Any updates on full, scale-out clustering?

And are there any plans on supporting columnar storage, for very wide tables?