I don't think we know yet whether this was targeted or indiscriminate. Of eight people killed, one was an 8-year-old girl. Many pictures appear to show women or people going about daily life. It sounds like there was a compromised shipment of pagers, but it is not clear that these were specifically seeded to Hezbollah. Yes, Hezbollah probably uses a disproportionate share of pagers, but if the pagers were sold to the public through commercial channels, then it was an indiscriminate attack.

As for "launching thousands of rockets at population centers and not caring whom you hit.", I think you are confusing Hezbollah with Hamas. Hezbollah's rocket attacks have been significantly more targeted and have been primarily concentrated on military targets in Northern Israel. Further, in many cases Hezbollah's attacks and actions have been telegraphed in advance to the Israeli military, precisely to avoid an escalation. In many cases, Hezbollah's actions have been focused on demonstrating their capability of inflicting significant, serious damage, rather than actually inflicting such damage. In other words, most of the Hezbollah actions since October 7 have been more akin to firing warning shots across the bow of an enemy, rather than actually attacking the enemy.

Don't forget, Hezbollah is a transnational terrorist organization, supported by Iran, operating both in Lebanon and Syria and claiming to represent Shiite Muslim populations in particular, and Muslim populations in general throughout the region. Why exactly would they care about which set of arbitrary lines early 20th century Europeans drew on a map?

I'm pretty sure that most of those 14 million people believed that if he became incapacitated, that the person currently running would have been his replacement. So, it's hard to argue that the will of those original 14 million is not being respected.

And yet, somehow this President was able to accomplish significantly more legislatively in his first two years of office than his predecessor did in his first four. ... And was on track to pass significant legislation on Immigration, until his predecessor that claims immigration as his top concern, decided to torpedo it.

Pretty sure the 8-year old girl was not preparing to tear Israel a new one. Yes, Hezbollah probably used an outside ratio of the pagers, but it sounds like the injuries extended to a significantly larger population than just Hezbollah.

The problem is that Hezbollah has something like 70,000 to 130,000 rockets, including cruise missiles, available to fire at Israel. Up to now, Hezbollah attacks have been relatively restrained, primarily due to internal pressures within Lebanon. The greater population fears war with Israel due to the previous war, (2006-7, I think), when significant portions of Lebanon were bombed back into the stone age. In the aftermath of that war, Hezbollah paid a steep price (in terms of recruiting and support), due to the damage wrought on Lebanese society.

Today's indiscriminate attack, may be perceived as an attack against Lebanon in general, rather than strictly against Hezbollah. This could turn public sentiment about war with Israel. Why withhold attacks or act with restraint, if your enemy is going to attack you anyway? That is the danger here. Hezbollah has the capability to perpetrate significant harm to Israel. They will lose a long-term war, but if such a loss is inevitable, why not move to position of inflicting maximum damage? If that happens, the damage to Israeli infrastructure will also be significant. In general, this is bad news all the way around.

The semiconductor market is where it is today due to the market. There have been many expensive solutions that promise faster and more advanced computing, but the market for such solutions is actually quite tiny. The problem for these technologies is that Silicon can do almost as good for a fraction of the cost. The number of Customers willing to pay a 10-100x premium for say a 2-3x increase in performance is actually quite small, especially when you consider that certain classes of problems, like what most modern servers do, are scaleable, so you can get the same 2-3x increase in performance by networking say 5-6 computers (or cores). This reduces the market to only those working on problems that are not easily scaleable or parallelizable.

Don't forget Motorola, which established the Semiconductor industry in the area starting in 1949. By the 1990's they were the largest non-government employer in the state and about half of those employees worked in Semiconductor manufacturing. Motorola later collapsed, but their legacy lives on in ON Semi and NXP, which still operate out of the area. The infrastructure they built, including the extensive equipment supplier base are likely what attracted Intel to the area. At the end of the day, you'd be hard pressed to find many sites in the U.S. with a more extensive semiconductor history than Phoenix.

Wisconsin has few if any semiconductor fabs and would largely be a "greenfield" operation, requiring TSMC to build up significant supply chain infrastructure as well as a talent base. By contrast, Arizona has been a major semiconductor manufacturing site for over 70 years (see my post above), has extensive infrastructure, including a mature supply chain and equipment supplier base, as well as a large and experienced talent pool.

