Scientists finally see the atomic flaws hiding inside computer chips

This is more an incremental advance than any kind of breakthough.

Back when I was working on electron microscopes at Cambridge Instruments (1983 to 1991) we sold about half of them to the semiconductor industry, and could already do atom scale resolution. Of course it took 20kV electrons to do it, so the chip wasn't much good after it had been looked at. The voltage contrast machines that I worked on used 2kV electrons to keep track on the voltages on working integrated circuits with nanosecond time resolution, and 2kV electrons didn't wreck the chips. I imagine that Cornell has come up with some kind of new trick, but the science daily press release doesn't say what it is, and it doesn't offer any kind of link to anything more informative.

The paper says '3D': 3D_atomic_scale_metrology_in_transistors_ptychoraphy__ s41467-026-69733-1_reference.pdf Was yours 3D?

Bill Sloman snipped-for-privacy@IEEE.org wrote: |-------------------------------------------------------------------| |"This is more an incremental advance than any kind of breakthough."| |-------------------------------------------------------------------|

I believe that a reasonable hypothesis is that few things are breakthroughs. I believe that a reasonable hypothesis is that many scientific works are incremental advances. An overwhelming amount of supposedly scientific publications are easily shown to have a significant error which any relevant competent scientist can detect from simply readings them (i.e. not a problem of graphs fudging withheld raw data), but gatekeepers do not even bother to check. Tax payers waste subventions supportings so-called scientists' prestige.

(S.

Only if you took enough 2-D images. The stages were tiltable, so it was a theoretical possibiity.

I've refereed scientific papers, and suffered from referees who didn't know as much as they thought that they did. The referees are gatekeepers, but they don't get paid for their work, and they are human, and do make mistakes.

I've also posted comments pointing out that some papers that did get published weren't referred as well as they should have been (not that one makes that point explicitly in a comment - it has to be accepted by the same editors who hadn't found adequate referees). There were't all that many papers that were bad enough to attract that sort of comment. Sturgeons Law says that 90% of everything is rubbish, but most bad papers are just anodyne, rather than wrong.