Anyone running SSD's yet?

[PCBONEZ]

Re: Anyone running SSD's yet?

PZBONEZ; Here is a novel idea for you.
Instead of copying an irrelevant PDF why not take the relevant one?
It clearly states that the B5 and K5 SSD drives has a 5 year warranty:

FOR ALL INTEL® SSD 320 SERIES AND 520 SERIES PRODUCTS WHERE THE LAST TWO DIGITS OF THE PRODUCT CODE (as foundon Box Label) ARE “B5” OR “K5”
if the Product is properly used and installed, it will be free from defects in material and workmanship, and will substantiallyconform to Intel’s publicly available specifications for a period of five (5) years beginning on the date the Product waspurchased in its original sealed packaging in the case of an Original Purchaser or the date of original purchase of a computersystem containing the Product in the case of an Original System Customer.

It's perfectly relevant. - Same series of drive.

Now lets look at what it really says:

...will substantiallyconform to Intel’s publicly available specifications for a period of five (5) years

And those specifications are here:
https://cdn.badcaps-static.com/pdfs/...bcea01b56c.pdf
And for every instance having to do with speed it says "Up to".
That means 0.00001 MB/s is fully in spec and they don't have to cover it.
~ Interesting how that works. - Ain't it...
.
So enjoy your how ever many years of sub-ATA performance.
It's in spec...
.

[Per Hansson]

Re: Anyone running SSD's yet?

PZBONEZ; Here is a novel idea for you.
Instead of copying an irrelevant PDF why not take the relevant one?
It clearly states that the B5 and K5 SSD drives has a 5 year warranty:

FOR ALL INTEL® SSD 320 SERIES AND 520 SERIES PRODUCTS WHERE THE LAST TWO DIGITS OF THE PRODUCT CODE (as foundon Box Label) ARE “B5” OR “K5”
if the Product is properly used and installed, it will be free from defects in material and workmanship, and will substantiallyconform to Intel's publicly available specifications for a period of five (5) years beginning on the date the Product waspurchased in its original sealed packaging in the case of an Original Purchaser or the date of original purchase of a computersystem containing the Product in the case of an Original System Customer.

The other nasty little thing about SSD is when they wear out they're gone.
You can forget about whatever was on it No change of data recovery at all..
~ So if you are not religiously and frequently saving you data and work to a more reliable media then at some point you'll be writing it off.
That isn't much of a concern with 'real' discs and RAID 1.

Thanks for not even commenting your other lie that I disproved above that SSD's die when all the cells have been exhausted.

You didn't disprove anything.
I spoke the absolute truth.
What I said was when they are dead there is no hope of data recovery.
Dead means DEAD - not degraded.
On a platter drive you can -still- recover data even if it's stone dead should it be important enough to pay for the service.
KC8 actually does that himself on occasion.

[acstech]

Re: Anyone running SSD's yet?

It matters because some of us are trying to make an informed decision on whether or not to buy an SSD, and / or recommend them to our customers.

Hypothetical scenario:

Say I have a 3TB Seagate HD where I keep all my data, but I want some extra speed to boot the OS and run programs. That's all I want to do; install Windows 7 and Linux to it, and run programs off it. All the data will be stored on the Seagate.

How much gets written to the disk just booting it up and running typical programs? Would this be enough to noticeably slow down the SSD before, say, 5 years of use?

The way I understand it, if you get a larger (higher capacity) drive, for a given usage (amount of data written) it will last longer than a smaller, lower capacity drive. If we knew usage, we could calculate the useful lifetime, and therefore buy the correct drive, right?

[shovenose]

Re: Anyone running SSD's yet?

Um... why does this matter?
If you don't want/like SSDs then go buy a mechanical hard drive.
If you want/like SSDs then buy some

[mariushm]

Re: Anyone running SSD's yet?

It's pretty clear to me.
All 320 series have 5 year warranty, EXCEPT:

1.8" drives
OEM drives
drives that have wear out indicator (the K5 or B5)

The 2.5" models in RETAIL or RESELLER boxes have 5 year warranty, the ones that have wear-out indicator (K5 and B5) have 3 years, the 1.8" and OEM drives have 1 year (or 2 years in most EU as it's the minimum legally allowed).
Why should they address the question of that guy, when it's very clear the 1.8" ssd drive has was an OEM one?

