worsening latencies and error rates in store for SSDs to come in the long run... but hopefully we'll all be onto using something else by then anyway...
http://www.tomshardware.com/news/NAN...upp,14728.html
Printable View
worsening latencies and error rates in store for SSDs to come in the long run... but hopefully we'll all be onto using something else by then anyway...
http://www.tomshardware.com/news/NAN...upp,14728.html
Well firstly if latencies do indeed double, that still leaves SSDs weeing over hard drives from a lofty height :)
Over that sort of time-scale there is a good chance that one of the other technologies like mram will take over from flash and give us a step change in performance & cost, but it would still be an SSD of sorts. Actually, form most people probably just improved cost will do them.
Well at the moment they are taking off
There are currently 1TB SSD units on the market, which don't experience any density issues (though they are very-very-expensive). At the current stage we are at most suppliers, and users, realize that SSDs aren't a large-scale storage solution. This area still belongs to traditional HDDs.
Also the article states "It's not going to be viable to go past 6.5-nm" and "feature capacities of ...16 TB"
Well we are currently at 3TB max (the highest i've seen, which cost sever £'000), and 25-nm. So theres a long-long way before we reach capacity, and frankly 16TB is more than enough for any conventional use (though storage requirements will undoubtedly rise).
But who's to say that this wont be different in 12 years time, any prediction that far ahead are bound to be rife with problems. Look back 12 years, where were we with storage back then? (Seriously where were we? I was only 10-years-old...). There's no way to tell what new technologies will come around, or for all we know there may be some break-through technology being worked on now...
Exactly - it'll be like someone whinging that we've hit the capacity for PATA transfer rates when SATA is available. It will happen, but something better will be around and everyone will move over to that.Quote:
Originally Posted by Samwood
A lot has changed in the last 10 years of computing, I highly doubt any potential issues we see with todays SSDs will still be relevant 12 years from now.
Reminds me of the "HDD's wont be able to pass X amount".
When there is demand, there is RnD.
What I find odd, is that while density may be a long term issue.....why keep making them all in thin 2.5" format? Lets have some the same size as 3.5" HDDs...you could easily cram 4 times as many chips in with the added bonus of increased bandwidth.
We'll be storing stuff on qubits by then.
Problem will be how to do you make a quantum disk drive portable.
Quote:
Originally Posted by shaithis
you mean like this? http://www.scan.co.uk/products/32gb-...-50mb-s-retail
TBH, I'd rather buy a 2.5" as it can be used in a laptop or desktop, Giving better flexibility.
surely permanent ram will take over sdd's
I think because when drives get that many flash chips, you are so far beyond what SATA can cope with that you really might as well stuff it on a PCIe card rather than waste most of your performance.Quote:
Originally Posted by shaithis
Only in embedded systems where the mindset has given battery backed ram for decades. Should make for much longer phone battery life for example.Quote:
Originally Posted by pmcx9
Problem with PCs is the mindset of "read off disk into ram" is very deeply engrained. Perhaps tablets will show us a way forward, but it will take a long time. In the mean time, we will get SSDs made out of permenant ram. They will be faster and more resilient than todays SSDs, people will be happy with that.