brennanpeterson

Why? It then proves nuclear is uneconomical?

Regulatory arbitration like this serves nothing?

Don't get me wrong, I would love to see more nuclear, but bad nuclear is not in fact better than none.

So prayer beads aren't meditation? Indian traditionalists are out. Walking isn't, so zen is out. Chanting isn't, so.i guess all bhuddism is pretty much not ok. Maybe there is no true meditation.

It is ok to be focused and mindful with different aids. Always has been. There are self focused (mindful) and also self-abnegating traditions, and all count.

No harm to.me if others have a different way. If music, or chants, or gardening, have at it.

Semiconductors are also incredibly risk averse, and requires fairly large gains to be worth the risk.

There are domains (or, mems, photonic sensing, high power) where innovation has a bit more flexibility for the early going.

Well, you are working from all sorts of misguided assumptions about how you convert, the yield, the efficiency, the shared capital, and history.

Old gen logic barely exists, and won't convert. The tools are wrong so it wouldn't yield. And there isnt the expertise (on tool recipes and integration) to do it.

Historically, memory is a sawtooth business. But history isn't a great guide here: 10 years ago, a new plant added meaningful capacity, as it came with a shrink of about 30%. Today....it doesn't. So it take huge capital to add a very small amount of capacity.

You can dream that things like 3d dram and 4f cells will help, but they are unlikely to offer enough with demand.

And finally, everyone running a dram plant has lived through these capacity boom bust cycles and the consequently layoffs and pay cuts. I suspect they are happy to take money this time. I would be in their place.

I do hope we see a new dram or dram-like designs. And that there are more efficient dedicated ai processors. And that models themselves become more memory efficient. But I don't really believe any of these come soon.

"If by some fiat I had to restrict all this writing to one sentence, this is the one I would choose: The summit of Mt. Everest is marine limestone."

John McPhee from the wonderful Annals of the former world

It would be nice to have portfolios, but systems are broken enough that that becomes hard to see. I suspect one of the reasons for the bias to hiring PHds in fields where it really isn't necessary is at least you have a work portfolio.

My work laptop will stall on resize constantly, and I suspect it is due to the mess of security and backup software. Windows does have an ecosystem problem.

I am also baffled by the multiple control points. I can log in to mail in 3 places. Settings have 3 with different uis....it is gross.

None for normal.compute, since energy density is still fundamental. But the interesting option is cryogenic computing, which can have zero switching energy, and 10s of GHz clock rates

Some neat startups to watch for in this space.

Group theory and crystallograpy without either word? I suppose I can look at this as an extension of group theory to glassy and partial.domains, but it doesn't appear to offer much more.

Columnar basalt formation has been understood for a long time, I really don't understand what this explained that wasn't already known?

I also quite liked https://ocw.mit.edu/courses/res-6-011-the-art-of-insight-in-...

Which is, I think, the successor and quite useful.

Variance, data rate/cost, and lithography.

You can do lithography small but slow and expensive. But small means you need a stack, which is even more expensive. At small sizes, defectivity/variation are really difficult.

So if you want a paradigmatic shift, you need low cost patterning, and the best way I can see is to use clever chemistry and a much different design style.

I wonder if that is why it worked as opposed to why it sold. At least, as one of the users in that time it seems to me that the itch was the document repository with some features, and the rest was fluff for purchasing to sign off on.

All the calendar, task, website stuff ended up dying away, because what we really needed was a good document management system, with optionally some simple signoff loops and notifications.

That was great. Yes, it was just unix like tools with a window. That is the craigslist of os improvements.

It really was the use case, but the simple one. I saw plenty of power users try to do complex and ultimately fragile uses that died away.

In the case of this user, I love the idea of turning an email to an action, but I also need to add that to my action board and assign it, and check people time, at which point the simple action only makes sense for individuals, not teams or orgs. and I need to add a couple missing actions, and summarize. So suddenly the all in one is a marginally useful tool that is also a straightjacket. And then I go back to outlook and jira and excel and trelli or whatever.

This isn't true. It used to be, as a new fab would appreciably add quantity. At 1M wspm in 2015, a new 100k fab at the most modern node would add effectively 20-30% capacity, and usually multiple.players at once, since all had cash.

Now, the relative shrink is tiny, so capacity adds are just wspm, in effect, and that gives 5%.

Put differently, you cannot invest your way out of the shortage, or into meaningful share....so you take profit.