There are older drives in the 320 series that didn't report through SMART the wear-out value, and Intel wanted to assure buyers that their data is safe by extending the warranty to 5 years.
The current models have indeed a 3 year warranty.

The info is here:

http://www.intel.com/support/ssdc/hp.../CS-032510.htm
http://www.intel.com/support/ssdc/hp.../CS-032511.htm

^ it's not hidden, it says right there on page with big font:

Intel® High Performance Solid State Drive
5-year limited warranty

The warranty information provided below is intended for the Intel® Solid-State Drive 520 Series and 320 Series Retail and Reseller SKUs where the media wear-out indicator (E9) is not applicable. Warranty information for all other Intel® Solid-State Drives can be found on the 3-year limited warranty page.

Later edit: this seems to contradict the document Intel has on the site... who knows, maybe you'll have to ask Intel directly : http://www.xbitlabs.com/news/storage...te_Drives.html

[PCBONEZ]

Re: Anyone running SSD's yet?

PCBONEZ
How about you provide some proof of your claims?

Sure.
The link you posted directs right to it. - Thanks.
.

And this guy can tell you what they are actually shipping. - B5 and K5.
It's right on Intel's site too.
http://communities.intel.com/thread/...ies%20warranty
- Notice how Intel didn't address it... No activity on the page is a good way to keep it out of Google.

Thanks for not even commenting your other lie that I disproved above that SSD's die when all the cells have been exhausted.

You didn't disprove anything.
I spoke the absolute truth.
What I said was when they are dead there is no hope of data recovery.
Dead means DEAD - not degraded.
On a platter drive you can -still- recover data even if it's stone dead should it be important enough to pay for the service.
KC8 actually does that himself on occasion.
.

[mariushm]

Re: Anyone running SSD's yet?

If one should last a decade it will spend roughly the last 8 years of it at around 33-70MB/s.
Many worse than and none better than a good SATA disc drive.
.
THAT is what the data CLEARLY shows for every one of the drives tested.
.

That might be the case IF you abuse the hell out of it by writing terabytes of data on it in the first two years.

But even if that's the case, the drive will keep working just great for high IO operations, which is a major advantage of SSD drives.

There are lots of situations where write speed is not important but fast disk access is, such as hosting and serving lots of small images, thumbnails, static documents that can't really be cached into memory.

For example, try saturating a 100mbps port (11.5MB/s) with hundreds of users requesting the avatar images and regular drives will choke, while the SSD drives will happily serve and saturate the port bandwidth.

[Per Hansson]

Re: Anyone running SSD's yet?

PCBONEZ
How about you provide some proof of your claims?

Confident in the enhanced reliability features of its recently introduced third-generation solid-state drive (SSD), Intel announced it has extended its limited warranty for the Intel® SSD 320 Series from three years to five years. The extended warranty term will apply to all Intel SSD 320 Series drives, including those already purchased. Additional limitations apply to enterprise usage levels.

This page is directly from Intel: http://newsroom.intel.com/community/...-intel-ssd-320


See here for more info and PDF's describing all the terms
http://www.intel.com/support/ssdc/hp.../CS-029645.htm

Thanks for not even commenting your other lie that I disproved above that SSD's die when all the cells have been exhausted.

[PCBONEZ]

Re: Anyone running SSD's yet?

If you think logically about the data, and how much regular people would write daily to tem, you will determine that the SSD drives are reliable enough to last for decades.

