LinkBack URL
About LinkBacksRead more.Memory specialist OCZ confirms UK pricing and availability of its price-busting new solid state disk range.
Read more.Memory specialist OCZ confirms UK pricing and availability of its price-busting new solid state disk range.
On play pricing is: £117.99, £179.99 and £319.99
I guess this would make a rather good system drive..
Crosshair VIII Hero (WIFI), 3900x, 32GB DDR4, Many SSDs, EVGA FTW3 3090, Ethoo 719
Ack.. pretty high UK markup.Originally Posted by dangel
On play pricing is: £117.99, £179.99 and £319.99
Yeah.. very nice small database sytem/game server too.I guess this would make a rather good system drive..
Well it's better than the quoted: 32GiB - £125, 64GiB - £199 and 128GiB - £349 in the article. But yeah, USD conversion anomaly expected.
Crosshair VIII Hero (WIFI), 3900x, 32GB DDR4, Many SSDs, EVGA FTW3 3090, Ethoo 719
How long do devices like this actually last as main system drives? (What with the limited write cycles inherent in SSD technology).
____
(='.'=)
(")_(")
1.5 million hours.
http://www.ocztechnology.com/product...ata_ii_2_5-ssd
Crosshair VIII Hero (WIFI), 3900x, 32GB DDR4, Many SSDs, EVGA FTW3 3090, Ethoo 719
Hmmm... I'm quite suspicious of that figure... I understood that most SSDs had quite a limited number of write cycles before they couldn't write any more. I wonder what the actual conditions were for that MTBF because 1.5M hours works out to over 170 years!
(Or to put it another way: 170 year MTBF, what's with the 2 year warranty?)
____
(='.'=)
(")_(")
I reckon they'd last (now) as long as a physical drive and that's all they have to beat in those terms. The *probability* of failure is lower too - no moving parts etc.
Crosshair VIII Hero (WIFI), 3900x, 32GB DDR4, Many SSDs, EVGA FTW3 3090, Ethoo 719
The 32Gb version is tempting as a HTPC system drive, then having the larger magnetic drive set to power down after a short period of no use.
The price is still prohibitve for me, but they're getting close to level a would pay....maybe 6 months time I'll take another look at the SSD market.
That caught my attention, as well, and yet, they're still coming up in the news, with no one making a note of it.Hmmm... I'm quite suspicious of that figure... I understood that most SSDs had quite a limited number of write cycles before they couldn't write any more. I wonder what the actual conditions were for that MTBF because 1.5M hours works out to over 170 years!
(Or to put it another way: 170 year MTBF, what's with the 2 year warranty?
My understanding is that if you do the math, the average SSD may only last a year under normal use. I believe that many of them only have a few hundred thousand writes in them, and really good ones have somewhere over a million. I don't remember the figures off hand, or even where I saw it, but that would explain why a "newer and better" lower priced drive (I don't get the idea that these magically super affordable OCZ drives are the bleeding edge, no compromises, best of the best) would have a two year warranty.
And in fact, the previous line of much more expensive OCZ drives is rated at 2 million hours MTBF. Many cheap drives even have 4 million hours MTBF.
Not exactly an apples to apples comparison, but motherboards don't have moving parts, either, and yet, the cheap ones die like crazy. There are simple explanations for that, but that's very likely the case with varying levels of SSDs, as well.
Those MTBF figures don't come close to indicating how long the drive will function to begin with. Ever had a HD last a good while and then die after apparently being worn out? Something tells me it didn't take over 100 years if you have experienced this. It probably takes more like one year of total use (maybe you had it five years, but it definitely wasn't in use 24/7) in many cases.
At least as far as normal magnetic disk drives go, I'm under the impression that the MTBF actually refers to how long the disks themselves will last, assuming maintence (i.e. replacement) of the other parts. In other words, your data will last that long, but there is NO WAY the drive is actually going to function for even a significant fraction of that length. And that's why your drive can "die," but those data recovery services are pretty much guaranteed to be able to get it back for you.
