Intel SSD 520 Issues

[davewolfs]

I owned a X25M which worked flawlessly for years. Decided to upgrade to a 520 and my sudden has become unstable. Random BSOD's.

Anyone else having issues with this drive? Any suggestions on what to do. Should I get a refund and go with something else?

 

[JoeComp]

I'd return it and get either a Plextor M3P or a Samsung 830.

Sandforce SSDs should be avoided. Intel fixed some of the Sandforce BSOD bugs, but obviously they did not get them all.

 

[mwroobel]

The 520 is a good drive, just make sure you have updated it to the latest firmware, which fixes a BSOD problem with this and other Sandforce drives.

 

[JoeComp]

The 520 is a good drive, just make sure you have updated it to the latest firmware, which fixes a BSOD problem with this and other Sandforce drives.

Please provide a link to the specific BSOD problem that you claim is fixed in some later version of the Intel 520 firmware. And what Intel 520 firmware version are you saying has fixed this particular BSOD problem?

Are you willing to bet the OP a decent SSD (Samsung or Plextor) that if he updates the firmware on his Intel 520 that it will fix his BSOD problems?

 

[mwroobel]

The 520 is based on an Intel-Modified Sandforce SF-2281 controller. The SF-2281 is used in a number of manufacturers' SSDs and has had a number of firmware updates released. These updates fix a number of problems, but the one most prevalent and most reported by users have been random but recurring BSODs. Install the latest Intel Toolbox to see if there is an update to your firmware.
No, I will not bet anyone on anything. I am simply making suggestions and observations and there could be 100 different things causing the BSOD. Other things to do is make sure you have all updates to your system installed, BIOS, firmwares, drivers, etc. and please post all the BSOD information you are receiving. If you can, also post a link to your dumpfile and I can give you more of an idea where the problem is coming from, or at least what is the app that is actually panicking the machine.

 

[JoeComp]

The Intel 520 is NOT "based on an Intel-Modified Sandforce SF-2281 controller".

The Intel 520 has a standard Sandforce 2281 controller. However, the firmware Intel is using has been customized to some extent, but it is not clear how much customization was done, or whether Intel made the changes, or whether Sandforce did at Intel's request. The changes apparently fixed some issues that Intel found during their testing, and made some minor performance differences.

But there are multiple issues affecting Sandforce SSDs, and Intel's 520 firmware does NOT fix all of the BSOD issues that people have with Sandforce SSDs.

Sandforce SSDs have a long history of firmware issues that cause seemingly random BSODs under various conditions and on various hardware configurations. The Intel 520's customized Sandforce firmware has clearly not fixed all of these issues, since there are still many reports just like this one.

This report seems clear cut. The OP had a system that was working well with an X25-M, then he switched to an Intel 520 and had BSODs. Unless there was some other change that the OP neglected to mention, it seems obvious that the Intel 520 still has BSOD issues in some systems.

I'm not surprised you are not confident enough to bet. Only a fool would be confident that any Sandforce SSD, even Intel's, has solved the numerous BSOD issues that have plagued Sandforce for a couple years now.

 

[mwroobel]

The Intel 520 is NOT "based on an Intel-Modified Sandforce SF-2281 controller".

The Intel 520 has a standard Sandforce 2281 controller. However, the firmware Intel is using has been customized to some extent, but it is not clear how much customization was done, or whether Intel made the changes, or whether Sandforce did at Intel's request. The changes apparently fixed some issues that Intel found during their testing, and made some minor performance differences.

But there are multiple issues affecting Sandforce SSDs, and Intel's 520 firmware does NOT fix all of the BSOD issues that people have with Sandforce SSDs.

Sandforce SSDs have a long history of firmware issues that cause seemingly random BSODs under various conditions and on various hardware configurations. The Intel 520's customized Sandforce firmware has clearly not fixed all of these issues, since there are still many reports just like this one.

This report seems clear cut. The OP had a system that was working well with an X25-M, then he switched to an Intel 520 and had BSODs. Unless there was some other change that the OP neglected to mention, it seems obvious that the Intel 520 still has BSOD issues in some systems.

I'm not surprised you are not confident enough to bet. Only a fool would be confident that any Sandforce SSD, even Intel's, has solved the numerous BSOD issues that have plagued Sandforce for a couple years now.