Semiconductor research and manufacturing has been going on in the Phoenix area since Dan Noble set up a Lab for Motorola in 1949. Motorola's 52nd street facility opened up in 1952 and started producing discrete Transistors. This business, which was called the Semiconductor Components Group was spun off into ON Semiconductor in the late 1990's and remains the corporate headquarters to this day. In the meantime, multiple fabs were established by Motorola in the 1960's and continued into the early 2000's. Motorola's Semiconductor Products Sector, which operated the Fabs was spun off into Freescale Semiconductor in 2004 and was later acquired by NXP, which continues to operate at least one Fab in the area.

In the 1970's ST Micro and Microchip opened fabs in the area and more recently, in the 1980's, Intel moved in to take advantage of the extensive Semiconductor infrastructure, which had been built up by Motorola and others. By the 1990's Motorola was the largest non-government employer in the State and about half of those employees worked in Semiconductors. This fostered a large Research interaction at the nearby Arizona State University, one of the largest Universities in the nation, which has many semiconductor related programs. While Motorola's fabs have mostly disappeared, Intel still has a major manufacturing base in the area. NXP has a fab. On Semiconductor has operations there. And many semiconductor equipment and materials suppliers operate out of the area. In short, it is an ideal location to build a semiconductor fab.

Well Said

You're forgetting about the lab space, lab equipment, computers, computer software, testing equipment and setup, and so on and so on, which can easily range in the hundreds of millions of dollars. Presumably, some of these innovations would also require vetting in silicon on the latest technology nodes. Again, developing an appropriate mask set and silicon in a foundry can easily cost tens of millions to hundreds of millions of dollars.

Inherently unstable planes are unlikely to have >200,000 successful flights before such software bugs are revealed.

We'll probably never know for sure, but the bigger issue here is that MCAS was apparently deemed a non-critical supplemental system, and thus did not get the rigorous FMEA analysis of other critical flight components. It seems that in the original design, MCAS was supposed to supply only a very modest correction, that likely wouldn't have caused the crashes. So, perhaps at that stage a low-risk determination was appropriate. However, after initial testing the correction factor was significantly increased (by 2-3x) and the new parameters were clearly not properly vetted.

Your description isn't terribly accurate either. Back in the days of planar semiconductors, the process node (eg. 45nm, 32nm, 22nm) referred to the minimum feature size, which was also at least approximately the minimum gate length. There are other features, particularly some metal line widths in the Back End Of Line (BEOL) that were also close to the nominal node size. The problem in modern technology is that as you scale down the gate length, short-channel effects, particularly Drain Induced Barrier Lowering (DIBL) begins to dominate. This effect degrades transistor performance, such that it is no longer practical to scale gate lengths much below ~20nm. We could easily print 7nm gates, but the resulting devices would be power hungry, hot and slow, not exactly the combination anybody wants!

Now having said that, some gate scaling is still possible, but whereas in the past, gates were routinely scaled by ~30% between each node, today gate scaling is on the order of a few percent and even that requires a lot of work to optimize the device junctions, so as to constrain DIBL. Nevertheless, there are still gains to be had by scaling other aspects of the device. Moving to Finfets, where the gate wraps around a "fin" of silicon provides two advantages. First, the fin itself has a vertical dimension, which corresponds roughly to the width of conventional planar device. This means that if we want more current per unit area, instead of widening the transistor, we can make the fin taller (within limits). So, we are kind of scrunching the transistors into a smaller space. The other and perhaps more important advantage is that because the gate wraps around a very thin fin of silicon, it has better control of the silicon channel, reducing DIBL and other short channel effects. Thus, thinning down the fin provides not only a density benefit, but also a performance benefit.

In today's technology nodes, the smallest dimensions no longer correspond to the gate length, but they do correspond roughly to the fin size. As far as I'm aware, all of today's 7nm processes consist of Fins in the 6-9nm range. I believe there are also some metal lines that are of similar size, which helps in providing more flexibility in routing the connections between various circuits.

Finally, regarding comparing technologies between fabs, the comparison is somewhat complicated by some specific design choices about the basic SRAM and Logic cells and how they are allowed to be connected. In particular, Intel's 10nm process, uses a very aggressive BEOL. This allows them to make small cells of comparable density to the 7nm process of Samsung and TSMC. By contrast, Samsung and TSMC have a relaxed BEOL, so they sacrifice some density, but retain a lead in the raw transistor performance. The end result is that intel's 10nm process and 7nm processes of their competitors are all very comparable.

I've actually seen quite a bit of R&D work outsourced to Russia (to former nuclear scientists) and more recently India. However, usually, this is theoretical or modeling and simulation work that doesn't require significant investments in sensitive equipment or the infrastructure to support it. On the other hand, most of the work IBM is talking about appears to be in advanced materials science. which tends to be pretty equipment intensive and is thus unlikely to be moved from their current locations in the U.S. and Europe.