Whats after intels "Skylake"

[HFS+]

topic

 

[MaxPayne63]

Alright, I'll say it.

SkyNet.

 

[Khato]

No comment as to the veracity of such, but the Wikipedia article claims it's Skymont - http://en.wikipedia.org/wiki/Skylake_(microarchitecture)

 

[Steelski]

Skyrim

 

[Idontcare]

Wiki says Skymont.

 

[Magic Carpet]

Imagine a 10 nm processor... would it still be 95w, IDC?

 

[T_Yamamoto]

Alright, I'll say it.

SkyNet.

 

[fixbsod]

Well it's only a die shrink so I doubt it would suddenly go from Skylake to SkyNET...although maybe that means SkyNET is even closer w/Skylake...

 

[Idontcare]

Well it's only a die shrink so I doubt it would suddenly go from Skylake to SkyNET...although maybe that means SkyNET is even closer w/Skylake...

ya never know, maybe Intel gets wicked serious with its "Tick+" efforts

 

[SickBeast]

OP with all due respect you are looking too far into the future.

 

[soccerballtux]

OP with all due respect you are looking too far into the future.

that's only 6-8years away....

 

[SickBeast]

It's too far.

 

[Phynaz]

Why? My next CPU upgrade may not be for that long.

 

[SickBeast]

It's impossible to predict that far is all.

Arm could take over by then for all we know.

 

[denev2004]

that's only 6-8years away....

Tech change rapidly aren't they
But still we can guess some details of which based on some paper we can see now.

 

[Khato]

Here's a hint as to why I wouldn't trust that Wikipedia article claiming it's Skymont to be correct. From the future CPU releases section on http://en.wikipedia.org/wiki/Intel_Atom

2013 : Silvermont - 22 nm.
2014 : Airmont - 14 nm.

Notice the naming trend?

 

[Edrick]

Do we even know ANYTHING about Skylake yet?

 

[Arachnotronic]

Do we even know ANYTHING about Skylake yet?

Nope. Nada, zilch, zip. Just like we know...next to nothing about Haswell

 

[dma0991]

Skymtl.

 

[ELopes580]

They will be reintroducing the Netburst arch on a 24 core CPU.

 

[gevorg]

a bit dated, but here is Intel's glance at future from 2010:

http://download.intel.com/pressroom/kits/research/2010/pdf/Moores_Law_Roundtable_MikeMayberry.pdf

 

[Idontcare]

a bit dated, but here is Intel's glance at future from 2010:

http://download.intel.com/pressroom/kits/research/2010/pdf/Moores_Law_Roundtable_MikeMayberry.pdf

Nice article, thanks for linking.

It is interesting to see what is specifically missing from the presentation, i.e. any reference or mention whatsoever to 3D xtors. 2010, Intel was obviously not going to tip their 22nm yet.

Now that they have, makes me wonder what the "real" roadmap is for the future, if they'd ever share it.

I liked his very last bullet:

Non-binary or alternate state computation

Ultimately that is where the scaling will have to head, physical scaling will be superseded by scaling of the logic in the electronic circuits themselves, multi-level cells in NAND flash is an early example of this, ternary computers and so on.

 

[Edrick]

Just like we know...next to nothing about Haswell

I would not say that. Sure there is a lot we do not know, but we do know a lot considering it is still a year away. For example, we know all the new instructions already.

 

[Magic Carpet]

Ultimately that is where the scaling will have to head, physical scaling will be superseded by scaling of the logic in the electronic circuits themselves, multi-level cells in NAND flash is an early example of this, ternary computers and so on.

Ternary logic? maybe :biggrin:

 

[tweakboy]

You wont be alive to know, HAHAHHA jk jk jk jk!!!