I don't work for Sandforce and have no agenda to push, it was you who suggested I bet and I have no interest in doing so. I said (in this and the other thread we are chatting in) that with the drives we have, where the BSOD's could be directly traced to the drives that for us 95%+ of the problems we did have have been alleviated by the various firmware updates. I didn't say that all problems in the world have been fixed. But at this point, the sandforce based drives are neither a garden of perfection nor a rotting cesspool of failure and corruption. Buy what you are comfortable with. I have standardized for MLC drives on the Samsungs at this point. As new drives are released we shall see. But being Sandforce based does not make the OP's drive a piece of crap nor does it guarantee his particular BSODs are or are not related. As I said, there are literally hundreds of things that could cause problems, hardware and software. If the OP posts more information on the BSOD and a dump, we can look further into the failing app and a more definite place to search for the problem.

 

[JoeComp]

Heh, even you have moved away from Sandforce SSDs to Samsungs, huh?

So my advice to the OP matched what you actually did.

It seems you are playing "do what I say, not what I do".

 

[mwroobel]

Our move to Samsung had nothing to do with Sandforce, it had to do with a contract we have with them and the pricing we received for 400 of them. I still have 200+ SF-2281 based drives in various boxes and am not running to the exits to replace them. They will be replaced as needed or when the boxes they are in have depreciated out and are replaced. We are also still buying 240GB 520's for a particular project where the price/performance ratio works for 16 drive R10 Arrays (10 of them) and they have are loved by the people tasked with supporting those boxes.

 

[mephistopoa]

The 520 is a good drive, just make sure you have updated it to the latest firmware, which fixes a BSOD problem with this and other Sandforce drives.

don't make false statements. 520 never had a firmware update, never had BSOD problems. The person could be using a faulty chipset that was not recalled and the problem would happen with any SSD on a 6GB SATA port.

 

[mwroobel]

don't make false statements. 520 never had a firmware update, never had BSOD problems. The person could be using a faulty chipset that was not recalled and the problem would happen with any SSD on a 6GB SATA port.

Actually, there was. We received a batch of 50 520's with 400h the week they were released, and intel sent us an update to 400i. I don't know if the OP had this same problem which is why I suggested it. With ANY otherwise-normal operating SSD's updating the firmware is always a good suggestion as a first troubleshooting step.

 

[mephistopoa]

I never managed to get in hold of a 400h, all of them I have tried so far were 400i from factory since very early in the launch.

Anyway, I don't like the idea of blaming the drive for something like this as I've tried many SSDs and so far the only shitty sandforce I've got were OCZ. Not just sandforce 2 but also the first one. I'm having so many of them simply dead after 18 months.

 

[JoeComp]

Anyway, I don't like the idea of blaming the drive for something like this

He said all he did was replace an Intel X25-M with an Intel 520. Unless he is not telling us something, then it appears the Intel 520 is at fault.

 

[baldrik]

I've been running 2 Vertex3s for over a year that have run in 3 different systems. They have not caused me a single problem. Anecdotal for sure, but with the firmware updates available, I see no reason to keep treating sandforce drives like they punched your momma in the babby maker.

 

[JoeComp]

I've been running 2 Vertex3s for over a year that have run in 3 different systems. They have not caused me a single problem. Anecdotal for sure, but with the firmware updates available, I see no reason to keep treating sandforce drives like they punched your momma in the babby maker.

The simple concept that you are missing is that it does not matter a whit whether you have had any problems with your SSD(s). If there is, for example, a 5% chance that a Sandforce SSD will have problems with a randomly selected computer, then there are going to be 19 computers that the Sandforce SSD works okay with for every 1 computer that the Sandforce SSD causes to crash.

Whatever the actual odds are (it is hard to tell), the relative odds are poor. All you have to do is compare the percentage of reviews that have problems for Sandforce SSDs versus other SSDs, and they are usually around 2 to 3 times worse for Sandforce SSDs as compared to Plextor or Samsung SSDs.

Sandforce SSDs are just a bad bet. Sandforce should be avoided.

The best SSDs now are the Samsung 830 for a good value, or the Plextor M3P (soon the Plextor M5P) for best performance and quality.

