SSD flash... can we not fit more on there?

[KemoKa]

Okay, so you know how we have all these SSDs that are basically high-end NAND disks all in RAID 0? I was wondering why we can't have much much smaller NAND chips like on these:

I know that SSDs have much higher-tier flash, but these tiny USBs can fit 64GB in a space smaller than the actual bus, so why aren't we doing this for SSDs and having massive arrays of these disks? Is is because the manufacturing process for the chips is second-rate? Is it the addition of points of failure? is it controller complications?

Just wondering...

[79wjd]

We can. just look at an M.2 drive. The issue is the amount of memory that the controller can control. 

 

But theres also the fact that the demand for anything more than 1tb is very small; and even 1tb doesn't have a particularly large demand. 

[LukaP]

we can. too expensive. plus everyone is moving to pcie now. make a 10TB pcie ssd and try to sell it to consumers.. nope

[Snickerzz]

we can. too expensive. plus everyone is moving to pcie now. make a 10TB pcie ssd and try to sell it to consumers.. nope

make a 10tb hdd and most wont buy that either

[SMURG]

What if I told you SSDs are only that size and shape to fit the mounting holes for 2.5" HDDs?

 

make a 10tb hdd and most wont buy that either

Really good point. I can't even see the point in 6TB drives for consumers, and I need more storage than most. When RAID is an option it's almost always better than a single-drive configuration anyway so the only people who are gonna be better off buying 6TB drives are gonna be people who are gonna RAID them for tens of TB of storage.

[Guest]

What if I told you SSDs are only that size and shape to fit the mounting holes for 2.5" HDDs?

 

Really good point. I can't even see the point in 6TB drives for consumers, and I need more storage than most. When RAID is an option it's almost always better than a single-drive configuration anyway so the only people who are gonna be better off buying 6TB drives are gonna be people who are gonna RAID them for tens of TB of storage.

 

I guess it's a case of "if anybody needs more than 2TB of storage they're probably going for a raid array anyway". If you need/want that much storage you probably value the data on it, and in that case you're going for redundancy. Those "prosumers" also have the space for a few more harddrives, they're not in an environment where you need the extreme density. Cheap 1TB SSDs would be amazing, but any more than that will probably stay server grade stuff for quite some time IMO. Enclosures that Raid-0 two M.2 SSDs will be pretty nice for notebooks, but you might as well get 2 normal SSDs for a desktop.

[IdeaStormer]

I can one up you, this is the smallest I know of and you can add as many as you have.

 

 

Or if you only need 4:

 

[hojnikb]

As it was pointed out before, the issue lies in the controller itself, as most consumer grade controllers can't adress more than 1TB, as it either reqires even more dram or bigger block size. So we're stuck with 1TB for now, until controller maker don't up that limit. Flash itself is very dense, as one can get 1TB in a mSATA aswell, which translated to only 4 packages or chips. So really, one could easily pack 4TB of flash on to 2.5" pcb, but thats just isn't possible with current controllers and wouldn't make much sense either, as it would be too expensive for most consumers to buy. So bigger sizes are reserved for enterprise for now (for example sandisk offers 4TB on SAS interface).