Canon plans to disrupt chipmaking with low-cost "stamp" machine | Hacker News Reader

Canon plans to disrupt chipmaking with low-cost "stamp" machine

arstechnica.com

61 pointsbornfreddy14 comments

I think one of the more disruptive aspects of nano imprint lithography is its ability to combine multiple patterns into a single mold and imprint the entire topography for a layer in one pass.

With optical techniques today, we often have to use multiple exposure passes to get the features defined in just the right way. This is obviously extremely challenging, especially as you go beyond double patterning into 4x+. This dramatically increases the costs since you are tying up the most expensive tools with all this extra work.

I think the end game with NIL is some hierarchical approach where you create a master stamp (one per product) out of the hardest material, use this to create an intermediate stamp (one per line), and then the intermediate stamp is used to create the working stamp (one per tool).

Isn't there an etching step somewhere so that the stamp doesn't need to physically deform the workpiece? Like stamp a soft "mechanoresist" layer, then cure the resist, then etch?

The same technique could be used to create copies of a master stamp without much wear on the master.

Excellent idea. I'm sure with so high stakes people will try every possible photoresist for that and find one like what you described.

Discussed at the time https://news.ycombinator.com/item?id=39165101

from january

Western government sanctions regimes on China and other countries are becoming increasingly farcical and self-destructive and the chips war is just one of several examples - solar and nuclear energy, electrified transportation and innovative agriculture are others. A good discussion of the chips issue is at Inside China Business:

https://www.youtube.com/watch?v=0CmDZ5s8Ypo

Fundamentally, the USA has no industrial policy, no manufacturing policy, no infrastructure policy, no STEM-centric educational policy, and the over-financialization of the economy has resulted in industrial monopolies that are more concerned with delivering profits to their investors than with maintaining a technological edge over competitors via investments in in-house research and development. The political leadership is elderly and increasingly incompetent, and really seems more dedicated to the retirement interests of the largest political demographic, the boomers, than to any other sector.

Thus, rather than focusing on winning technological races on a global scale via a complete restructuring of the economy and a shift in priorities away from investor profits, the strategy appears to involve kneecapping the competition by any means available. This won't end well for the citizens of Western governments.

I agree that the tariffs and trade embargoes are destructive. However I'm not convinced that the central planning policies are optimal either. The US's over-financialization can be viewed as a destructive policy of the central bank. It is also worth noting that China has tariffs, capital controls and a central bank incentivizing misallocations in the real estate sector.

In short, it is all bad, all around with some areas of "less bad".

> This won't end well for the citizens of Western governments.

Neither does giving China everything they want, though. I agree that America isn't poised to replace all the things we import from China, but denying them from strategic sectors like compute and chipmaking makes perfect sense. China has to aquiesce to the principles of globalist security if they want to reap the rewards from international partners like ASML and Nvidia. America has no reason to give China the slightest foothold for as long as they seek political isolation from global partners.

On the flip side - there is a path for America to re-affirm it's position as a high-tech manufacturing hub China couldn't hope to replicate. It will take a while to get it working, but with enough funding America can import everything China can't have and build their manufacturing sector off that. I don't have high hopes for it, but I don't have high hopes for China catching up on EULV in the short-term either. Sanctions against China make sense from every perspective except short-term profitability.

Relevant Asianometry deep-dive from a year ago: https://www.youtube.com/watch?v=8UdNB3ZY4Ks

Second the Asianometry shout-out: Jon makes deep-dive video essays around the semiconductor ecosystem. Recommended.

Using "disrupt" like this has not aged well, and people who still seek to "disrupt" a market sound like they're reacting to trends 10 years out of date. After everyone else realized that disruption is bad, these guys are jumping on a bandwagon that's already over. That's the impression the article conveys by using such a charged word.

While “disrupt” is definitely a charged word, I think it’s fair to use in this case. Chip making is monopolistic and kind of stagnant. The existing power resting on a few (like two) enormously valued companies and the difficulty to enter the market makes me root for the underdogs here. It’s a situation that seems appropriate to disrupt.

There’s also the concept of word chasing, where one label is deemed inappropriate to use so everyone switches to a different one which also quickly falls out of favor for the same charged reason. Rather than addressing WHY something is charged, it’s far easier and very pointless to just avoid the word entirely, aside from short term good feelings.

Is it stagnant, or near-optimal?

We’ve seen eye-watering improvements in chipmaking for decades now. Could be that it is close to perfect at this point and we need new science to make further improvements.

What's the acceptable new lingo? "Canon plans to make chipmaking seamless"?

Reply on news.ycombinator.com