Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

menziep

macrumors 6502a
Original poster
Jan 21, 2006
527
1
Is there any news on the the Mobile Version of the Intel Core 2 Quad?
 

Pressure

macrumors 603
May 30, 2006
5,182
1,545
Denmark
menziep said:
Is there any news on the the Mobile Version of the Intel Core 2 Quad?

Currently there is no Quad Core processor on any processor road maps.

I think it may be doable at 45nm manufacturing process but not before.
 

Butthead

macrumors 6502
Jan 10, 2006
440
19
Pressure said:
Currently there is no Quad Core processor on any processor road maps.

I think it may be doable at 45nm manufacturing process but not before.

Doable, but not likely (not sure why, maybe power consumption constraints?) until 32nm process, a few years down the road.

http://www.mikeshardware.co.uk/RoadmapQ407.htm
"el Penryn mobile CPU is expected to be released in H2 2007 on a 45nm process. Penryn is the successor to Merom / Gilo and is expected to be a dual core CPU based on a 45nm process and containing either 3MB or 6MB of L2 cache."

Other sites have also speculated than Intel will only have dual core CPU's for the foreseeable future. That far down the road, maybe AMD will have a quad core mobile chip that Apple will use???
 

Anonymous Freak

macrumors 603
Dec 12, 2002
5,604
1,388
Cascadia
Not on any current roadmaps.

A friend of mine at Intel was just laid off. The latest processor roadmaps he saw include 8-core chips for certain in the server space, and 4-core for certain in the desktop space, with an 8-core desktop chip that hadn't been formally approved yet. Only 2-core in mobile, though. That isn't to say that Intel won't release quad-core mobile chips, just that they have no plans to do so at present. This covers new-chip development for the next two years, so it means processor releases through about 2010.

But with Intel's recent demos of massively-multicore chips, it seems likely that sooner or later, we'll see 4 or more cores, even on mobile.
 

Butthead

macrumors 6502
Jan 10, 2006
440
19
ehurtley said:
A friend of mine at Intel was just laid off. The latest processor roadmaps he saw include 8-core chips for certain in the server space, and 4-core for certain in the desktop space, with an 8-core desktop chip that hadn't been formally approved yet. Only 2-core in mobile, though. That isn't to say that Intel won't release quad-core mobile chips, just that they have no plans to do so at present. This covers new-chip development for the next two years, so it means processor releases through about 2010.

But with Intel's recent demos of massively-multicore chips, it seems likely that sooner or later, we'll see 4 or more cores, even on mobile.

With desktop systems, even though Intel is trying to improve the watt/performance paradigm, you still have die size to consider in a mobile chip that isn't much of a constraint on a desktop system, yes? So you'd need to go down to around 32nm process (2008 or later) before you could put 4 cores into a small enough package to use in a laptop...I would think.

Hmm, perhaps they will someday be able to stack them one on top of another, dual sided or something; to keep the over all 'footprint' small enough?
 

ricgnzlzcr

macrumors 6502a
Jun 7, 2005
802
0
Does smaller nm just mean that they can fit more processors in the same space that current processors use?
 

Anonymous Freak

macrumors 603
Dec 12, 2002
5,604
1,388
Cascadia
Essentially, yes. Smaller nm means more transistors for a given die size.

For example, the original Pentium Processor had 3.1 million transistors on a die that was 295 sq. mm in area, produced on a .8µ (800 nm) process. By contrast, the latest Core 2 Duo is on a 65 nm process, and crams 291 million transistors (almost 100x as many,) in 143 sq. mm, or less than half the area.

THAT is what Moore's law is talking about. 10,000 transistors per square millimeter in 1993 vs. 2,000,000 transistors per square millimeter today.

And, yes, at present, a quad-core chip would be larger than the mobile package Intel currently uses.

Stacking dies on top of each other runs into the problem that the heat generated by the 'bottom' die has to pass through the 'top' die to get to the heatsink, and the top die is generating heat of its own, too.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.