On Possible Tech Impacts on Tech Due to the Taiwan Quake of April 2nd 2024.

[Uttamattamakin]

Summary

Just the facts. TSMC "Still assessing, and Framework temporarily paused production.  This will have an effect since the region of the world these products are produced in was disrupted by the quake.  Our first thoughts must be for any and all whose lives were lost or lives altered by this event. 

 

Quotes

Quote

 

Framework is stopping production until further notice, while Apple and Nvidia chipmaker Taiwan Semiconductor Manufacturing Co. (TSMC) has stopped production at some of its factories and evacuated employees, Bloomberg reports.

 

"Our production in Taiwan is temporarily paused to prioritize worker safety while our manufacturing partner checks for any impact from the earthquake today," Framework wrote on X, formerly known as Twitter, late Tuesday night. "We're hoping for the best for our friends, colleagues, and partners across Taiwan."

TSMC plans to resume production overnight beginning Wednesday evening, per Bloomberg.

 

 

Quote

Taiwan’s Strongest Earthquake in 25 Years Kills 9 and Injures Hundreds. The magnitude-7.4 quake was followed by more than 200 aftershocks. Dozens of people were trapped. Two buildings in the city of Hualien teetered perilously.

 

 

My thoughts

It is probably too soon to know what effects this will have on the tech sector.  Thankfully this 7.5 magnitude quake hit Taiwan which has high building standards and an economy that can meet those standards in practice. The buildings being strong enough to not totally collapse likely saved a great many lives.  Compare to the Quake which greatly damaged Por Au Prince Haiti some time ago which was about the same magnitude.  Well-built modern structures stood but a great many did not live in such structures.   

 

Vibration and shaking is probably not good for precision Manufacuring equipment or for things that were in production at that moment.   This likely won't be some big systemic effect.  I really hope it won't be.  I would not be surprised if there weren't some defective parts in the supply chain for a while. 

 

Sources

https://arstechnica.com/information-technology/2024/04/tsmc-still-assessing-chipmaking-facilities-after-7-4-magnitude-quake-hits-taiwan/

https://www.pcmag.com/news/framework-pauses-production-earthquake-hits-taiwan

https://www.nytimes.com/2024/04/03/world/asia/taiwan-earthquake.html

Reuters (tweet above)

Interesting Engineering (Tweet above)

[OddOod]

TSMC has said they will resume production in affected factories tonight

via https://www.cnn.com/asia/live-news/taiwan-earthquake-hualien-tsunami-warning-hnk-intl/index.html

[Forbidden Wafer]

Daaaamn. Hope taiwanese friends are safe.

On a different note, everytime I see a borked building like this I wonder why aren't most buildings built on top of a radier (concrete slab). I know it is more expensive than deep foundations for large buildings, but I still prefer a raft over something that can get sheared by differential settlement. 

[Uttamattamakin]

1 hour ago, OddOod said:

TSMC has said they will resume production in affected factories tonight

via https://www.cnn.com/asia/live-news/taiwan-earthquake-hualien-tsunami-warning-hnk-intl/index.html

The big news there is that their EUV tools are apparently undamaged.  Those are what makes our CPU's and GPU's etc.  I can't imagine that any wafers that were on the line are just ok.  The defect rate in those will have to be higher. 

 

That the tools themselves are ok is great news. 

 

Then there are all the indirect effects of the disruption around TSMC, and other companies, that will have an effect. 

 

IMHO

This is why the world needs to try to create more than one place that is really good at semiconductors.  Taiwan has used being indispensable for semiconductors as a national security priority though.  If we see an invasion suddenly an 8 core CPU in your rig will be like this. 

 

[StDragon]

1 hour ago, Uttamattamakin said:

The big news there is that their EUV tools are apparently undamaged.  Those are what makes our CPU's and GPU's etc.  I can't imagine that any wafers that were on the line are just ok.  The defect rate in those will have to be higher.

Those machines have all sorts of calibration routines they run post installation. I wouldn't be surprised at all if ASML designed their equipment with robustness to withstand the journey of **shipping and earthquakes for any fabs located around the ring of fire.

**For shipping, complex machinery (including large office multi-function printers) often have pins and clamps that must be removed before operation. Though they still have to be robust beyond what simple temporary fasteners can provide. 

[leadeater]

1 hour ago, StDragon said:

Those machines have all sorts of calibration routines they run post installation. I wouldn't be surprised at all if ASML designed their equipment with robustness to withstand the journey of **shipping and earthquakes for any fabs located around the ring of fire.