Or The Crossing, which, at least for the first third is his sparest and best writing. At least, I prefer the marriage of the gothic sensibility and poetry with the classic western.

Not sure on limbs, but for fast bone and tooth repair it works.

https://www.nature.com/articles/srep31724

We aren't far at all from on chip combs or pseudo combs, and that will be fine. More for sensors generally, but you can also have your clocks.

Please tell me how to make a water prism to test compressive strength and deformation resistance. Water is an incompressible fluid, that is different.

These are well understood terms in the field. Unfortunately, this illustrates the bounds of ai in subfields like materials: it confuses people.

The sram density is a pretty good equivalent. You can arguably do the average of sram and some logic.

If you take the square root of that...you pretty much end up with (modulo a linear scale) the existing nodes.

That gets you size. You then need power and speed, which are a bit trickier to compare without a standard/reference device.

Spin in is an interesting tech history. As for cvd low-k, it is mostly how much C is in your silicon, and likewise how you setup the damascene etch stop. Intel was low-ish k in about 2002 on 130nm.

I am not so sure tsm was first. Depends on how you define lowk.

Probably done with 3 separate litho/etch layers, where they etch and process in groups of 110 or so.

Each of those layers can have a cell, so if you have a tlc device at a 100nm pitch, you have a density of 321*3/(1e-4)^2 bits/mm, or about 1e11bits/mm2.

Fun reference: atomic density is 1atom/.5nm, so 1/5e-7^2, or 4e12/mm2 ish.

Not too far away.

It exists, though I don't know of any free or open versions

But there are great books in this area?

https://www.statlearning.com

https://www.stat.cmu.edu/~cshalizi/ADAfaEPoV/

There are other fine ones, but these are very good.

Lithography lives in the thin film approximation anyway. Timiosheko is a good reference. There are papers from Barnett or Nix that are very nice, but edges will probably end up a fem solver domain.

True! I went a little far in the name of 'eli5'. I think it roughly holds that you gain about a factor of 1.5 in routing density by removing the power distribution, so you can relax some critical patterning. But I havent looked closely in a long time.

I am not sure where that would come from. There is nothing about dsa that means this.

Dsa is one of many patterning assist technologies, just...an old one. Neat, but not 'new'. You use patterning assist to make smaller, more regular features, which is exactly what the 16a vs 18a refers to.

That has somewhat less to do with performance, which is tied as much to material, stress, and interface parameters. Nothing gets better from being smaller in the post dennard scaling era, the work of integration is making better devices anyway.

Patterning choices imply different consequences. For example,.a.double euv integration can take advantage of spacer assists to reduce ler and actually improve cdu even with a double expose. Selective etch can improve bias, spacer trickery can create uniquely small regular features that cannot be done with single patterns. Conversely, overlay trees get bushier, and via CD variance can cause horrific electrical variance. It is complicated, history dependent, and everything is on the developmental edge.

Half your wires deliver power, half deliver signal. So if you do both on the same side, you need twice the density of wires. If you split the delivery into two parts, you get double the density without needing to make things smaller.

The benefit of X-ray is they go through everything (so you can see metal). The problem is they go through everything, so you have little signal.

For metal/bump voids, hard to inspect any other way

The harder problem for X-ray inspection is that 1) resolution is difficult to scale economically (possible, but hard). And 2) if any version of wafer or die direct bond dominates (I tend to believe it will), then x-ray won't be good enough anyway, so hard to spend the investment on this.

You are correct. You can do mask edits or short runs in some cases beam.

Even more interesting is mask level overlay and CD correction using ultrafast absorption based stress fields. There are some heat technologies out there

Very capital intensive means operating costs (wages) are a small percentage.

A trailing fab has very few people running the operation. This will be a trailing fab.

It isn't complex at all. Take the average density (average of sram and logic), take the square root of that, and you have (barring a factor) the node.

The best way to understand the dimension is to look at the contacted gate pitchm2 pitchcell height,and take the square root. The pitches obviously make an area, the cell height sets the 'design' component. So you get a reasonably scaled number.

Putting it that way also clarifies technology. Why euv, to tighten contacted gate pitch. Why cobalt: to tighten up metal 2 pitch (without killing resistance). Why backside metal: lowers cell height. There is of course more to it, but it is a good way to coarsely understand

A good way to know if it is marketing is if someone talks about feature sizes instead of pitches. Patterning is done by pitch. For example, you don't do euv to make smaller features. You do it to make more complex layouts at tight pitch.