http://www.pcmag.com/article2/0,1895,2324600,00.asp

Quote:

Gelsinger's predictions: # 1.) Moore's Law continues, unsurprisingly. According to Gelsinger, Moore's Law is like driving down a foggy road at night: a headlight allows one to see some distance ahead, but how the driver negotiates the road is up to him. Intel is "on the way to 22[-nm]," Gelsinger said. "We have a good view of what 14[-nm] and [15-nm] will look like, and a good idea of how we'll break through 10[-nm]. Beyond that, we're not sure."

I thought it was still seen as likely that 16nm or 12nm would be the end for power and performance sensitive applications implemented in CMOS.

3 more predictions are sited and as allways Gelsinger shows very strong believe in compatibility as a key driver for theis chips. I kind of feel that price, power and performance really is much more important than he seem to think.

So I wonder when they will start talking about the 22nm generation in terms of codenames and dates? Doing some searching, it appears that the 32nm Gesher/Sandy Bridge was first disclosed in April 2006 (link below) so we are about due for disclosure on the following generation, assuming they are sticking with the two year cycle.

http://www.xbitlabs.com/news/cpu/display/20060428162855.html

> So I wonder when they will start talking about the 22nm generation in
> terms of codenames and dates?

The codename for the (22nm) shrink of Sandy Bridge can already
be found on the web. Hint: look for Intel's job ads.

Assuming all goes well (i.e. Nehalem ships in late 2008 or in early
2009, Westmere ships in late 2009 or in early 2010, Sandy Bridge
ships in late 2010 or in early 2011, and 22nm development won't
encounter unforeseen delays), the part should ship in late 2011 to
mid 2012, followed by another 22nm part about a year later.

Lets see whether Intel can maintain their tick-tock model without
delays, through times of recession and reduced competition, and
in the face of ever-increasing programming challenges.

Lots of things can happen to chips that are several years out into
the future. Perhaps IPF will finally take over the x86 world?

> Doing some searching, it appears that the 32nm Gesher/Sandy Bridge
> was first disclosed in April 2006 (link below) so we are about due for
> disclosure on the following generation, assuming they are sticking with
> the two year cycle.

Well, they already disclosed one tidbit -- the Sandy Bridge shrink
is going to introduce FMA (which isn't part of the initial AVX).