv connections or worrying about using the tv when someone else may want to use it, no swapping carts (I have a VecFlash), the unit is fairly portable so I can move it to any room I want, no long waits for an OS to boot up (like emulation on a PC).

 

Time to stop dreaming now.....

[carmel_andrews]

I was reading some website that had a brief synopsys of atari history, it mentioned that during the 'brief' JTS period, JTS offloaded (sold) some of atari's patents and IP's to generate some income

 

I'd be curious to knowing what patents were sold and how much did JTS/Atari get for them and did the sale of those patents stop atari from reproducing hardware based on those patents (assuming that atari wanted to return to hardware manufacturing)

 

Apparently just before Atari got into bed with JTS, Atari were in talks with amiga owner Gateway over atari giving gateway (for a price) some of atari's IP's and patents etc (sort of glad it didn't happen, after all, look what happened to the amiga, gateway turned it form a hardware platform into little more then an o/s platform)

[doctor_x]

i love that this is still getting replies.. great thread!

[falcon_]

 

I would definitely support a new Atari console, if what I read about the specs would make me want to support it. In fact, I have an idea for the layout of such a new systemm. It would involve a Tilera 64-100 core CPU, ...

 

Unfortunately, I do not trust Infogames to deliver this piece of hardware at acceptable levels of quality control(And I will never refer to them as Atari SA-Oh no, I already did ). They seem content to milk the name for all its worth without any risk or cost to themselves and their shareholders. ...

 

But assuming a screwdriver for a moment that my concept came to fruition, those writers are less of a threat than you think. The honorable ones who poopoo it before having system and game in hand will publicly apologise, signing their names to the articles before going home to commit unassisted seppuku .....

 

I love a good rant!

 

_falcon_

[Lynxpro]

Personally I would settle for getting any big American company to come back into the console wars with a Silicon Valley, CA development team comprised of bright individuals from past companies such as Atari, Sega, 3D0 and present industry software and hardware companies. I’d ask them to develop an American console to compete with the big 2 Japanese consoles and maybe shift a little of the home console energy and enthusiasm away from Japan and back to the Europe and the Americas where it's been cornered for so long.

Microsoft is a big American company. Nobody is going to build high-volume, mass-market electronics in Silicon Valley anymore. Hell, Atari itself was an American company that moved operations to the Far East in what. 1982? If Atari couldn't continue to build in America then, and Microsoft can't build in America now, by what magic do you propose to make it profitable to do so in the future? How are companies going to pay $20/hr plus benefits in America when they can get the same labor for $0.60/hr in China, with the additional savings of no insurance, no unions, no lawsuits, and no EPA (just dump the toxic waste in a gully)?

 

 

The US need to get serious about bringing manufacturing back to our country [and the same goes for the EU nations]. It is strategically important to have a manufacturing base, ergo, it is a national security concern. As a government employee, I find it totally unacceptable that we buy computers from Lenovo which is owned by the Chinese People's Liberation Army. First of all, their quality level is worse than Dell's [or HP's], and second, I don't like the idea of funding a competing nation's military that we could have to face off against in the next 20 years over Taiwan. This is very similar to what happened to Britain prior to WWI. Capital left the UK and funded German manufacturing. The Germans then pumped their goods into the free market of the UK and the Germans took all the sterling they gained to build up their military power. The eventual war bankrupted the British Empire.

 

Microsoft won't build in the US because it is more profitable to do so overseas. The same with Apple. Yet both companies are sitting on $30 billion+. It irks me that both of these companies are so big and Atari ended up the way it did. If anything, it should be the other way 'round.

[+wood_jl]

The US need to get serious about bringing manufacturing back to our country [and the same goes for the EU nations]. It is strategically important to have a manufacturing base, ergo, it is a national security concern. As a government employee, I find it totally unacceptable that we buy computers from Lenovo which is owned by the Chinese People's Liberation Army. First of all, their quality level is worse than Dell's [or HP's], and second, I don't like the idea of funding a competing nation's military that we could have to face off against in the next 20 years over Taiwan.

