Tenchu: Where have you been? Intel has been teaching this ever since Centrino came out.
This may be. But, it's not what I think the future portends. From the same clip you get:
"Intel is keeping the wraps on their plans," says Gordon Haff, an analyst with technology consulting company Illuminata. Intel CEO Craig Barrett has ordered the company to be more conservative about product details until the company is sure they work, he says. For example, it's unclear which microarchitectures Intel will base its multicore products on or how much cache memory they'll contain.
<edit> This is not good news for AMD. It's obvious that Intel has something up it's sleeve in addition to all the publicly available standards they are trying to promote. AMD might have jumped on the blatently obvious but what's the encore?
Meanwhile, competitor Advanced Micro Devices Inc. has been discussing more detailed plans for its multicore products. "It says something about Intel's strength in the market that they're willing to let AMD get out in front on announcements and then catch up," Haff says.
<edit> Truly, they don't seem concerned. EMT64 is one clear example. Whereas this is needed in the future, any one implementation is needed, not two. The other standards appear to be more critical.
Also uncertain is how much business customers will pay for computers that use the new multicore technology. Microsoft hasn't said whether it will license Windows for PCs and servers with multicore processors at a higher price than for single-core machines, Haff says. "If Microsoft decides to treat dual-core desktops as two processors, that's going to be an issue for end-users and for Intel," he says.
At least two Intel technologies in the development pipeline won't see daylight until Microsoft ships its next version of Windows, code-named Longhorn and expected in 2006. According to Otellini, Intel's LaGrande Technology and Vanderpool Technology, which can deliver more secure computing and allow systems to run multiple computing environments at once, won't appear until Longhorn ships.
<edit> There are two clear signals here. One, is that it's clear that Microsoft is waiting for Intel ( and vv) in releasing various technologies. If you don't have the computers on which to put the technologies, how can you sell software?
<edit> The other point that I think is significant is that incremental increases in uP speeds won't generate enormous increases in demand (and manufacturing capability) that Intel needs. How I read this is that in the normal progression of uP technology, the fastest technology today becomes tomorrow's lowest. Dual core technology becomes the highest. What a lot of people are not thinking, is that dual core will only be incrementally higher in price than single core and single core soon transfers to dual core itself. I think Intel has this figured out wrspt capacity and the economics of manufacturing dual core and multicores. I also think that this will vastly stimulate the computer industry and applications for multi-core. While this may take 3-4 years to play out, it will provide a new upward surge in general computing. AMD has a whale of a problem on how to keep up with massive increases in production capability and implementation of add-on technologies.
Smooth
News 
Market Data 
Discover 
AI
