Are SSDs Doomed? Researchers Say "Yes"

I once had a teacher. He once told me that no matter the problem, to trust in the collective ability of mankind to solve whatever issue is most pressing. Imagine any problem that seems insurmountable and visualize the collective intelligence of the entire world and focus it at one specific point - if it is solvable we will find the answer. The only caveat is "now let there be profit."
 
are they saying that SSDs are going to get SLOWER as they become more capacitous (not sure if that's a word) or just that the speeds won't be increasing proportionately with the capacity?

the former seems a little ridiculous...
 
Granted I have no clue how they're stacked, but seems like you make a circuit board with chips, test it, then stack multiple numbers of those circuit boards all connected, essentially multiple SSDs inside of a single case you'll get get what you need. Or go more horizontal, why does it have to be a 2.5" architecture? why not a 3.5" like a normal HD? or why not a fatty like a 5.25" after all how many DVD/BR drives do you need?

I get it, making things tiny is big, but if it's going to fit inside of a computer case what's the big deal?

There are 3.5" drives. I know OCZ makes one, that's essentially is two SSDs in raid. Who knows why everyone targets 2.5" however. And the chips are stacked vertically. The most I've heard of is 16, but that hasn't made it past the lab, probably due to the issues I stated earlier. I'd imagine testing each layer before building the next would probably be more expensive than just doing say four stacks always and relying on die shrinks.
 
Who knows why everyone targets 2.5" however.
There are a few reasons.
-Laptop and some ultra high end server drives use the 2.5" form factor. Everything else is 3.5". It is a lot easier to design for 1 form factor and scale up from there using adapters and brackets. This keeps costs down for the hard drive manufacturer by keeping designs simple since they can make a "one size fits all" product.
-They could produce large capacity SSD's in 3.5" form factor, but there is not enough demand for them due to high costs (which are coming down now).
-The memory manufacturers have been shrinking the manufacturing process extremely quickly as of late, at least on the memory side of things where the transistors and chips are easier to design and produce. Producing a large capacity drive in 3.5" form factor only to have the same capacity in 2.5" form factor a year later is a waste of R&D resources at the moment.


As demand for SSD's and cost to manufacture them goes down, I have no doubt we will start seeing more 3.5" large capacity drives.
 
So SSDs only have 12-15 years left in their lifecycle -- of improvements in size and speed -- before they reach their reasonable technological limitations yet are already the fastest devices on the market. 'Oh Darn'. So what, SSDs will tap out at the mere 1TB/sec transfer speeds and maybe 2-4TB drive sizes in 12 years? *Queue sarcastically Darth Vader NOOOOOOOOOOOOOOOOOOO sound effect*.

I'm certain memresistors technology will be commercially developed and starting to show up on the shelf around that time to keep consumers buying new technology. After all, I don't think any hard drive manufacturers want people not to have a reason to buy new hard drives. If they can't improve the size or speed and the old drive still works, most people wouldn't upgrade. Hard Drive manufacturers will just shift over to memristors in 12 years time if not sooner if the technology proves reliable and cheap to manufacture.

Besides, mechanical hard drives had very slow improvements in speed compared to what SSDs have achieved in a few short years. During the 70s, 80s, 90s, and early 2000s and we didn't really complain much. I remember the first SSDs having an amazing! 80MB/sec transfer speed and now top of the line SSDs are up to 350-400MB/sec after a few years of development. Still pretty incredible considering how it tooks us decades with traditional hard drives to progress from 5MB/sec to 40MB/sec. Even then, traditional HDDs still had significantly longer seek times and less operations per second.
 
The trick is to make sure they stay nice and cool.

A lot of drives that come in OEM computers have issues because they don't keep the drives and other components cool. Their theory is that as long as they stay within "operating temps" there shouldn't be a problem.

Google really had a good set of data from their data centers about causes of hard drive failure.
linky
 
Ha, like old fashioned spinning Magnetic disc's will beat it? I doubt it very much!
 
If that really is an issue and there's no way to circumvent it, then I don't see why they can't just stick to making them an optimal size. Technology does not HAVE to keep going up. Floppies went to 1.44 and stayed that way for a long time. CDs stayed at 700MB for a while, etc...

I don't see SSDs as a replacement for HDDs, instead I see them more as a second option, if practical for that specific situation. They make a great OS drive, but perhaps not a great mass storage drive. I really tihnk SSDs and HDDs can both coexist. They have their own best use scenarios.

