Rich,
Thank you for the summary of the WSJ piece - I look forward to reading it tomorrow [I don't have an on-line subscription].
I am sure that DEC would have done its homework. The patents they would have cited would have been ones for which at least surface similarity could be shown to whatever techniques Intel used. However, that is different from showing intentional infringement, and it is different from showing that one or more of the claims were actually infringed.
Digital did not, of course, invent speculative execution [nor out-of-order execution], and some form of branch-prediction method is of course necessary to do the "speculation". If Intel accidentally or incidentally infringed, I would expect that the impact would be quite limited as to damages [as I mentioned in another post] and as to remedies [e.g., it could license IBM's branch-prediction method, i.e., assuming that it doesn't already have cross-license rights].
IMHO, that is why Intel is holding off commenting until it has thoroughly examined these patents - they would want to assess how central the claimed patents are to their design, and so how best to respond. If they deem the patents peripheral and the claims weak, they may decide they have a strong hand and may choose to respond to DEC quite strongly legally, as well as with a heavy PR counterattack. We just don't know, at this point.
Best regards,
Arno |
