Thermal resistance in electronics packaging is often dominated - not by thermal resistance of the bulk material (e.g., copper), but by the thermal interfaces. I have low confidence in graphene as a mature and deployable technology. They would likely use graphene as an interface layer, and they will have to watch yield very carefully. They likely already do x-ray every cpu after the mainboad is assembled to inspect the die for voids. As far as I know, graphene can't be bonded or physically attached well, there could be ageing problems with the interface if it isn't studied closely. The risk is that 4 or 5 years down the road, air voids could develop between the CPU and heat spreader /chassis and cause overheating in aged devices.