**For shipping, complex machinery (including large office multi-function printers) often have pins and clamps that must be removed before operation. Though they still have to be robust beyond what simple temporary fasteners can provide. 

They will for sure have a calibration verification process that can be run to make sure it's all ok. The more common reason the fabs shutdown is the supporting utilities get turned off like water supply, because they have to check the pipes and make sure they aren't ruptured. Same with all the robotic overhead transport, gotta check absolutely everything.

[Mark Kaine]

4 hours ago, Forbidden Wafer said:

Daaaamn. Hope taiwanese friends are safe.

On a different note, everytime I see a borked building like this I wonder why aren't most buildings built on top of a radier (concrete slab). I know it is more expensive than deep foundations for large buildings, but I still prefer a raft over something that can get sheared by differential settlement. 

hey, i don't know but I think a "swimming" skyscraper or something might be even more devastating than one that just collapses in itself.

[leadeater]

4 hours ago, Forbidden Wafer said:

On a different note, everytime I see a borked building like this I wonder why aren't most buildings built on top of a radier (concrete slab). I know it is more expensive than deep foundations for large buildings, but I still prefer a raft over something that can get sheared by differential settlement. 

Because for even something as 'lightweight' as a residential house cause slumping of slab foundations, we just ignore that it happens because it's typically not a problem. A very large very tall building would simply topple over, no earth quake required.

 

Leaning Tower of FiDi

Quote

The building leans 3 inches (76 mm) to the north as a result of the method used to construct its foundation: instead of using the piling method like other neighboring skyscrapers, a soil improvement method were used where chemicals or other material are added to the soil to strengthen it. As of 2023, only half of the finishes, including windows, have been installed.

https://en.wikipedia.org/wiki/161_Maiden_Lane

 

There is a really good reason why it's done the way it is.

[Kisai]

6 hours ago, Forbidden Wafer said:

Daaaamn. Hope taiwanese friends are safe.

On a different note, everytime I see a borked building like this I wonder why aren't most buildings built on top of a radier (concrete slab). I know it is more expensive than deep foundations for large buildings, but I still prefer a raft over something that can get sheared by differential settlement. 

Asian countries often don't have zoning systems like US and Canada do, thus many times buildings are built to be disposable rather than repaired because it's cheaper to build, and they don't treat it as an investment.

 

In the case of Taiwan specifically, 50,000 buildings were damaged in 1999 due to lax building code. The building that collapsed "Uranus Building" was built in 1988. There was previously a 1986 7.4 earthquake in the same place.

https://archinect.com/news/article/147653406/taiwan-earthquake-tin-cans-found-as-fillers-may-have-caused-high-rise-to-collapse This is from two months ago.

 

Often buildings (see "tofu dreg projects" in PRC) collapse, or crumble because of corruption in the layers of the project management. When a building suffers catastrophic collapse, its' because something wasn't done right, not done to code, not inspected, made of inappropriate materials, or people leading the project just ran off with the money.

 

When you hear about cost-overruns in big projects, this likely is because the project scope creeped somewhere. Better materials required, things having to be excavated deeper, deeper piles, etc.

 

At any rate, to address the "concrete slab" comment. No. Any building larger than 4 stories has to be pile-driven into bedrock so it doesn't move and doesn't "fall over", this is why "tall skinny" towers are garbage, because they have to drive the piles down so deep and the foundation already has to be extremely big to support the structure.  In cities along the pacific coast, the towers need to have piles driven in order keep them anchored in the wind, and when you start getting into the 50-story heights, you need counter-weight systems to keep them from swaying in the wind.

 

Of course if we stopped making buildings with flat faces , the wind wouldn't be an issue. But people don't seem to want to live in buildings shaped like domes and silos.

[Forbidden Wafer]

4 hours ago, Kisai said:

At any rate, to address the "concrete slab" comment. No. Any building larger than 4 stories has to be pile-driven into bedrock so it doesn't move and doesn't "fall over"

... 

But these toppled buildings aren't even that tall. Some piles to stabilize the soil and the radier on top should make it work. 

 

It's not all of them, but I know multiple buildings that are 20-30 floors near the beach, constructed on top of sand, that don't have foundations that go to the bedrock. They use very large feet's and some floaters below them, as far as I can remember. Some others use just a lot of piles, but not that deep, and sometimes suffer with settlement and need to be  fixed.  