 

[dilidolo]

The simple concept that you are missing is that it does not matter a whit whether you have had any problems with your SSD(s). If there is, for example, a 5% chance that a Sandforce SSD will have problems with a randomly selected computer, then there are going to be 19 computers that the Sandforce SSD works okay with for every 1 computer that the Sandforce SSD causes to crash.

Whatever the actual odds are (it is hard to tell), the relative odds are poor. All you have to do is compare the percentage of reviews that have problems for Sandforce SSDs versus other SSDs, and they are usually around 2 to 3 times worse for Sandforce SSDs as compared to Plextor or Samsung SSDs.

Sandforce SSDs are just a bad bet. Sandforce should be avoided.

The best SSDs now are the Samsung 830 for a good value, or the Plextor M3P (soon the Plextor M5P) for best performance and quality.

I have over 10 Sandforce based SSDs from different manufactures, including OWC, Corsair, Patroit, Intel. None of them had single issue so far. I also have many other SSDs such as Intel E, M, M4, Corsair Nova, etc. Again, none of them has single issue so far.

I think a lot of the problems come from mother BIOS and BIOS setting, plus controller driver.

I saw someone mentioned Sandforce SSD has problem with SAS controller running at 6Gb. I have a few connected to LSI SAS, they all run at 6Gb.

 

[baldrik]

Whatever the actual odds are (it is hard to tell), the relative odds are poor.

So you don't know what the odds are, but you know the odds are poor.

I can't tell if you didn't think that statement all the way through, if you're trolling me or if you're just a Marvell fanboi.

1) I did not miss the point. I said my evidence was anecdotal. I also have the same anecdotal evidence for my Crucial M4s - 2 drives, no problems.

2) Thanks for explaining that 5% is 1 in 20. I never took math past the 3rd grade, so I'm glad someone finally got around to teaching me this. I can't wait to tell my DnD buddies about this concept. Minds will be blown.

3) Comparing the number of reviews tells me nothing. Drive A could have the exact same failure rate as Drive B, but drive A has 5 times more bad reviews. This could very easily be because Drive A sold 5 times as many drives. It could also be because the wind blew from the east on Tuesday and blew Mr Johnson's hat off, which made him irate and the only vent for his infantile rage was the internet. If you are basing your statements off of reviews, you're evidence is really just as anecdotal as mine.

4) What you're failing to take into account is the failure rate for drives now. If massive improvements to the firmware have been made that corrected the issues that were causing failure, as they have, then the failure rate for drives will be much lower. If I manufacture cars with leaky gas tanks that have a 5% chance of exploding because of said tank and then I fix said tank, my chances for explosions are going to plummet. Which is a shame, because I like explosions.

 

[jwcalla]

Well before a certain firmware revision, 100% of Crucial m4s would experience a BSOD after six months.

My recommendation to the OP is to contact Intel or visit one of their customer support forums.

 

[JoeComp]

So you don't know what the odds are, but you know the odds are poor.

I can't tell if you didn't think that statement all the way through, if you're trolling me or if you're just a Marvell fanboi.

Or, you did not understand what I was explaining. I said that I don't know what the actual odds are. In other words, I don't know whether it is 1% or 5% or 10%. But I DO know what the relative odds are. As I said, it is about 2 to 3 times worse for Sandforce as compared to Plextor or Samsung. Since you seem to have trouble understanding this simple concept, I will say it in yet another way. If the chances of major problems with Plextor is called "X", then the chances with Sandforce are about 2X or 3X.

 

[jwcalla]

Or, you did not understand what I was explaining. I said that I don't know what the actual odds are. In other words, I don't know whether it is 1% or 5% or 10%. But I DO know what the relative odds are. As I said, it is about 2 to 3 times worse for Sandforce as compared to Plextor or Samsung. Since you seem to have trouble understanding this simple concept, I will say it in yet another way. If the chances of major problems with Plextor is called "X", then the chances with Sandforce are about 2X or 3X.

Wait, where are you getting that info? Does that include SF-1200 drives? Plextor and Samsung are relative late-comers compared to SF.

 

[JoeComp]

Well before a certain firmware revision, 100% of Crucial m4s would experience a BSOD after six months.

No, the problem occurred after 5200 hours of power on time. That is 7.1 months.

And a firmware update completely fixed the issue. 0% of Crucial m4 SSDs have a BSOD due to the 5200 hour issue with the updated firmware.

