Intel and AMD

[hash]

I don't know if this topic belongs here. But I was eager to know as to how many years at a time does Intel and AMD plan for. I mean like right now Intel would have already planned what its going to release in 2011 . I just want to know how many years they plan at a time?

I woke early in the morning and started to search google for answers but nothing convincing. Anyone?

 

[505090]

With all the research and development i would think easily the next 25 are being worked on with the plans getting loser the farther out.

 

[hash]

Amazing. I consider people who work there are really lucky,they'd probably have 8-core at hand already.
I mean from Single core - Duel Core - Quad core like in 6-7 years? And they've already done 25 years planning for bettering the current ones while the current is more than enough.WOW!

 

[Reynod]

One plans to stay on top ... the other plans to stay in business.

 

[fazers_on_stun]

You could try looking at both companies' roadmaps which generally go out 2-4 years; however these (particularly AMD's) change subject to market conditions. For example AMD used to plan an ultra-low-power CPU codenamed Bobcat, for the netbook market, but dropped it last year as unprofitable. After seeing Intel sell tons of theirs (the Atom), AMD decided recently to get back into that market so they once again have Bobcat on their roadmap. Yeah, they wasted a lot of time and opportunity and profit but given their financial condition they can't afford to be in every market all the time, so tough choices have to be made, plus their crystal ball was a bit cloudy a year ago .

 

[jsc]

I mean from Single core - Duel Core - Quad core like in 6-7 years?

More like 3 - 4 years.

 

[Dekasav]

They plan several years out. 3-4+ years out is being researched/engineered, but further is probably planned by both companies.

Though both do like to change plans periodically, especially AMD.

And we're at 6-core....

 

[JAYDEEJOHN]

Thing is, things have to change, as we are hitting the shrinkage/smaller node barrier using silicon. How this affects both companies, or allows for more competition altogether, no one knows. Getting beyond 11nms isnt going to be easy, and we will likely see different materials/approaches then, or , even the death of x86 computing

 

[rambo117]

oh man, titles like that ALWAYS scare me. more like "fanboy fight to the death"

but yes, just as jaydeejohn said, the smaller the process, the harder it will be to manufacture.
so in a way, technology does have its limits unfortunately. they will have to resort to optimizations... but thats far away, so no need to worry(or look foreward to) yet

 

[JAYDEEJOHN]

In 4+ years, we will be at 16nm or thereabouts, below that, we may have to change. If x86 just doesnt cut it anymore, thats a possibility, if we cant find a better porcess, take a different approach

 

[Dameon_Bananaman]

as far as i was aware x86 isnt a "architecture" as such. its basically a set of instructions so amd and intel can design their processors in soooo many different ways and it still be x86 because of the set of instructions programmed into it.

but jaydeehjohn is right, when the traditional approach isnt getting much thurther, then take a completly differnet approach like IBM have "supposidly" claimed is possible which i read a long time ago, but with IBM you take things lightly because they are good at claiming stuff and never doing anything about it. such as stacking cores on top of eachother and instead of using copper tracks to communicate between cores, use light beams at differnt frequencies which makes the core speed of the cpu much higher.

all these have been claimed by IBM but that was years ago.

Also all companies have short and long term planning and amd and intel are no exception. it would be stupid to assume otherwise that they are not only working on the next generation of cpus, but the generation after that and possibly the one after as well. processors take a long time to design and test.

 

[JAYDEEJOHN]

The arch is designed for x86, and may have to be changed to something else, thats better for usage. Sounds extreme, but if the process end fails, whats left?
Weve all read about nano this and carbon that, but in high volume mass production? Time will tell