RE: Hari
Rich wrote;
> Really the only major issue is how to deskew the lanes and
> where the deskew
> function is performed. 8B/10B codes really can not/should not
> be split into a
> finer granularity than 10-bit code-groups. I believe that you
> did not intend
> otherwise.
My point regarded the fact that 8B10B coding rules include tracking
disparity and using that for error detection. I somehow thought that
disparity was not being maintained on a per-channel basis, but looking
back at your presentation, I see that it is. Never mind! :)
Regards,
Dan Dove