In contrast, Sandforce SSDs BSOD crashing problems are not due to a single clear cause, so it is difficult to fix the issue completely. As is obvious, since there are still Sandforce BSOD issues years after they were first reported.

Sandforce has a long and terrible history of BSOD problems. Some of the causes have been fixed, but others remain. Due to the bad history and chances of remaining problems, Sandforce SSDs are a bad bet.

 

[mwroobel]

Or, you did not understand what I was explaining. I said that I don't know what the actual odds are. In other words, I don't know whether it is 1% or 5% or 10%. But I DO know what the relative odds are. As I said, it is about 2 to 3 times worse for Sandforce as compared to Plextor or Samsung. Since you seem to have trouble understanding this simple concept, I will say it in yet another way. If the chances of major problems with Plextor is called "X", then the chances with Sandforce are about 2X or 3X.

And you came to this 2X or 3X absolute claim HOW (Clairvoyance perhaps?) In my particular case, where I quoted percentages I went back in our trouble ticket system, saw how many were drive related, how many were then closed and complete and how many were outstanding. That is where I came up with the 95%+ (95.6 actually) and gave you the breakdowns of the remaining open/closed tickets. You (comically, and rather incorrectly) called me an OCZ shill before (and then I pointed out to you where I said to stay AWAY from OCZ) and now it seems like you are trying to blast Sandforce (with a particular agenda?)

 

[JoeComp]

And you came to this 2X or 3X absolute claim HOW (Clairvoyance perhaps?)

I mentioned that in my post where I first gave the number.

And I did not call you an OCZ shill. I said your comment sounded like one. There is a difference. I knew you avoided OCZ, which is why I made that parenthetical comment.

You would do better to try to read my posts and understand them before ridiculing them for things you have misunderstood.

 

[baldrik]

Or, you did not understand what I was explaining. I said that I don't know what the actual odds are. In other words, I don't know whether it is 1% or 5% or 10%. But I DO know what the relative odds are. As I said, it is about 2 to 3 times worse for Sandforce as compared to Plextor or Samsung. Since you seem to have trouble understanding this simple concept, I will say it in yet another way. If the chances of major problems with Plextor is called "X", then the chances with Sandforce are about 2X or 3X.

Dude, I don't think you understand what you're explaining. I'm just trying to figure out how you can come up with this "2x or 3x" when you have no numbers to base it on. None at all. Oh, that's right, you said, "bad reviews." You might as well say there are "2x or 3x" more tooth fairies than there are hobgoblins. You have just as much evidence of that.

It's also rather obvious that the part of my post that you quoted is the part where you stopped reading. I also don't think you know what anecdotal means.

 

[mwroobel]

I mentioned that in my post where I first gave the number.

And I did not call you an OCZ shill. I said your comment sounded like one. There is a difference. I knew you avoided OCZ, which is why I made that parenthetical comment.

You would do better to try to read my posts and understand them before ridiculing them for things you have misunderstood.

I have not misunderstood anything. Why you seem to be taking this adversarial "Sandforce sucks and everyone else is wrong" stance is beyond me. Whatever your biases are, you should make your determinations on specific cases and not just "All Sandforce-based drives suck and are bad and kill kittens(final being my sarcastic addition, before you demand I like to examples of Sandforce drives killing kittens), etc" rants.

 

[Meeho]

1. Forums

2. http://forum.notebookreview.com/sol...78644-ssd-reliability-study-return-rates.html

3. http://www.avsforum.com/t/1420361/interesting-ssd-reliability-study

Google translate:
"However, OCZ is well behind what reliability: on average there will be 3x as many drives back at Intel, and Crucial."

"It is unfortunate that no more results are in contrast to conclude that the vertex 1 and 2 are unreliable, and the Intel X25-M G2 very reliable, I can at this time. The issues have a similar picture: X25-m good, bad vertex."

"We can draw the following conclusions:
a.. The Indilinx Barefoot controller is very unreliable.
b.. Followed by the SF-1200. This controller is not very reliable according to these statistics.
c.. The SF-2200 improves a lot, but there is a remarkably high number of RMA% which means that many people have returned due to disk issues, or because they did not like the color. That is also a possibility
d.. Finally, both the Marvel and intel controller are relatively reliable."

