SI
SI
discoversearch

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor.  We ask that you disable ad blocking while on Silicon Investor in the best interests of our community.  If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.
Technology Stocks : Intel Corporation (INTC) -- Ignore unavailable to you. Want to Upgrade?


To: willcousa who wrote (185431)12/9/2007 10:43:36 AM
From: robert b furman  Respond to of 186894
 
Hi Will.

Right on.

I believe the evolution will feature speed rated Graphics chip - (much like Intel won on Megahz rating on mpu) and speed rated memory as well.

I think this will prove to be the reason Intel bought into Micron.

As micron differentiates itself with Intel's future plans/needs - it will remove itself from the Asian insanity of commodity pricing of their memory chips.

Not to a big degree - but it will provide somewhat better margins(certainlt better than below cost).

IMO

Bobbe



To: willcousa who wrote (185431)12/10/2007 12:45:19 AM
From: greatplains_guy  Respond to of 186894
 
"It is based on their continued integration of graphics into the CPU. AMD has announced they will have an integrated product - I forget the date for that. I expect at some point they will figure out how to integrate memory as well. Wi-max?

It is the way I would expect the processor business to evolve regardless of who is doing it. It is a great way to add value to the product."

I agree that the quest to integrate more functionality will continue. It certainly helps throughput when the external IO bottle neck is eliminated. The current push of INC is to put more processors into one package.

I still remember clearly a Sales Engineer for a major OS enhancement software company telling a room full of techs that increasing parallelism within a computer faced the law of diminishing return. He quoted statistics like 50% increase in computational power from one to two. Then he suggested an increase of around 25% for the third and 13% for the fourth.

The value of small marginal improvements holds as long as buying additional computers (servers) is not economically justified given the relative cost of marginal increases between more cores and more servers. Due to the diminishing return there is a point where more boxes produces greater returns than increasing the number of processors in each server. Of course software licensing can skew that as well.