If one should last a decade it will spend roughly the last 8 years of it at around 33-70MB/s.
Many worse than and none better than a good SATA disc drive.
.
THAT is what the data CLEARLY shows for every one of the drives tested.
.
~~~
.
The 5-yr LIMITED warranty on the Intel 320 is hiding. [Can't find it in writing.]
Word has it:
- It only applies to the full retail boxed version which carries a different PN than the others.
- The others only have a 1 yr warranty. (As discovered by people that had them fail and were denied replacements.)
- Intel isn't shipping many (if any) of the full retail box version.
- Only the OEM PN is going out in significant numbers.
.......That tells me the 5-yr warranty claim is mostly a marketing ploy.

Won't surprise me if other companies are doing the same.
.

[mariushm]

Re: Anyone running SSD's yet?

The SSD drives do what they're advertised to do: low access times, high IOPS, UP TO ### MB/s.
They don't say the transfer speed is going to be fixed at a certain value for its whole life.

The tests above where users keep writing data to the drives until they start to run out of write cycles are not compatible with real world use of the drives. They do show that the drives can support a lot of writes on them, until their performance starts to degrade.

If you think logically about the data, and how much regular people would write daily to tem, you will determine that the SSD drives are reliable enough to last for decades.

You say that it's bad that a 64GB SSD drive dropped it's speed when it got to writing 450 TB on it. But do the math: in real world, how fast will a regular user copy 450 TB on a 64 GB drive? That's equal to completely erasing and filling up the drive about 8000 times - if a regular user would fill up the drive each day, it would still be 8000 days or about 20 years until the speed degrades. Even with a moderate 60 TB value, that's still 960 complete fills of the drive, and since a user doesn't fill the drive completely every day, that's a perfectly safe to assume it will last 3+ years.

It's like saying a formula 1 car is bad because it can fail faster or more often than a truck when climbing a mountain. Yes, it may fail faster, but you're not climbing a mountain several times each day and when you do climb it, you probably want to climb it fast.

We can play the same game with regular hard drives, where a similar test can be made to abuse them. Start a benchmark with a queue depth of 10-20, with lots of seeks and small writes - you'll probably find 1 out of 10 drives crashing their heads after 1-2 months of constant abuse. Not only the servo would cause issues but the drives would also heat a lot, which increases the chances of failure.
I won't determine out of this test that regular hard drives are bad, because regular users don't see a queue higher than 2-3 and don't write data to them 24/7.

[PCBONEZ]

Re: Anyone running SSD's yet?

I usually try to keep my mouth shut about the gigantic arguments you bring up against a technology, about how much it sucks just because YOU don't like it or see the point of it, but this is just too much bullshit for me to keep quiet.

EVERY SSD gives you warnings when shit hits the fan and it runs out of spare sectors, and everyone using a SSD should have either a manufacturer (like Intel Solid State Drive Toolbox) or 3rd party health tool (like CrystalDiskInfo or SSD Life (Pro) for it. (note: screenshots are from the SSD endurance test, after writing more than 450 terabytes of data nonstop)
Even if you don't, it'll start slowing down and "dropping out" (acting like someone pulled the SATA plug) to indicate something's wrong.

Even if you did manage to run completely out of spare sectors, the drive is still readable for a certain period of time. You just can't write a single byte to it anymore.

If you even bothered to click the link in post #15, you would have seen that this is exactly what happened to the Samsung 470 series 64GB SSD.
At first, writing slowed down from 113MB/s to 108MB/s (after 450TB of written data), then it slowed down a bit more to 105MB/s (472TB), then the write endurance test stopped with an error (at 478TB). Write speed crumbled down to 12.8MB/s and the last reported value was 5.3MB/s.

After unplugging the drive and letting it sit for a while, it was detected normally again (had dropped out), all the data was still there and the CRC checksums on the static data that was on there during the whole endurance test were OK too.
Almost exactly 1 month later, the guy tried the SSD again, and it refuses to be detected properly (constantly in busy state).

In other words: even if you manage to kill the drive by running out of spare sectors, the data on it at the time remains readable and intact to make a backup. From then on, only time kills the drive (can't be written to, so cells can't be refreshed by its controller either)

