Just to be clear Nividia was an example. Think in terms of any of the hyperscalers/AI players that Mitsubishi serves that are waiting for the 400G EML to be integrated into a 1.6T / 3.2T pluggable. 

Think about the funding that CAI has received and the time frame that they are working toward. It really is not out of the question and is probably why it was included as a statement in that report.