trust me, I know this all too well in a past life I use to design product "torture" tests and the biggest issue does the test match the real world. I use to have a saying I can design a test ever product will fail and one that every product passes, but in less they represent how the product is used it all meaning less. Engineer tent to test to make products fail but many time the failures are meaningless.
As I pointed out, I believe the real issue why display break has more to due with the stresses the display was put under before it was impacted. We use to do cumulative tests, individual products which have gone through a series of tests one right after the other and those products usually failed in different ways than the ones which only saw a single test (stress) condition.
This was all the issue I had with Apple on the whole voiding the warranty of the phone if they got wet. All the engineers live in Calif and never experience something we on the east cost experience ever winter. Many building are heated with Steam or Hot water so they tend to be hot and moist. If you are outside in the very cold and walk into the building and you electronics are still cool from being outside as so as they hit the warm moist air they are soaking wet, this is real world condition Apple never considered and most manufacturers never consider.
Exactly. When I used to design and manufacturer professional audio equipment, we used to do stress testing of various kinds. One was a shaker table. That was really nice, but I questioned the usefulness. True, in shipping, something can get shaken heavily, but it's a rarity that it would cause failure, of shorten the life of the product, or even cause specs to change. When in use, the product is on a shelf, or in a rack.
So sometimes those tests look good, and cause the manufacturer to spend extra time and money to get the product to pass those tests, but if the test doesn't relresent real world use, what the point?
And as you mention, they can fail to do tests that do matter. After all, engineers are sometimes human too, they make mistakes.
If we talk about stress as being a point of failure, then we have to aske if the device, or part is being stressed last the point of no return, or if whether that stress can be repeated almost infinitely. So bending a part less than its ability to return to the same initial condition, may not be a problem no matter how many times it's done. But bend slightly past that point may cause it to fail rapidly. We see this particularly with metals, ending up in stress cracking.
I don't see anywhere that Corning claims the glass is "harder", as AI's headline suggests. Did I miss that? Yes, it's stronger and more resistant to drop damage, but is it actually "harder"?
Actually, dropping a phone held next to your ear would seem to be unlikely. More common, I'd think, would be the phone slipping out of a pocket, or falling to the ground when you tried to put it into or take it out off a pocket, or falling off a table, or falling off your lap... all of which meet the 1-meter (3-ft) criteria quite nicely.
Seen many times: person (let's say, a lady) tries to phone (say, with grandma) with phone stuck between face and shoulder, while holding a thing (say, a pan), in a hand and doing something else (say, open a cupboard) with the other. Then the phone drops on the kitchen's floor and the glass shatters. And to top it off, the husband's dinner is burnt.
Yes, I know my example is delightful. I love it too.
Intel is not just claiming that 10nm will be next, they're claiming 7nm after that, and even 5nm! But the experts in that field aren't too sure that even 7nm is possible, much less 5.
On the nanometer scale, it does indeed appear we're approaching zero.
Fortunately, the SI provides another prefix: picometer. Which handily solves the problem, because now your sentence reads "not just claiming that 10,000 pm will be next, they're claiming 7,000 pm after that, and even 5,000 pm!" Clearly we'd have 4,999 pm to go before we actually reach zero!
Except... the SI provides yet another prefix: femptometer. And then attometer, and zeptometer, and even yoctometer.
On the one hand, I recoil at the absurdity of even claiming to measure such distances. However, a century ago our current technologies would have produced the same response. So... time will tell.
Fortunately, the SI provides another prefix: picometer. Which handily solves the problem, because now your sentence reads "not just claiming that 10,000 pm will be next, they're claiming 7,000 pm after that, and even 5,000 pm!" Clearly we'd have 4,999 pm to go before we actually reach zero!
Intel is not just claiming that 10nm will be next, they're claiming 7nm after that, and even 5nm! But the experts in that field aren't too sure that even 7nm is possible, much less 5.
On the nanometer scale, it does indeed appear we're approaching zero.
