Toshiba and WD NAND Production Hit by Power Outage: 6 Exabytes Lost

This doesn&#x27;t make sense:<p>&gt;Toshiba Memory and Western Digital on Friday disclosed that an unexpected power outage in the Yokkaichi province in Japan on June 15 affected the manufacturing facilities that are jointly operated. &#x2F;&#x2F;<p>Surely that&#x27;s not the reason, it would have to be &quot;and local [backup] power failed, and the failovers for that failed too&quot;??<p>Toshiba manufacture generators too, it&#x27;s not like they&#x27;d need to go far to get backup power designed for them.<p>There must be more to this? (Which explains why people are assuming it&#x27;s suspicious, I guess; and this site is making 35% of global NAND output).<p>FWIW, I hadn&#x27;t realised that it takes ~2months to process a wafer in to a chip.

Here&#x27;s an article from one month ago discussing the over-supply of NAND and DRAM (and the effect it has on pricing): <a href="https:&#x2F;&#x2F;www.forbes.com&#x2F;sites&#x2F;tomcoughlin&#x2F;2019&#x2F;05&#x2F;25&#x2F;nand-dram-supply-and-pricing" rel="nofollow">https:&#x2F;&#x2F;www.forbes.com&#x2F;sites&#x2F;tomcoughlin&#x2F;2019&#x2F;05&#x2F;25&#x2F;nand-dra...</a><p>I can&#x27;t help but feel very skeptical about the timing of this event, given the history of price-fixing in the industry.

Important to quote from the comment section<p>&gt;Five fabs and an R&amp;D center, outage was after the batteries also ran out.<p>For perspective, the batteries at GF&#x27;s leading fab can run the 1&#x2F;3 of the systems for only a few minutes. That&#x27;s the scale we&#x27;re dealing with.<p>I think before we do all sort of conspiracy theory, we need to look into reason for why was there an outage in Yokkaichi.

I would be most grateful if someone could please explain what sort of tools are likely to be used here, and why a power loss to those tools would ruin days&#x2F;weeks&#x2F;months worth of output relative to the time they were offline?

I recall the story of Micron&#x27;s first Chinese fab: 1 millisecond out phase brownout and they loose few megabucks instantly, and like that during every electrical event.<p>Giant UPSes are not an option in the industry because fabs eat oodles of electricity, and it is cheaper to loose a megabuck once a year than build a stabilisation&#x2F;ups plant

So... does that mean they&#x27;ll be hiking NAND prices, just like with HDD prices after the Thai floods?

First time I have to really think about Exa&lt;unit&gt;.<p><pre><code> Giga &#x2F; Tera &#x2F; Peta &#x2F; Exa </code></pre> 6 Millions Terabytes of solid state memory.. quite a mass.

I know that NAND involves no exotic raw materials, so does that enable them to recycle any of the damaged&#x2F;lost wafers? I don&#x27;t know very much about the physical processing&#x2F;preparation of the raw silicon and such that goes into making a wafer, could you simply grind up or perhaps chemically dissolve everything back to base components and re-create a fresh wafer?

I wonder what failed in their redundant power supply because they surely have something.<p>I hope the postmortem will be public !

This seems weird that 13 minute outage can kill month and a half production.<p>I wonder if this is standat hi-tech factory process reliability.

Conspiracy: this is how the NSA buys their disk space.

The fragility of the supply chain.

The stock of these companies really took a hit. Sarcasm.

Every time I see Godzilla he’s ensnared in and tearing down power lines. This was bound to happen sooner or later.