 

Of course, we don't have quakes over here, so may not be safe in those conditions. 

[CarlBar]

4 hours ago, Forbidden Wafer said:

But these toppled buildings aren't even that tall. Some piles to stabilize the soil and the radier on top should make it work. 

 

It's not all of them, but I know multiple buildings that are 20-30 floors near the beach, constructed on top of sand, that don't have foundations that go to the bedrock. They use very large feet's and some floaters below them, as far as I can remember. Some others use just a lot of piles, but not that deep, and sometimes suffer with settlement and need to be  fixed.  

 

Of course, we don't have quakes over here, so may not be safe in those conditions. 

 

Yeah most likely an earthquake would make that sand undergo liquefaction. Bye bye any ground support.

 

Though TBH most diagrams i've seen of tall tower construction have had the piles tied together at ground level by a slab anyway, the piles provide the support, but the slab distributes it around the building.

[leadeater]

1 hour ago, CarlBar said:

Yeah most likely an earthquake would make that sand undergo liquefaction. Bye bye any ground support.

 

Though TBH most diagrams i've seen of tall tower construction have had the piles tied together at ground level by a slab anyway, the piles provide the support, but the slab distributes it around the building.

It's sort of obvious, mostly when pointed out, but the piles even if not super deep or down to rock provide so much more support because the surface area interacting with the soil/ground is so much greater. Many many piles gown down far exceeds what a straight slab only would do and if a slab starts subsiding/slumping for any reason then it's compromised and gets non-linearly weaker and less supportive.

 

As you say sing slab foundation supports be piles is the standard far as I know, from what a lot of mega project documentaries of building sky scrapers where they explain all the whys etc. I also watch one on the NYC leaning tower and why that failed so badly and quickly.

 

Earthquakes are just bad and the magnitude often isn't so directly related to building damage because you also need to look at ground movement itself in the virtual and horizontal as well. Up down good, side to side badddd.

[Kisai]

5 hours ago, Forbidden Wafer said:

But these toppled buildings aren't even that tall. Some piles to stabilize the soil and the radier on top should make it work. 

 

It's not all of them, but I know multiple buildings that are 20-30 floors near the beach, constructed on top of sand, that don't have foundations that go to the bedrock. They use very large feet's and some floaters below them, as far as I can remember. Some others use just a lot of piles, but not that deep, and sometimes suffer with settlement and need to be  fixed.  

 

Of course, we don't have quakes over here, so may not be safe in those conditions. 

Very simple and to the point.

 

There's a dozen other videos from other geological survey's around the US/Canada that basically demonstrate liquefaction the same way.

 

Hence piles need to hit bedrock for the building to not sink. A building might have 10 stories of underground parking and be 50 stories tall, but if it's not sitting on bedrock, it has piles reaching several stories down beyond the ground floor. Likewise buildings built on reclaimed land/landfill need that.

 

 

[Uttamattamakin]

On 4/3/2024 at 5:28 PM, StDragon said:

Those machines have all sorts of calibration routines they run post installation. I wouldn't be surprised at all if ASML designed their equipment with robustness to withstand the journey of **shipping and earthquakes for any fabs located around the ring of fire.
 

I was referring to wafers that were being worked on as the ground beneath them was shaking violently.

 

13 hours ago, leadeater said:

It's sort of obvious, mostly when pointed out, but the piles even if not super deep or down to rock provide so much more support because the surface area interacting with the soil/ground is so much greater. Many many piles gown down far exceeds what a straight slab only would do and if a slab starts subsiding/slumping for any reason then it's compromised and gets non-linearly weaker and less supportive.

Very true.  Consider the skyscrapers of Downtown Chicago.  I don't think they are founded on bedrock because the bedrock here is very VERY far down.    The area of Chicago where the John Hancock building is, streeterville, was at the bottom of lake Michigan at first.  It started out when a boat, headed by a captain Streeter, ran aground.   Sand built up around it ... then it became a sort of island shantytown/ red light district.  Now it has skyscrapers with just the sort of deep foundations you describe. 

 

https://informedinfrastructure.com/31619/building-skyscrapers-on-chicagos-swampy-soil/

 

Don't know how well that would hold up if the New Madrid fault had a big 8.0 magnitude quake.   Something which hasn't happened since just before the war of 1812.   Lets just say I would not want to be on lower Wacker drive under all of that when the shaking happens. 

[SplAdam]