"Based on the above results, I can advise hetvolgende: an SSD with a SF-1200 or SF-2200 controller you can, from reliability considerations should not take."

 

[mwroobel]

1. Forums

2. http://forum.notebookreview.com/sol...78644-ssd-reliability-study-return-rates.html

3. http://www.avsforum.com/t/1420361/interesting-ssd-reliability-study

Google translate:
"However, OCZ is well behind what reliability: on average there will be 3x as many drives back at Intel, and Crucial."

"It is unfortunate that no more results are in contrast to conclude that the vertex 1 and 2 are unreliable, and the Intel X25-M G2 very reliable, I can at this time. The issues have a similar picture: X25-m good, bad vertex."

"We can draw the following conclusions:
a.. The Indilinx Barefoot controller is very unreliable.
b.. Followed by the SF-1200. This controller is not very reliable according to these statistics.
c.. The SF-2200 improves a lot, but there is a remarkably high number of RMA% which means that many people have returned due to disk issues, or because they did not like the color. That is also a possibility
d.. Finally, both the Marvel and intel controller are relatively reliable."

"Based on the above results, I can advise hetvolgende: an SSD with a SF-1200 or SF-2200 controller you can, from reliability considerations should not take."

Understand that I am not taking any particular side here (other than saying the "Sandforce=D00M statement is overblown." That said, these numbers seem to be taken from an unnamed french etailer and another dutch retailer with a sample size of about 1500 units. It was taken in the time frame when the bulk of the drives that were sold hadn't had better-working firmware yet released for the majority of the drives. One of the studies seems to breakout general rma'd drives from complete failures. Nowhere does it state that they were returned for firmware related problems, BSOD's etc.. just that they "died". It is also an extremely small sample size, like saying that 75 out of 1500 people have asthma, so 5% of the world has asthma. You need a larger sample group, and a more widespread demographic to make arbitrary statements.

 

[JoeComp]

It's also rather obvious that the part of my post that you quoted is the part where you stopped reading. I also don't think you know what anecdotal means.

I was trying to be polite, but since that obviously means nothing to you, I will be blunt.

Your post betrays your complete ignorance of basic math. I don't know about you, but I learned about percentages in elementary school. It amazes me that you feel qualified to make comments like you did without understanding the most basic math concepts such as percentages. No doubt you don't even realize how ignorant you are and will come back with some nonsense reply. But I don't feel like wasting my time educating you, so you will have to find someone else to tutor you, or just remain ignorant.

 

[JoeComp]

I have not misunderstood anything. Why you seem to be taking this adversarial "Sandforce sucks and everyone else is wrong" stance is beyond me. Whatever your biases are, you should make your determinations on specific cases and not just "All Sandforce-based drives suck and are bad and kill kittens(final being my sarcastic addition, before you demand I like to examples of Sandforce drives killing kittens), etc" rants.

And you misunderstand again. I am not surprised you have so much difficulty, since it is clear that you don't read posts carefully and you do not think much before writing.

And please do not make up quotes and misattribute them to me. I never said "Sandforce sucks" nor "everyone else is wrong".

I said that Sandforce is a bad bet, since it is 2 or 3 times more likely to have problems with randomly chosen computer hardware, as compared to Plextor or Samsung.

 

[mwroobel]

And you misunderstand again. I am not surprised you have so much difficulty, since it is clear that you don't read posts carefully and you do not think much before writing.

And please do not make up quotes and misattribute them to me. I never said "Sandforce sucks" nor "everyone else is wrong".

I said that Sandforce is a bad bet, since it is 2 or 3 times more likely to have problems with randomly chosen computer hardware, as compared to Plextor or Samsung.

<Sarcasm> Well, then everyone on the board must bow to your superior intelligence.</Sarcasm> I still don't understand your abrupt and condescending attitude towards a number of people here (myself included.)

 

[JoeComp]

I still don't understand your abrupt and condescending attitude towards a number of people here (myself included.)

It is good that you admit your misunderstanding. Next step is to try to read carefully so that you will understand. But I see that you apparently value quantity of posts over quality, so I don't think there is much chance of that. Probably about the same chance as a Sandforce SSD having problems with randomly chosen hardware. Or even 2 or 3 times lower, the chance of a Plextor or Samsung SSD having problems.

