Considering TSMC corporate culture and staffing practices, and the scale of raw material and supply chain complexity needed for advanced integrated circuits manufacturing - most of which China dominates - I think we're in a bind here. It was the overlap of American capital and engineering and East Asian resources and labor that made the current scale of compute possible.
As China absorbs Taiwan when their area denial capabilities become credible enough to keep the American navy away, we can expect an end to cheap compute. Either nationalists in Taiwan degrade assets enough or China simply shuts us out. But the second half of the 21st century won't be full of enterprise saas and a supercomputer on every pocket.
Hot take: If that means on-shored computing HW jobs back in the west, I'm all for it. I'm only upgrading my phone/PC every 6 years or so anyway, I can live with them being more expensive if they're made in the US/EU and provide local jobs.
Them being more expensive could also mean more focus on SW improvements and optimizations instead of throwing more HW resources at bloated and poorly written SW, more focus on longevity and repairability, and less e-waste for the environment, at the cost of corporate profits and the consoomerist attitude focused on hyping up this year's must-have "new-shiny" that's just 2% better than last year's "old-shiny". All I see is a win-win.
Fair points. Although if computers become more expensive the overall result is everyone becomes poorer.
Also, the energy transition towards decarbonization entails smart grids and intense use of computers for design, deployment and operation of a more diverse fleet of energy generating and storage assets.
People with the right skills might be better off, that's true.
>if computers become more expensive the overall result is everyone becomes poorer
How so? This might have been the case in the '90's and '00's when computers were still in adoption for home users, but nowadays, almost everyone already has a usable PC and a smartphone, even in developing nations in Africa.
Them being more expensive will mean only businesses and rich kids will get to play with the latest shiny, while Average Joes will have to stick to older tech and focus on sustainability and repairability instead of upgrading every 2-3 years.
I could be wrong, but I see this generating more HW repair jobs for technicians in every community like TV repair shops in the old days, and even more SW jobs needed to improve SW to run on older HW in order to target more consumers.
I don't see how this makes everyone poorer, other than execs at companies making a living selling a new HW widget with 2% improvement every year while abandoning SW support for last year's widget to force everyone to throw their old widget away, cough, Quallcomm, cough.
If things are more expensive, people can have less things and are thus worse off. This is circular, poor = less able to acquire = things are relatively more expensive.
American generals have stated it could be as early as 2027 or 2030. When a country says they'll take land as soon as they can, I'm inclined to believe them. China missile deployment on South China sea is quickly becoming overwhelming.
The US has bases both north and south of Taiwan, in Japan and the Philippines respectively.
I’m against China seizing territory in the South China Sea and believe that the US needs to focus on rebuilding its MIC (as shown in the Ukraine conflict). We should address both of those serious issues promptly.
China says it's coming. US military says it's coming. Policymakers in US and EU are adapting to this shift. How am I getting downvoted and told off for trying to discuss the consequences to tech markets in a tech oriented forum on a thread about the consequences of it?
I’m not telling you off: I think you raised a common sentiment and I was responding.
I’m disagreeing that the saber rattling from two countries both experiencing civil unrest and economic hardship will translate into an armed conflict for Taiwan (generally) and that the missile buildup in the SCS is “overwhelming” (specifically). My point about US bases adjacent to Taiwan is a specific response: we already have forces with both access to the Pacific and clear lanes to Taiwan, positioned and ready.
The job of the DOD is to let the nation know our military status — and they’re letting us know that the building of US fleets versus Chinese fleets implies a weakness in the 2027-2030 range, which could be utilized by China. But that doesn’t automatically translate into an armed conflict during that period.
More broadly we’re shifting policy to compete (eg, by on-shoring) — but again, that’s different than an armed conflict.
I think navigating this conflict requires we not be hyperbolic in either direction: dismissal nor concern.
I agree with the content of your comment. I hardly think I was hyperbolic when stating "compute might become more expensive, thoughts?". A knee-jerk "whoa calm down" response to people raising such topics will inhibit healthy discussion. Someone said this was a non concern and I defended the value of thinking about it and discussing it. So at that point to say things are getting blown out of proportion is not warranted at all.
Foreseen doesn't mean fixable. The scale of production achieved in East Asia is unique and doesn't look reproducible elsewhere at the same price points.
As China absorbs Taiwan when their area denial capabilities become credible enough to keep the American navy away, we can expect an end to cheap compute. Either nationalists in Taiwan degrade assets enough or China simply shuts us out. But the second half of the 21st century won't be full of enterprise saas and a supercomputer on every pocket.