It’s been quite common to see comparative advertising in the States and UK for years. As long as it’s factually correct and not misleading it is allowed.
This is one of my favourite examples of a campaign that really sticks the knife in. Using 1 IBM server for the named task is faster than using 4. It doesn’t improve processing speeds. Not so with Oracle: add several servers together and speeds improve 3.5 times.
The beauty with this ad, which is plain Fact A compared to Fact B, is the source of the data: an IBM research paper presented at a conference. The data was handed on to the Oracle marketing team, who knew a good story when they saw one.
It would be nice to know if the IBM person who presented the paper is still working for them.