Yes, there is potential for competition. IBM's plan is to make an optical transceiver that's cost competitive with copper.  However, there's still the question of IF and WHEN they can do it.

Although IBM anticipates the final product will be cost competitive with copper, they don't have a working prototype yet (sound familiar).  They will need to demonstrate performance and cost competitiveness with the functioning prototype. All this may happen in 2018, assuming they don't run into any technical difficulties along the way.

As in Rainer's note, direct product comparison still has POET with the advantage of monolithic fabrication, NO assembly required.  IBM's solution has discrete components packaged together (assembly required) therefore increased production costs.  Time will tell if they can get the total package cost down to what they hope it will be.

Assuming we can get our transceiver prototyped and in production as planned, POET could still have the advantage of first mover in this particular product line even if IBM is successful with theirs sometime in 2018.