And a quote from that guy to explain the 478TB written data number a bit:


... and that SSD wasn't one with the most written-to data either (A Crucial M4 64GB managed 768TB, a Corsair Force 3 120GB 1015TB..)

Just because you don't see the point in something or don't like it doesn't mean it's crap.
Especially if you can't even be bothered to look stuff up on the topic.

.
I have a right to my opinion just like everyone else in here and if you don't happen to like my opinion that's your problem.
-
-
I base my opinion on practical data and critical thinking, not advertising hype or baseless extensions of it.
And I don't feel a need to LIE about my own experiences to discredit someone else's opinion.
.
.
Yes I read that page you linked to in post #15. - Long before you posted it too. - It's an extension of add hype.
My conclusion was those people are a pack of blithering idiots.
They have NO CONCEPT of the real world implications of the data they are collecting.
As such most of their data is pointless and their (and readers) conclusions are bogus.
---
And anyone that knows how to analyze data would realize that thread actually SUPPORTS my opinion.
---.
Here is an automotive analogy to their 'test'.
The SSD-Car goes faster than the Disc-Car so it's supposedly better.
So lets see how long the SSD-Car lasts to see if it's actually worth the higher cost.
They run the SSD-Car to death.
In ~2 years it blows a gasket, it's now running on 1/2 it's cylinders.
It's now much slower than the Disc-Car which has many years left in it before it 'slows down' at all.
- Anyone that is not off in La La Land would call that the end of the comparison.
- NOT THESE GUYS!
They keep running the car on 1/2 it's cylinders at sub Disc-Car speeds to see how much many more miles they can squese out of it before it won't run at all.
POINTLESS - STUPID - BOGUS
Once the gasket blew the Disc-Car was the better car. - END OF STORY.
-
Once the SSD-Car becomes slower than the Disc-Car the comparison is over.
The ONLY thing that made the SSD-Car better was that it's faster.
It's NOT faster anymore.
It's no longer better.
Continuing to log data after that is: A measure of how long the SSD-Car sux in comparison to the Disc-Car.
-
-- The point where MWI vs Host Writes goes flat and horizontal <=> when the gasket blew.
After that the transfer rates on every one of those drives was no more than the ATA specs.
Some of them -barely- beat ATA-33.
THAT IS THE END OF THE DRIVE BEING FASTER THAN A DISC DRIVE.
And that's when having an SSD Drive becomes POINTLESS. - It's NOT faster anymore. - ZERO advantage to SSD.
-
The drive you like was just over ATA-100 after the curve went horizontal.
(Much better than the others but still sucky for SSD. OTOH: Looking at some of his posts I think that guy doesn't know how to find the correct drive speed and that's why his was so much different than the others.)
BUT CONGRATS! - Of your 478 TB you moved ~ 60TB fairly fast,,, then 418TB at the speed of a 10 year old IDE interface.
I'm so impressed.....
-
If the whole point of getting and SDD is 'speed' then you need to throw it away when the curve goes flat.
That would be at around 60TB for your champion drive..
If you are still using it after that point 'and' claiming it's better than a SATA Disc drive - you are a lair.
The content of your link PROVES THAT for multiple SSD drives. - So thanks again for the link...
-
-.
If you want a drive that's faster than a disc drive for ~2 years then slower than an ATA Disc drive forever after - then more power to ya.
Limp it along for a decade at sub-ATA speeds for all I care.
~ Enjoy it.
.

[Rob Northen]

Re: Anyone running SSD's yet?

I've used one for 302 days now, in a server for my security system. Its on 24/7, but with very low write activity. Its great for me, cause it makes no noise and its small in size, so I could pick a smaller PC cabinet.

[mariushm]

Re: Anyone running SSD's yet?

Exactly, Per. PCBONEZ just needs to keep up with technology and not just repeat some assumptions someone made a long time ago in the past.

Read this: http://www.anandtech.com/print/2738

Old MLC cells used to have up to about 10k erase cycles (which should not be confused with writes), newer smaller MLC cells have about 3-4000 cycles, but the wear leveling algorithms have improved a lot, so the long term reliability remains about the same.

Controllers that can do a lot of throughput (like the Sandforce chips) use compression to reduce the writes and in all controllers the wear leveling algorithms reduce the need to do erases to make pages available.

Also, controllers like Sandforce keep track of how much data they write to cells and at some point, they actually decrease the write speed to cells in order to keep the cells good until the warranty expires - ex. user writes 5-10 TB to a 120 GB Sandforce based SSD drive with about 250 MB/s and then the controller starts throttling the writes to 220 MB/s or less for the next 1-5TB or so of data and then keeps going lower...