So who knows what in the world they're referring to with the MTBF of SSDs? I've never seen it explained. They have no moving parts, but obviously, there are other issues with the longevity of SSDs, and I imagine the case is very similar to the MTBF rating of HDDs. Your DATA may survive that long, but without knowing the amount of write operations it is expected to have (which OCZ does not mention on their site), the possibility exists that the drive could cease functioning properly very quickly, as write operations, at least as a permanently limiting factor, are likely not taken into account.
I have no idea how these things really work. Maybe, in some far fetched way, they're serviceable in a manner that allows them to continue writing? It'd be nice if some site like Hexus did an in depth explanation, as SSDs are rapidly becoming a more viable outright replacement of HDDs. It's to the point now where people really need to know what's going on, and unfortunately, it doesn't appear that the manufacturers themselves are too quick to take the initiative of informing their would be customers.
Salazaar (03-07-2008)
I'd be very surprised if it was less than a year. The write cycle thing isn't per disk, but per cell. How often do you completely erase and rewrite every single byte of data on your disk?
Personally, I probably wouldn't put a heavily-used page file on one, but even that's pretty over the top from a write cycles perspective.
It's not like that as far as I know - MTBF is a measure of the drive functioning properly. However it's an artificial figure that's inflated upwards somewhat. If you test 10,000 drives and one of them fails after 5 hours then they put the MTBF at 50,000 hours. Obviously they're not saying every drive can do 50,000 hours - that's not even tested, but it makes sense in that case to have a warranty of say 10 hours so that you are covered from having the duff drive.
Is that really what they do? Because from a statistical point of view that is just so plain wrong it's almost funny! Very 'inventive' interpretation of the results I think...
Anyway, for an SSD fence-sitter like me this new range can only be good news - it's guaranteed to get a lot more people buying them and that will mean a whole load of real-world testing to draw on. Getting awfully close to affordable now too!
I can't remember the exact thing, but it wasn't far off the same principal.
I know the MTBF assumes "maintenance," which can mean pretty much anything they want it to. Their definition of "failure" is not what most peoples' is.
I think to say that the MTBF is, "inflated upwards somewhat," is just a wee bit of an understatement. It's not even really remotely accurate.
The problem is that they call it the MEAN time between failures, and that it's rated in HOURS OF USE. I imagine it's quite possible that some drive could last 170 years...if it were barely used, and was quite an exception to the AVERAGE. That sure as heck wouldn't be leaving it running for 170 years, either. But what they're saying is that that IS the average, which doesn't begin to make sense to me. If a drive were to die very quickly or be DOA, as definitely is not uncommon, even with the very best quality control, that would destroy their average. So are we to assume that some drives last a million years?
However in the world it is derived, it's almost a totally pointless figure, and even more so with SSDs. The only thing it MIGHT tell you is how much the build quality differs between various models by ONE manufacturer (as I figure is the case with these OCZ drives), but that's still not a remotely accurate way to gauge it, just a HUGE generalization, at the very best. They still could have fudged it, as the numbers are inherently fudged to begin with. Of course they're going to give more expensive drives a higher MTBF. How bad would it look if it were lower? But we'd never be able to know the difference.
All flash memory is not created equal!
There are 2 main types:
SLC and MLC (single and multi level cells) SLC stores one bit per cell, MLC doubles it.
SLC is faster, and has around 1-5 million write cycles ( similar to hard disks).
MLC is slower and cheaper but less reliable (sometimes as few as 10,000 write cycles i.e. 100 times less!)
So writing continously you could wear an MLC cell out in six months, SLC would take 50 years!
Now the question is, which SSDs use MLC, and which SLC?
Price and preformance usually tell you the answer to that one...
Remember the limited write cycles is per cell, not per disk, and all controllers have had wear levelling (using different cells) for a few years now - thats partially the reason fro the slower writes.
Any "bad" cell can be mapped out just like a bad block on a hard drive ( but more transparently)
Last edited by mikerr; 03-07-2008 at 09:36 PM.
There are currently 1 users browsing this thread. (0 members and 1 guests)
Posting Permissions
Reply With Quote