Fortunately, the SI provides another prefix: picometer. Which handily solves the problem, because now your sentence reads "not just claiming that 10,000 pm will be next, they're claiming 7,000 pm after that, and even 5,000 pm!" Clearly we'd have 4,999 pm to go before we actually reach zero!
Except... the SI provides yet another prefix: femptometer. And then attometer, and zeptometer, and even yoctometer.
On the one hand, I recoil at the absurdity of even claiming to measure such distances. However, a century ago our current technologies would have produced the same response. So... time will tell.
I don't think the units were ever the issue, but rather the indivisibilty of the building blocks and the observation that, with a 0.5 nm atomic spacing, we are approaching the physical limits of circuit miniaturization.
Comments
Exactly. When I used to design and manufacturer professional audio equipment, we used to do stress testing of various kinds. One was a shaker table. That was really nice, but I questioned the usefulness. True, in shipping, something can get shaken heavily, but it's a rarity that it would cause failure, of shorten the life of the product, or even cause specs to change. When in use, the product is on a shelf, or in a rack.
So sometimes those tests look good, and cause the manufacturer to spend extra time and money to get the product to pass those tests, but if the test doesn't relresent real world use, what the point?
And as you mention, they can fail to do tests that do matter. After all, engineers are sometimes human too, they make mistakes.
If we talk about stress as being a point of failure, then we have to aske if the device, or part is being stressed last the point of no return, or if whether that stress can be repeated almost infinitely. So bending a part less than its ability to return to the same initial condition, may not be a problem no matter how many times it's done. But bend slightly past that point may cause it to fail rapidly. We see this particularly with metals, ending up in stress cracking.
http://news.rice.edu/2012/09/20/glass-half-full-double-strength-glass-may-be-within-reach/
I don't see anywhere that Corning claims the glass is "harder", as AI's headline suggests. Did I miss that? Yes, it's stronger and more resistant to drop damage, but is it actually "harder"?
Actually, dropping a phone held next to your ear would seem to be unlikely. More common, I'd think, would be the phone slipping out of a pocket, or falling to the ground when you tried to put it into or take it out off a pocket, or falling off a table, or falling off your lap... all of which meet the 1-meter (3-ft) criteria quite nicely.
Seen many times: person (let's say, a lady) tries to phone (say, with grandma) with phone stuck between face and shoulder, while holding a thing (say, a pan), in a hand and doing something else (say, open a cupboard) with the other. Then the phone drops on the kitchen's floor and the glass shatters. And to top it off, the husband's dinner is burnt.
Yes, I know my example is delightful. I love it too.
Actually I find it easier to let the gorillas play wherever the hell they please. Damn dirty apes!
But but... gorillas aren't native to America.
Native to the America's?
Yes, definitely.
Intel is not just claiming that 10nm will be next, they're claiming 7nm after that, and even 5nm! But the experts in that field aren't too sure that even 7nm is possible, much less 5.
On the nanometer scale, it does indeed appear we're approaching zero.
Fortunately, the SI provides another prefix: picometer. Which handily solves the problem, because now your sentence reads "not just claiming that 10,000 pm will be next, they're claiming 7,000 pm after that, and even 5,000 pm!" Clearly we'd have 4,999 pm to go before we actually reach zero!
Except... the SI provides yet another prefix: femptometer. And then attometer, and zeptometer, and even yoctometer.
On the one hand, I recoil at the absurdity of even claiming to measure such distances. However, a century ago our current technologies would have produced the same response. So... time will tell.
Fortunately, the SI provides another prefix: picometer. Which handily solves the problem, because now your sentence reads "not just claiming that 10,000 pm will be next, they're claiming 7,000 pm after that, and even 5,000 pm!" Clearly we'd have 4,999 pm to go before we actually reach zero!
Also.... Isn't INTEL the "world expert"?
I don't think the units were ever the issue, but rather the indivisibilty of the building blocks and the observation that, with a 0.5 nm atomic spacing, we are approaching the physical limits of circuit miniaturization.