Basically, if the drive is advertised with 3 year warranty, it tries to make sure that by the end of those 3 years there will still be a few erase cycles to keep the drive working. When the erase cycles are gone, the drive becomes read only and data is not lost. But in 3-5 years, you're expected to buy a newer drive anyway.

This sounds ugly, but a regular user will never write 2 TB of data on a 120 GB every day.

[Per Hansson]

Re: Anyone running SSD's yet?

Yes, I wanted to stay out of this thread too, I fully agree with Scenic.
PCBONEZ you are a great guy but sometimes I would just want a temporary ignore filter that hides your posts from specific topics, please, no hard feelings meant.

We actually even went over this before, and I tried to keep quiet the longest time then too: https://www.badcaps.net/forum/showthread.php?t=17338

When you say things like: "SSD are expected to last 2-3 years with 'average' use."
Intel upgraded the warranty from 3 to 5 years on all of their current & old previously sold SSD 320 series drives!
As it is I bought a Transcend 32GB SSD for my IPCOP firewall April 2008, it still works fine.
That's quite close to 4 years ago, and it's online 24/7 since it's a firewall...
It's also one of the shittiest SSD's money can buy, being based off a JMicron controller which should be avoided like the plague...

The other nasty little thing about SSD is when they wear out they're gone.
You can forget about whatever was on it.
No change of data recovery at all.
~ So if you are not religiously and frequently saving you data and work to a more reliable media then at some point you'll be writing it off.
That isn't much of a concern with 'real' discs and RAID 1.

You said the same thing in the thread I linked above, I wrote this reply for you back then: https://www.badcaps.net/forum/showpo...2&postcount=72

And your comments about harddrives having a 5 year warranty is pretty moot too.
Lately HDD manufacturers have reduced their warranty lengths.

And before you say that I'm trying to just defend my SSD purcache understand that I don't even have a SSD in my primary system, I have 3x1TB harddrives in RAID-5 and 1xSegate Cheetah 300GB 15k RPM drive for the OS and apps.
And as it stands 2x of those 1TB harddrives have failed and needed to be replaced under warranty.

I do however install SSD drives in clients computers, and even in industrial systems costing millions of dollars.
Simply because they are more reliable than mechanical harddrives.
So far I have not had a single drive fail, but then I only buy Intel drives...

[Scenic]

Re: Anyone running SSD's yet?

The other nasty little thing about SSD is when they wear out they're gone.
You can forget about whatever was on it.
No change of data recovery at all.
~ So if you are not religiously and frequently saving you data and work to a more reliable media then at some point you'll be writing it off.

I usually try to keep my mouth shut about the gigantic arguments you bring up against a technology, about how much it sucks just because YOU don't like it or see the point of it, but this is just too much bullshit for me to keep quiet.

EVERY SSD gives you warnings when shit hits the fan and it runs out of spare sectors, and everyone using a SSD should have either a manufacturer (like Intel Solid State Drive Toolbox) or 3rd party health tool (like CrystalDiskInfo or SSD Life (Pro) for it. (note: screenshots are from the SSD endurance test, after writing more than 450 terabytes of data nonstop)
Even if you don't, it'll start slowing down and "dropping out" (acting like someone pulled the SATA plug) to indicate something's wrong.

Even if you did manage to run completely out of spare sectors, the drive is still readable for a certain period of time. You just can't write a single byte to it anymore.

If you even bothered to click the link in post #15, you would have seen that this is exactly what happened to the Samsung 470 series 64GB SSD.
At first, writing slowed down from 113MB/s to 108MB/s (after 450TB of written data), then it slowed down a bit more to 105MB/s (472TB), then the write endurance test stopped with an error (at 478TB). Write speed crumbled down to 12.8MB/s and the last reported value was 5.3MB/s.

After unplugging the drive and letting it sit for a while, it was detected normally again (had dropped out), all the data was still there and the CRC checksums on the static data that was on there during the whole endurance test were OK too.
Almost exactly 1 month later, the guy tried the SSD again, and it refuses to be detected properly (constantly in busy state).

