My busiest time of the year continues. A few interesting links:
- I'm sure you've heard of wordle. There are some other free games that are similar in design (and not co-opted by the New York Times) that are also good to keep your brain tuned up. Mathler is cute (and knows order of operation), and worldle, while difficult to pronounce, is fun if you want to keep up with your geography.
- The US presidential science advisor resigned this week, apparently because he is just an awful person to work with. That prompted this article from Stat, which advances a thesis that I don't buy, that this resignation shows that we are leaving the era of "big ego science". I hope that we are finally entering an era where bullying and pushiness are not automatically tolerated in high profile positions, but drawing some sweeping conclusion from Lander's departure is not reasonable to me. I do know from interactions with folks like my colleague Neal Lane that it is possible for top-flight scientific leaders to be both highly accomplished and genuinely nice people. I hope someone in that mold ends up taking the reins.
- The US House passed the 2022 version of the America COMPETES act. The US Senate had passed the related US Innovation and Competition Act (USICA) last summer. Now it's up to the conference committee to try to work out a compromise bill that can pass both houses. If this passed and the House actually appropriated the funds, it would be big news for NSF and the DOE Office of Science. I'm a bit cynical about the prospect of this happening, and both bills have issues, but it's better to have this at least in front of Congress than languishing off-stage.
- There are rumors that Nvidia's next big chip will be built on the TSMC "5 nm node" process (where the numbering really doesn't mean that a critical lengthscale is 5 nm) and hold 140 billion transistors (!!). If anyone asks you whether nanotechnology is meaningful, point to things like that as examples of nanoelectronics.
8 comments:
Prof. Natelson, the Senate version of the COMPETES act contains $52B in subsidies for fab construction in the US. Do you think it’s a good use of taxpayer dollars to give subsidies to companies like Intel that have been lagging behind in Moore’s law scaling? Wouldn’t you rather have the money go to pure research rather than corporate bailouts?
Dear Anonymous,
Worldwide, chip production is expected to increase dramatically in the coming decade. Now, just think of what will happen if 90% of chip production is controlled by countries that may not necessarily remain on friendly terms with the United States.
Intel may be a node or so behind their competitor right now, but is still producing competitive processors for data center and client applications.
The U.S. government needs to both fund academic research, and to support its industry in a strategic way. Both are important.
I'm on the same page as Gordon here. I've been saying for years that it is a serious risk to national security and the global economy to have such a high fraction of chip fabrication being done in only a couple of locations. With the pandemic we've seen how badly screwed up supply chains can get when everything is "just in time" and no one has any inventory on hand, and that's without any military conflict dimension. Bear in mind, too, that other countries subsidize massively. $50B is, I believe, small compared to what the PRC government did in terms of favorably underwriting silicon-for-solar in China. I want more funding for basic research, but I also think the US needs to look strategically at how important it is to have more and better local manufacturing infrastructure.
The original Anon poster here. Thanks Gordon and Prof. Natelson for your replies. One benefit of the $53B corporate welfare is that it will create more jobs for folks who get their PhDs in nanoscale related fields like EE, Physics, Chemistry, Mat Sci, etc. That does warrant consideration as well.
Anon, if you'd like to point fingers, please make sure you point them at the right direction. That would be those responsible for the deregulation and financialization of US corporations. Tragically, it is the same ilk that today tries to patch a gaping wound with a band aid.
Anon @ 5:29 AM, Wall Street culture can be part of the problem but I also think there is a significant delta between the quality of the Taiwanese engineering workforce and the US engineering workforce in terms of technical smarts, work ethic and general effectiveness. There was an EETimes article about the difficulty that TSMC is having in Arizona. https://www.eetimes.com/tsmcs-arizona-culture-clash/
I wonder how much of Intel’s woes can be due to lack of investment vs lower quality engineers than TSMC.
Steve Chu was far more obnoxious than Lander but he was never fired by Obama nor did I see any such criticism of him. Maybe it is best to put a woman who raised kids to that position since the bar is set so high and so vague that being a male prof automatically disqualifies anyone. Smalley, Lieber and Weinberg were also far more abusive than Lander. In fact, you would be hard pressed to find a lenient male at top echelons of academia. I hope I am not the only one who is aware of this worst kept secret.
Anon @ 1:40 PM, one of the effects of the financialization was the decimation of a highly skilled workforce.
Post a Comment