In reality, I'm sure they'll find a way to circumvent this anyway. I remember at one point reading that they could no longer make HDD density smaller, otherwise the magnetic fields would interfere with each other given the size of the bits. That was around the time 120GB drives were the biggest you can get.

What would really be a breakthrough is if they can find a technology to make something as fast as a SSD, but have unlimited read/write operations like HDDs. But even as it stands now, SSDs can get TONS of writes to them, so they do fine as an OS drive. Keep HDDs in raid for large scale high I/O data and you get the best of both worlds.
 
THIS JUST IN:

NEW TECHNOLOGY will REPLACE OLD TECHNOLOGY in the FUTURE!!!

Scientists in white lab coats say, "We don't know how to create future technology yet. Right now, we're stuck with technology from the present, but in the future, we'll have technology from the future. However, we expect real engineers to run into issues with SSDs in the future. At that time, they will either have to solve the issues or develop new technology."
 
I think I will be doing just fine in 2024 with my 132TB* 2.5" SSD. Doubt I will be able to use all that storage anyway since media files probably won't be getting larger and they use the vast majority of my storage space currently.

*(Assuming storage capacity doubles every 18 months starting with current base of 512GB).
 
Memristors will take over in 5 years.

[Most recently discussed technology with a demonstratable proof of concept] will take over in [random number between 1 and 20] years.

I called it here folks.

Good thing you did, else it wouldn't be happening!

Pfft. You can call it here all you want but you're not even within the first million to say Memresistor is going to be the next tech.

You feel special now?
 
What I'm worried about with NAND memory is the write endurance. With 34nm flash the write endurance was 5,000 cycles, with 25nm we're now at 3,000 cycles. It's going to get to a point where SSDs will have to be disposable devices (at disposable prices), or they will have to stop using smaller manufacturing processes and instead stick to a larger one, then optimize it to lower costs.

One thing's for certain, we're going to be seeing lots of stuff happen in this segment.
 
THIS JUST IN:

NEW TECHNOLOGY will REPLACE OLD TECHNOLOGY in the FUTURE!!!

Scientists in white lab coats say, "We don't know how to create future technology yet. Right now, we're stuck with technology from the present, but in the future, we'll have technology from the future. However, we expect real engineers to run into issues with SSDs in the future. At that time, they will either have to solve the issues or develop new technology."

That sounds like it could make awesome breaking news on a site like The Onion.:D
 
... since media files probably won't be getting larger and they use the vast majority of my storage space currently.

I would think they will get larger with 4K Ultra-HD displays being sold this year at the highest-end models, higher-end models next year, mid-end models the year after and low end models probably two more years after. 3D movies are also a bit larger than 2D movies.

This year alone, there's been something like 15-3D movies released in theatres already and 3D is here to stay I'm afraid this time. It seems every 2-3 weeks a new 3D title is released and almost every local theatre I know of has 3D screens. It'll be offical 3D is here to stay once we see something like 'The Walking Dead Season 3 or 4 - "Now in 3D" or 'Criminal Minds 3D'. IE Major large-budget syndicated TV shows shot in 3D.

So I'd say media files will get larger over time. Isn't standard definition suppose to go the way of the dinosaur sometime in early next decade where even if you liked downloading sd rips, you might be forced to dl hd rips for new content?
 
@Arentol: Also, larger media files will probably be needed/created by digital cameras. D-SLRs and even regular cameras are starting to record videos in 1080p. It won't be long till we see these cameras that are already capable of taking burst shots at up to 20+MP start to record 4K ultra-hd video since the sensors are clearly somewhat capable of receiving that much data. The real challenge would just be improving the sensor to work to store it real-time but will certinately happen before 2024 ;)
 
Like the lack of mechanical parts with SSD, but honestly I see no appreciable increase in speed with this thing. I have a somewhat early model Intel 40g think SLC maybe MLC, but between loading my USB devices and other necessary drivers my boot speed is only slightly increased.

SSDs certainly have a large amount of development to be done upon which previous posters have addressed, but for now I stick to good ol' high density mechanical drives.
 
Memristors will take over in 5 years. I called it here folks.

That was called 5 years ago anyway...and they are late with it.
I say single atom transistors will increase ssd speed and density at least 100 times in the next 5 years.
 
what happened to using Diamonds as storage when they found out they could make them in alot less time...
 
Back
Top