In other words: even if you manage to kill the drive by running out of spare sectors, the data on it at the time remains readable and intact to make a backup. From then on, only time kills the drive (can't be written to, so cells can't be refreshed by its controller either)

And a quote from that guy to explain the 478TB written data number a bit:

That is equal to 8212 times writing the 64GB capacity of the drive. Which is equivalent to writing the entire 64GB capacity of the drive, every day, for 22 years. Or to writing 287GB every day for 5 years.

... and that SSD wasn't one with the most written-to data either (A Crucial M4 64GB managed 768TB, a Corsair Force 3 120GB 1015TB..)

Just because you don't see the point in something or don't like it doesn't mean it's crap.
Especially if you can't even be bothered to look stuff up on the topic.

[PCBONEZ]

Re: Anyone running SSD's yet?

I just wanted to try one out. Thanks for throwing us all under the bus.

I didn't do that, you did.
The information is all out there if you look.
.
If you only look at info from the salesmen or at reviews that are supported by the drive manufacturer's advertising what do you think they're gonna say?
.
.
If you look in similar places about Asus they are all going to say Asus is great too.
.

[Mad_Professor]

Re: Anyone running SSD's yet?

No you didn't.
You 'invented' and experience and that seems to be very common among those that tout the attributes of those things.
.
You even admitted your bulk storage is on a file server and:
1: an SSD isn't going to help with that and that makes it slower than a local [real] hard drive.
2: if you weren't using an SSD you wouldn't HAVE to do that.

I'm not about to pay $125 for a drive that's only going to last 2-3 years.
- I don't care if anyone likes that or not.
- If you read up on them someplace that is truly objective that's all they last.
I pay a lot less for much larger drives with 5 year warranties and they generally actually last more like 10+ years.
.
.
The other nasty little thing about SSD is when they wear out they're gone.
You can forget about whatever was on it.
No change of data recovery at all.
~ So if you are not religiously and frequently saving you data and work to a more reliable media then at some point you'll be writing it off.
That isn't much of a concern with 'real' discs and RAID 1.
.
.
When they get down to like $40 [NEW] I 'might' get one to play with one as a toy [cause I a hobbyist PC geek] but it would never be used for anything the least bit important. Ever.
- If they never get that cheap I won't be feeling any loss.
- The tech just isn't 'there' yet and I'm thinking tech is on the verge of going in another direction soon enough making SDD a mute point.
[The 8-track drives [or Beta-Max] of the New Millennium..]
.

I just wanted to try one out. Thanks for throwing us all under the bus.

[PCBONEZ]

Re: Anyone running SSD's yet?

Not here to argue semantics, just state my experience.

No you didn't.
You 'invented' and experience and that seems to be very common among those that tout the attributes of those things.
.
You even admitted your bulk storage is on a file server and:
1: an SSD isn't going to help with that and that makes it slower than a local [real] hard drive.
2: if you weren't using an SSD you wouldn't HAVE to do that.

I'm not about to pay $125 for a drive that's only going to last 2-3 years.
- I don't care if anyone likes that or not.
- If you read up on them someplace that is truly objective that's all they last.
I pay a lot less for much larger drives with 5 year warranties and they generally actually last more like 10+ years.
.
.
The other nasty little thing about SSD is when they wear out they're gone.
You can forget about whatever was on it.
No change of data recovery at all.
~ So if you are not religiously and frequently saving you data and work to a more reliable media then at some point you'll be writing it off.
That isn't much of a concern with 'real' discs and RAID 1.
.
.
When they get down to like $40 [NEW] I 'might' get one to play with one as a toy [cause I a hobbyist PC geek] but it would never be used for anything the least bit important. Ever.
- If they never get that cheap I won't be feeling any loss.
- The tech just isn't 'there' yet and I'm thinking tech is on the verge of going in another direction soon enough making SDD a mute point.
[The 8-track drives [or Beta-Max] of the New Millennium..]
.

[japlytic]

Re: Anyone running SSD's yet?

From what I have read, ReadyBoost under Windows Vista onwards would be disabled on an SSD (if it is the system drive) because it has little or no effect on performance.
I am not sure if ReadyBoost would work on a standard HDD with an SSD being used for ReadyBoost.

[colormebad]

Re: Anyone running SSD's yet?

Im running the extra 40gig SSD that came with my Asus barebone kit...I set it up like they said using the Intel Smart Responce Technology..Its a good bit faster using the ssd with my 350 sata drive...