By using this site, you agree to our Privacy Policy and our Terms of Use. Close
Pemalite said:


You would hope so. But it's still not a true 7nm process. :P

I don't really care if it's not, what I care about is what the offerings are going to be ... 

If Intel's competitors are offering 7nm technology that has comparable densities to their 10nm technology in a similar timeframe then I don't see it as a big deal plus since when did we consumers need to about this when that's the fabless chip designers issue which I assume that their educated enough on these matters ...

Pemalite said:

 

Well the deal was for 14nm plus it's vcariations like LPC, LPP etc'.

No reason why they couldn't expand on that at a later date if they need to. If.
Also I think Global Foundries also has a few deals with IBM for patents anyway? I would assume the same holds true with their Samsung connections.

GF are in it with a long haul with 7nm and plus they expect to iterate upon it too so I don't see where Samsung comes in the forseeable future ... 

Pemalite said:

 

Intel still has an edge in transistor technology though and is still a few years ahead of it's competitors. 

https://www.extremetech.com/computing/246902-intel-claims-three-year-advantage-10nm-process-wants-change-define-process-nodes

Intel 14nm process is actually a 14nm process. Samsung's isn't. Intels 10nm process will actually be a 10nm process, Samsung's isn't.

https://www.extremetech.com/computing/249075-foundry-futures-tsmc-samsung-globalfoundries-intel-gear-7nm-beyond

*snip*

As for EUV, while it's great to hope for it... At the moment it's still fairly expensive to implement (More so than multi-patterning approaches it seems). Right now Samsung seems to plan for it's introduction with it's CLN7FF+  process in 2019-2020.

Intel might introduce it much later at 5nm.
 

Actually Intel's 14nm process is actually closer to 16nm and their 10nm process is closer to 11nm so like it or not their numbers also don't add up ... 

By the time Intel is ready with 10nm (sounds like Canonlake is delayed to 2018), Samsung will respond in 6 months with 7nm (they'll start with DUV then go EUV with 6nm in 2019) which will have a 22% density improvement ...  

Pemalite said:

 

As for Investment. Intel is still competitive with Samsung on that Front. So to counter Samsungs $7 billion dollar investment... I give you Intels $7 billion dollar investment. :P
http://www.pcgamer.com/intel-is-pouring-7-billion-into-7nm-chip-production-plant-in-arizona/

Intel is also opening up it's fabs for select 3rd partys as well.
https://www.theverge.com/2016/8/16/12507568/intel-arm-mobile-chips-licensing-deal-idf-2016
http://www.pcworld.com/article/209573/article.html

So that should help drive some additional investment for Intel
.

Intel's investment is over a 2 year period whereas Samsung's investment is over a 1 year period so not only is Samsung spending twice as much as Intel in the same period but this does not factor in IBM and GF R&D so Samsung has even more capital to draw from ... 

It's great that Intel has started to produce ICs for other chip designers but it's too little and too late so their days at being leaders in transistor technology are over ... 

Pemalite said:

 

Hence the "Caveats".


Still. If there is a demand/need that process can be improved upon for high-end, power hungry chip designs.
But even that 12nm process isn't a real 12nm process anyway.

I don't know about you, but I am tired of all the bullshit "nm" marketing angles that TSMC, Samsung, Global Foundries and IBM use. It gives people the incorrect idea.

Then you should be mad at Intel too for misadvertising but these numbers have no meaning for us when it's for engineers who are the ones designing the chips ...

Pemalite said:

 

Hopefully Navi brings the goods.AMD doesn't have a graphics side. They aren't even really competing against nVidia right now. 

Vega will help bridge the performance gap... But that was needed 12 months ago.

Vega needed to more than double Fury X's performance to be a sound winner, that's a hard task when AMD didn't increase the amount of pipelines in their design, so most of their gains comes from efficiency gains and clocks.

It won't, AMD's only salvation is through Microsoft ... 

It wasn't AMD who solely topped Intel with the K8, they had help from Microsoft standardizing AMD64 in their Windows ABI ... 

They need to keep looking to Microsoft again so that they change the spec once more in DirectX to begin to be even remotely competitive with Nvidia. AMD needs to seriously streamline some more of their graphics hardware extensions into official optional features for DirectX ... 

The only way I can see Vega competing against Pascal is with shader model 6 games and there's none of them so far. It's just another lost generation for AMD at this point ... 

The biggest leverage AMD has so far against Nvidia is that Microsoft is on their side so they should use that partnership that they have with Xbox to extend into changing the DirectX spec ...