 

Absolutely. In essence this country (and many others) are now wholeheartedly dependent on Chinese manufacturing. If war were to break out (heaven forbid) and relations/trade severed, it's kind of scary to think what a predicament the U.S. would be in. Aside from Taiwan, Chinese interests are being asserted globally. (For example, I do believe they operate the Panama Canal now).

 

Large corporations are all for Chinese manufacturing and the fat margins it provides. Those entities which benefit from it maintain a nice campaign by extolling the virtues of "free trade" because the word "free" has a nice ring to it. Yeah, "free trade" sounds nice, but "fair trade" sounds even better; it's too bad it doesn't exist, and you never hear calls for it. Any such calls would immediately be demonized as a threat to freedom. Meanwhile the Chinese government manipulates their currency to keep it cheap so their products are. You really have to hand it to the Chinese - they are extremely intelligent and resourceful.

[Lynxpro]

Absolutely. In essence this country (and many others) are now wholeheartedly dependent on Chinese manufacturing. If war were to break out (heaven forbid) and relations/trade severed, it's kind of scary to think what a predicament the U.S. would be in. Aside from Taiwan, Chinese interests are being asserted globally. (For example, I do believe they operate the Panama Canal now).

 

 

I would take the stimulus funds and invest them in manufacturing here in the States. Take about $100 billion of it and assign DARPA to administer it into going into companies that would build fab plants here in the States. We need to make the chips here. Same goes with memory chips and flash. We have more than enough fabless design companies but not the manufacturing to go along with it unless one counts Intel. The problem with counting Intel is eventually their designs are going to hit the wall and then the ARM derivatives will take over and most of them are built in Korea [until the Chinese catch up in that category too and surpass the Koreans]. Look at Apple. They are now creating their own ARM chips - thanks to their acquisition of PA Semi - but the chips themselves are being built in Korea by Samsung. That is still unacceptable.

 

With all the talk about fascism/corporatism going on with the Federal Government owning a good chunk of GM, I'm not seeing the Feds actively managing the company. Their management team are paying themselves bonuses while rejecting offers for selling off Saturn and Hummer. I don't understand the logic behind taking a smaller monied offer versus shutting down the entire operations. $1 is better than none. They also need to be forced into selling off OnStar. That needs to be owned by a mobile phone company and/or some company like DirecTV, not GM. Hummer could've been turned around had GM not been idiots and given the vehicles GM diesel engines. That's a very popular mod for Hummers. Even Arnold Schwarzenegger has one of his Hummers retrofitted with a diesel engine that runs on bio diesel. Saturn should be a joint venture between GM and VW. Maybe GM could get a license to TDI from VW for it.

 

Bringing back production to the US could help revive the possibility of someone going out and creating a modern Atari here in the US again. An upstart that could challenge not only foreign competition but the big American multinationals as well. That is what the Feds should be doing instead of trying to prop up Intel and Microsoft.

 

 

 

 

I believe in the concept of free trade between equals. For example, the US and the EU/EFTA/EEA should have free trade for everything but agriculture [since that's always the sticking point]. We should not have "free" trade with China. "Fair" trade is a term that has been hijacked by the various lefties who don't bathe and will scream at you for buying coffee at Starbucks versus the local mom & pop coffee shop even though Starbucks gives their employees health care, stock options, and tuition expenses while the mom & pops do not.

[atarian63]

This is a discussion the wife and I had coming home from the office tonight. A subject beaten to death by everyone in the media.

Sadly I just don't see any way for manufacturing to come back. You cant fund it with borrowed chinese cash,as it would be artificial and only survive till the cash runs out. Tax free wont work but could slow down what little is left to bleed out. Multinationals already get nearly tax free status,and the cheap labor cant be beat.

Which brings us to the real and I think unsolvable problem. Once the cheap labor genie was out of the bottle years ago with the chinese etc making themselves available and the cheapest.. there is no stopping the manufacturing bleed. Once a major company in a catagory goes cheap, evetually all others must follow in order to compete. Not to mention the sales end of the problem,like Walmart which insists on low prices and not much margin from it's suppliers, virtually sealing the deal on the need for cheap manufacturing. Sadly I think the only thing that will fix this cycle is a MAJOR reset of the world economy. Sadly I think that horror is well on it's way.

Please beam me back to 1980 now please.