 

[jwcalla]

I said that Sandforce is a bad bet, since it is 2 or 3 times more likely to have problems with randomly chosen computer hardware, as compared to Plextor or Samsung.

You were just asked to back up your numbers, that's all, since you seem so certain. All we have is what you've read from user reviews. Whatever that means.

Either way, Plextor was very late to the consumer game and I'm not sure how Samsung's first-gen drives fared. But I would put a decent amount of trust into a modern SF drive. I think they're all fairly comparable at this point.

JMO of course.

 

[bAMtan2]

a lot of talking in this thread from people who dont know what they're talking about. bunch of thread pissing hijackers arent even replying to the thread starter who isnt even here

 

[mwroobel]

a lot of talking in this thread from people who dont know what they're talking about. bunch of thread pissing hijackers arent even replying to the thread starter who isnt even here

In post #5 I made some suggestions, and asked for more info and a link to a dumpfile where we could get more precise info to help. Another here decided to turn it into a "Reasons not to buy anything that has a Sandforce controller" dissertation/pissing match.

 

[baldrik]

I was trying to be polite, but since that obviously means nothing to you, I will be blunt.

Your post betrays your complete ignorance of basic math. I don't know about you, but I learned about percentages in elementary school. It amazes me that you feel qualified to make comments like you did without understanding the most basic math concepts such as percentages. No doubt you don't even realize how ignorant you are and will come back with some nonsense reply. But I don't feel like wasting my time educating you, so you will have to find someone else to tutor you, or just remain ignorant.

Chief, I know what relative odds are and yes, I can manage to mash 2 brain cells together and figure out a percentage or two. What I've been trying to point out is that you need actual numbers to figure them and that anecdotal evidence found in reviews does not count. Unless you've got a large test pool yourself or access to the manufacturers RMA numbers, then you're just spouting the "2x or 3x" out your ass. Or possibly you read that number in some other post and you're spouting that number out of someone else's ass. Which sounds entirely unsanitary. I suppose it was all a roundabout, obtuse way of asking for a citation. And no, I don't mean the Chevy.

 

[John721]

Decided to upgrade to a 520 and my sudden has become unstable. Random BSOD's.

Anyone else having issues with this drive? Any suggestions on what to do. Should I get a refund and go with something else?

I'm running two 520s right now. Absolutely stable with no bsods, and on an overclocked system to boot.

I'd say swap it for another 520 and see what happens. You chose it over the competition for reasons that must have been good at the time.

If that doesn't work out, I'd probably keep an eye out for Plextor's new M5 Pro. 5 year warranty and some very strong component quality by the look of the previews. If early adoption isn't your style, Samsung seems to get a litany of positive reviews and has been around.

 

[hmz]

My 520 works great without issues.

 

[JoeComp]

Either way, Plextor was very late to the consumer game and I'm not sure how Samsung's first-gen drives fared. But I would put a decent amount of trust into a modern SF drive. I think they're all fairly comparable at this point.

You need to get your facts straight. Plextor was not "very late" to the consumer SSD market. They have been in it for years now. They had the M1, M2, M3, and now M5 models.

And it is foolish to trust a "modern Sandforce" SSD. Most Sandforce SSDs still have broken TRIM, and have had for months. In the past week or so some of them are getting around to fixing TRIM, but still quite a few have not yet released v5.03+ or v5.04 firmware to fix the TRIM bug. Sandforce has years of history of having bugs and not fixing them for months or years.

Sandforce should be avoided. The best SSDs are Plextor M3P for best performance and quality, or Samsung 830 for a good value.

 

[JoeComp]

In post #5 I made some suggestions, and asked for more info and a link to a dumpfile where we could get more precise info to help. Another here decided to turn it into a "Reasons not to buy anything that has a Sandforce controller" dissertation/pissing match.

When I said you sounded like a politician (or an OCZ marketing droid) you did not have to take it to heart and become one completely! Heh, you start a flame war and then blame me, who responded to the OP in post #2 with good advice, and I have consistently echoed that same advice in my subsequent posts, even while responding to your posts which are sprinkled with incorrect information.

 

[FrgMstr]

Bunch of adults acting like children are about to get themselves banned from this forum for good. So I suggest you talk about the subject rather than each other.