Toshiba Celebrates Quarter-Century NAND Flash Anniversary

[exfileme]

NAND flash turns 25-years-old this year, and Toshiba wants to celebrate.

Toshiba Celebrates Quarter-Century NAND Flash Anniversary : Read more

 

[Plasmid]

Good now lower the prices

 

[jrharbort]

[citation][nom]Plasmid[/nom]Good now lower the prices[/citation]
Patience, young one.

http://www.tomshardware.com/news/ssd-hdd-solid-state-drive-hard-disk-drive-prices,14336.html

 

[alidan]

[citation][nom]jrharbort[/nom]Patience, young one.http://www.tomshardware.com/news/s [...] 14336.html[/citation]
that number is complete bs.

the reason solid state is falling fast is because there is a market for it, be it mobile devices or desktops. harddrives had to innovate, and get better over the years, solid state was more or less invented into processes that were already on track...

unless they come up with a way to fit more data per square mm besides nm shrinks, there is a base cost of about 50000$ a wafer, and about 21 tb a wafer, granted i REALLY want the size of a ssd chip (waffer, not the black shell) and how much space it stores so i can tell better how much ssds really cost, but there is a base cost that will not go down, and the only way to get it down... well... intel made a 450 wafer plant correct? if a 450 costs the same as a 300 to make, than thats one way to trim costs, the other is nm shrink, which even if we get to the 5nm range, they are JUST able to start competing with the prices pre flood, assuming the tech takes off, we may have 60tb drives around the same time for 2-300$, which is far cheaper than they can offer.

i got off on a tangent there, but point stands, ssds will never be cheaper than a hdd, and we will forever need more space... just watch what happens if xbox gets a bluray drive, you will see many games take up 25-50gb for no reason other than no compression.

 

[blazorthon]

[citation][nom]alidan[/nom]that number is complete bs.the reason solid state is falling fast is because there is a market for it, be it mobile devices or desktops. harddrives had to innovate, and get better over the years, solid state was more or less invented into processes that were already on track...unless they come up with a way to fit more data per square mm besides nm shrinks, there is a base cost of about 50000$ a wafer, and about 21 tb a wafer, granted i REALLY want the size of a ssd chip (waffer, not the black shell) and how much space it stores so i can tell better how much ssds really cost, but there is a base cost that will not go down, and the only way to get it down... well... intel made a 450 wafer plant correct? if a 450 costs the same as a 300 to make, than thats one way to trim costs, the other is nm shrink, which even if we get to the 5nm range, they are JUST able to start competing with the prices pre flood, assuming the tech takes off, we may have 60tb drives around the same time for 2-300$, which is far cheaper than they can offer. i got off on a tangent there, but point stands, ssds will never be cheaper than a hdd, and we will forever need more space... just watch what happens if xbox gets a bluray drive, you will see many games take up 25-50gb for no reason other than no compression.[/citation]

Flash can't scale down to 5nm because it's insulators can't fit in that small of a process node. Basically it can't handle leakage at that node and would not work. It would need to be replaced by SONOS/SHINOS to get insulators that can get thinner than Flash's insulators and still work. Even then, I don't know if they can get that small for 5nm.

[citation][nom]naomiquexi[/nom]harddrives had to innovate, and get better over the years, solid state was more or less invented into processes that were already on track...[/citation]

At least we know ways to keep on with the increasing of hard drive data density and thus capacity and performance. With Flash, we know it's limited and needs replacement because there's little that we can do about it besides replacement with another technology, even if it's one that is very similar such as SONOS and SHINOS (they're basically Flash, except made from different materials, if I remember correctly).

 

[eddieroolz]

Didn't know Toshiba was the one that invented a NAND chip. Thanks Toshiba!

 

[alidan]

[citation][nom]blazorthon[/nom]Flash can't scale down to 5nm because it's insulators can't fit in that small of a process node. Basically it can't handle leakage at that node and would not work. It would need to be replaced by SONOS/SHINOS to get insulators that can get thinner than Flash's insulators and still work. Even then, I don't know if they can get that small for 5nm.At least we know ways to keep on with the increasing of hard drive data density and thus capacity and performance. With Flash, we know it's limited and needs replacement because there's little that we can do about it besides replacement with another technology, even if it's one that is very similar such as SONOS and SHINOS (they're basically Flash, except made from different materials, if I remember correctly).[/citation]

i know, the real wall is something like 6 or 7nm, but i did the math down to 5nm to make it easier on me, and even than they just barely were able to match the cost gb.

well there are things you can do to increase flash density, and that would be making the chips 3d... if they can make the layering take less time than a full new wafer, it would drive cost down by a fair amount. hell layering it would actually make ssd competitive with hdds in size, and possibly cost per gb if it can be done fast enough.

but i never take that into account when i run numbers, i just take whats actually happening into account.

 

[blazorthon]

[citation][nom]alidan[/nom]i know, the real wall is something like 6 or 7nm, but i did the math down to 5nm to make it easier on me, and even than they just barely were able to match the cost gb. well there are things you can do to increase flash density, and that would be making the chips 3d... if they can make the layering take less time than a full new wafer, it would drive cost down by a fair amount. hell layering it would actually make ssd competitive with hdds in size, and possibly cost per gb if it can be done fast enough.but i never take that into account when i run numbers, i just take whats actually happening into account.[/citation]

The only problem that I see with die/chip stacking like that is power usage. SSDs are known to be low power devices, but how much power will one use if it has four times (or any other reasonable number) more power sucking silicon electronics?

 

[alidan]

[citation][nom]blazorthon[/nom]The only problem that I see with die/chip stacking like that is power usage. SSDs are known to be low power devices, but how much power will one use if it has four times (or any other reasonable number) more power sucking silicon electronics?[/citation]

well... does a ssd constantly need full power to everything, or is that more of a just when its read and written? at least from my understanding the power usage comes from the controller more than the actual space itself.

here
http://www.newegg.com/Product/Product.aspx?Item=N82E16820148527
that is a 512gb drive
active power usage on it is 0.28W

even if power scaled with the drive, it would take a 10tb drive to rival today's hdds in power usage.

 

[blazorthon]

[citation][nom]alidan[/nom]well... does a ssd constantly need full power to everything, or is that more of a just when its read and written? at least from my understanding the power usage comes from the controller more than the actual space itself.herehttp://www.newegg.com/Product/Prod [...] 6820148527that is a 512gb driveactive power usage on it is 0.28Weven if power scaled with the drive, it would take a 10tb drive to rival today's hdds in power usage.[/citation]

Ahh, good point.