Re: Those prices may prove reasonable for the server market, but I doubt there will be an option for desktop and mobile anytime soon.
You mean the prices of FB-DIMMs? Sure, it will be a server oriented technology for the first generation, but in a couple of years, the second generation will minimize the power and die size of the buffer, which is really the only cost adder to the DIMMs (other than vendor price premium, which will also disappear over time). The analysts seem to suggest that Intel's intercept with integrated memory controller will occur sometime in late 2007, so I would think it should reach mainstream pricing by mid-2008, plenty of time, since Intel will launch the first FBD product later this quarter.
Re: Lack of integrated memory controller makes it more difficult, which is one of the reason AMD investors have been happy campers lately.
If Intel were to integrate direct connect DDR-n, it would hurt their chipset business and complicate their product transitions, just like the lack of DDR-2 has had a small but noticeable impact to AMD these past couple quarters. FB-DIMMs solve the transition issues by putting the buffer in the middle of the DRAM and memory controller. It's the right answer for Intel's high volume business.
On the other hand, reaching performance levels without an integrated memory controller may be difficult, but Intel has already done so with Yonah, at least for the majority of apps. Even if Merom does not live up to the hype, even a small improvement ought to put it clock for clock up to the performance of the X2.
And if it does live up to the hype....
Market Data 
Markets 
AI
