News Western Digital and Kioxia Report 3D NAND Contamination, Impacts at Least 6.5 Exabytes

[Admin]

Kioxia and Western Digital reported that two factories have suffered from contamination issues that scuttled at least 6.5 exabytes of flash memory.

Western Digital and Kioxia Report 3D NAND Contamination, Impacts at Least 6.5 Exabytes : Read more

 

[JarredWaltonGPU]

6.5 EB of NAND. That's a lot of SSDs! That's basically 6.5 million 1TB SSDs, though obviously it wouldn't have been purely for 1TB models. Glad I don't need another SSD right now!

 

[Endymio]

Kioxia currently shipped [sic] 8.5% of the words [sic] SSD capacity in 2021, while Western Digitial [sic] accounted for 15.4%.

Editor needed on Aisle 5, stat!

 

[Alvar "Miles" Udell]

Sounds to me like it's probably a water contamination issue that's reducing the fab's ability to produce the wafers rather than existing wafers being ruined. Probably a large overestimation of the potential impact for legal purposes though.

 

[tennis2]

Ya know, we haven't had a catastrophic event (fire, blackout, flood, etc) recently to drive up storage costs. SSD prices have been decreasing to "wow that's super cheap!" levels. I guess we were due.

These things are so predictable.

 

[Scott Kay]

Sounds to me like it's probably a water contamination issue that's reducing the fab's ability to produce the wafers rather than existing wafers being ruined. Probably a large overestimation of the potential impact for legal purposes though.

That's an interesting take. I initially assumed contaminated manufactured product or source material. Never thought of water contamination as a potential factor. Did you have any additional information leading you to believe the problem was related to the water supply for the fab?

 

[Endymio]

Ya know, we haven't had a catastrophic event (fire, blackout, flood, etc) recently to drive up storage costs. SSD prices have been decreasing to "wow that's super cheap!" levels. I guess we were due.

These things are so predictable.

Are you suggesting they intentionally sabotaged several months of their own production, to create a temporary price increase that their competitors will primarily benefit from, due to that same lack of production?

 

[tennis2]

Are you suggesting they intentionally sabotaged several months of their own production, to create a temporary price increase that their competitors will primarily benefit from, due to that same lack of production?

Just observing that these events rarely seem to happen when storage prices are high....

 

[William Pyke]

6.5 EB of NAND. That's a lot of SSDs! That's basically 6.5 million 1TB SSDs, though obviously it wouldn't have been purely for 1TB models. Glad I don't need another SSD right now!

Wouldn't that be 65 million 1TB SSDs? Googling 6.5 Exabytes to terabytes brings up 6.5e+7 terabytes, and using wolframalpha.com 6.5e+7 is 65 000 000.

 

[Alvar "Miles" Udell]

Ya know, we haven't had a catastrophic event (fire, blackout, flood, etc) recently to drive up storage costs. SSD prices have been decreasing to "wow that's super cheap!" levels. I guess we were due.

These things are so predictable.

Except for Samsung's fab shutdown in Texas, the drought in Taiwan...

 

[Endymio]

Wouldn't that be 65 million 1TB SSDs? Googling 6.5 Exabytes to terabytes brings up 6.5e+7 terabytes, and using wolframalpha.com 6.5e+7 is 65 000 000.

No. A petabyte is 1,000 terabytes, and an exabyte is 1M terabytes. The nice thing about the SI prefixes is that you can perform all such conversions in your head ... no "wolframalpha" needed.

 

[William Pyke]

No. A petabyte is 1,000 terabytes, and an exabyte is 1M terabytes. The nice thing about the SI prefixes is that you can perform all such conversions in your head ... no "wolframalpha" needed.

Thanks for the clarification!

 

[thelastwinchester]

That's barely any significant impact on total percentage of SSDs, but hey anything to drive prices up.

 

[JarredWaltonGPU]

That's barely any significant impact on total percentage of SSDs, but hey anything to drive prices up.

I believe something like 200 EB of consumer SSDs were shipped last year, so this would represent around 3% of total SSD shipments. Or maybe that was just in one quarter? But there’s a cascading effect so it will certainly impact prices for the next quarter or three.

 

[thelastwinchester]

I believe something like 200 EB of consumer SSDs were shipped last year, so this would represent around 3% of total SSD shipments. Or maybe that was just in one quarter? But there’s a cascading effect so it will certainly impact prices for the next quarter or three.

Lets not forget that capacity increases every year over increasing demand which they take into account and will provision extra for it. So, at worst this should make the SSDs stay at current prices.

 

[Icu812]

I'm guessing the QC department has been fired or where they the first victim of the "pandemic".

 

[helper800]

I believe something like 200 EB of consumer SSDs were shipped last year, so this would represent around 3% of total SSD shipments. Or maybe that was just in one quarter? But there’s a cascading effect so it will certainly impact prices for the next quarter or three.

I feel like even if it did not actually contribute to increased costs for the manufacturers they would all increase prices anyways because the public perception is that it may for a quick money grab.

 

[Endymio]

I feel like even if it did not actually contribute to increased costs for the manufacturers they would all increase prices anyways because the public perception is that it may for a quick money grab.

It doesn't matter whether or not the producers' costs increased: if it reduced supply, then prices must rise to compensate.

 

[helper800]

It doesn't matter whether or not the producers' costs increased: if it reduced supply, then prices must rise to compensate.

Its not that they must rise, its that they rather just do because the demand placed on it has not moved. I was suggesting something more sinister like collusion and market manipulation. Like the people who actually change the prices on the products were told to do so by an amount above and beyond compensation for the reduced supply.

 

[Endymio]

Its not that they must rise, its that they rather just do because the demand placed on it has not moved.

The demand curve doesn't shift, but the supply curve does. Either the price rises, or the shelves run dry: basic economics.

I was suggesting something more sinister like collusion and market manipulation. Like the people who actually change the prices on the products were told to do so by an amount above and beyond compensation for the reduced supply.

There's an entire host of misconceptions packed into that viewpoint, starting with the belief that producers somehow magically know how much they "should" change prices in response. Any rational producer always charges the price that maximizes their own profits. If that price is too high, for any reason whatsoever, they'll wind up with an unsold surplus, and be forced to lower the price regardless.

 

[Deleted member 14196]

So stick with Samsung huh

 

[helper800]

The demand curve doesn't shift, but the supply curve does. Either the price rises, or the shelves run dry: basic economics.

There's an entire host of misconceptions packed into that viewpoint, starting with the belief that producers somehow magically know how much they "should" change prices in response. Any rational producer always charges the price that maximizes their own profits. If that price is too high, for any reason whatsoever, they'll wind up with an unsold surplus, and be forced to lower the price regardless.

1. This is essentially what I meant to say. I made a distinction that its not that they must rise but rather they do naturally. The price rising is not mandated. Prices either do or the item becomes unavaliable.
2. Agreed, however, I suspect that because news information spreads so rapidly there comes a public expectation of rising prices which makes produces believe they can raise the price beyond what was required to correct for the supply loss.