The Trump Administration seems to have almost completely cut off Huawei’s access to American semiconductors or semiconductors made with American equipment, with the possible exception of Qualcomm, which designs wireless chips.
This will be effective in hurting Huawei in the short to medium term, but there are possible downsides. The first is that Huawei and every other Chinese company or government ministry involved will redouble their efforts to establish independence from American-related suppliers. They will pour billions more into the effort and they also will probably redouble their efforts to steal semiconductor know-how and to recruit Chinese or Chinese-Americans who have access to that know-how to come home through the Thousand Talents campaign. American companies and American government agencies should get ready for that.
It might be a smarter strategy to keep China dependent on American semiconductors. That’s one of the last points of leverage we have. Under this scenario, American-related semiconductor companies might slow down the speed at which innovations are passed along to the Chinese. The products would be just half-a-step behind the cutting edge. This would have to be coordinated between the U.S. government and the chip companies, and that would have to be done very quietly. (There also should be a conversation about how the semiconductors are being used–do we really want to sell into the facial recognition systems being developed for use internally and internationally?)
The upside of this strategy is that American chip companies could continue to make money in China, pleasing investors, employees and shareholders. But we would slow the pace just a bit to maintain a strategic advantage, discourage Chinese efforts to reach parity, and learn how to direct the flow of chips away from the most abusive uses. But it would take industry and government to trust each other and cooperate, a radical idea in today’s environment. But that’s what it is going to take if we are going to blunt China’s global technological ambitions.
