But it remains to be seen how many niches will be big enough to justify an application-specific CPU design, considering growing costs in all stages of processor lifecycle. Now, of course, everyone is talking about asymmetric designs – and this will indeed be a revolution, if it really happens.ĪMD seems to believe that it will be able to offer a lot of designs simultaneously targeting specific market niches. Most radical design changes will benefit only a small fraction of applications and as a result will not be cost-effective. The current business model is to make one design for all markets and sell huge numbers of chips. It is the nature of CPU to be evolutionary. You could argue that even Netburst was basically a logical evolutionary step in the direction of deeper pipelines. What kind of revolution do you expect? Itanium was probably the last really revolutionary design because of its static scheduling.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |