I have had the 240gb M4 for nearly 6 months and I have never had one problem with it. I don't doubt that it happens to people, but I think it is more directly related to the motherboard chipset then to the SSD. My MB isn't extremely new and doesn't get me 'max' throughput, but I think that has kept it 100% stable.
Do you have over 5000 hours on yours? D/L something to read the time on your drive. I bet you are no where near the 5k mark that is why.
My M4 has only 1100ish hours on it. This seems to only affect those with 5200+ hrs on their drive. I saw a lot of people saying they had intel for their controller but I don't think that is the problem. That and crucial admited they have a fix for it so I am guessing its a simple code flip somewhere.
I wonder how all these people got to 5000 hours. If you subtract the amount of time the drive spends asleep or powered off totally, you're not left with much more than 8 hours a day. And at 8 hours a day it would take two years for this issue to show up! Even if it was somehow running 24/7 without ever sleeping it would still take more than 6 months.
It's not impossible. If it's your boot drive, it won't enter sleep mode unless you sleep the OS too. Some people like to leave their computer on 24/7 (e.g. for downloads, rendering etc) - let alone if you use it in a some kind of a server.
Definitely a lot hours and for many users, it will take years to achieve.
I don't think Plextor, Samsung or Intel SSDs w. Intel controllers have had such problems. A website that has taken a leadership in SSD tech overviews and reviews SHOULD REPORT WHICH SSDs WE CAN COUNT ON instead of making not-very-useful comments like "Pretty much every SATA 6Gb/s SSD offering has experienced some issues"
PRETTY MUCH: Not a useful term in a tech article. Ever.
Yeah, they don't want to piss of their advertisers haha. Samsung & intel have better records. Although to be fair until this issue so did Corsair. I love my intel 320 even tho its only 3Gb/s.
This has nothing to do with our advertisers. Our ads are provided by a third party so we don't even know who the advertisers are ;-) I use adblock on top of that so I'm 100% clueless about our ads as I have never seen them.
Like I said in my reply to rochlin, we have no reliable data about what is the most reliable and trouble-free SSD. If we had, you would know about it. Heck, I would love to know about it too! However, I don't think it would be good journalism to say what is good and what is bad SSD based on forum posts.
The only bias on Anandtech is the links to the Dailytech website.
Seriously, Anand, Kristian, you can't try to claim Dailytech is impartial and unbiased, and because of that they really shouldn't be linked on your website.
I agree. While I think I have never read a whole DT article, I don't think they deserve the visibility they get. I guess it was different when DT launched but nowadays, they are totally different. We have no impact on what they post and vice versa. If they post crap, it's linked to our site automatically, even if we strongly disagree with the article.
We have some plans on improving pipeline visibility which may (and hopefully will) reduce or even remove the visibility of DT on AT. I'll be sure to bring this up in our next chat.
DT does have interesting things once in a while, but the users are so caustic I just stopped reading it. The article just kind of provides the topic for the ensuing flame war.
They have when it has come up.. But as a website that relies primarily on actual hands-on testing.. even running one of these drives 24/7 would take 200 days to bring this issue up.. so they probably have no first hand evidence to report on.
The problem is that there are no studies on this. Google "[enter the SSD here] issue" and you will find something, regardless of the SSD. It's always hard to come up with an objective article without any concrete figures. How can we prove that one SSD is more reliable than the other without having any backup?
Anand tried to prove and find the SandForce bug for months. This Crucial issue is even harder because it requires 208 days of 24/7 running. Besides, how do we know that never SSDs, e.g. Samsung 830 and OCZ Octane, don't suffer from a similar issue? Only time will tell, I guess.
If you want my opinion on what SSD to get, it's Samsung 830. From what I have seen around the forums, it has no widespread issues. Then again, I thought the same about Crucial M4 a few days ago, and Intel 320 before the 8MB bug.
I don't want to sound like a partypooper but it's still missing on-time tests. This Crucial issue isn't about the amount of data that is written, it's purely on-time related (at least according to Crucial).
Also, it would be better that all tests were done on the same machine and the SSDs were brand new. I know I sound boring again but benchmarks and other tests are just nitpicking. You can question all the results if they are not done under the same circumstances. The last thing we want is an egg on our face for reporting questionable results. We don't like rumors and sensationalism like some other sites do. If we aren't sure about something, we prefer leaving it out :-)
Well, that statement is "pretty much" right. Intel 320 series had the 8mb bug, sandforce 2200 has had a constant stream of issues, and now crucial m4 as well. The only ones left standing are samsung 830 (only 2 months out so far) and intel 510 series.
Intel 320 isn't a 6Gbps drive, and since it has no other outstanding issues, and had its warranty upped to 5 years, I would safely bet that Intel 320 is currently the most reliable SSD available.
This is good criticism, and I hope the author takes note. Anandtech pushes SSDs /hard/, but I don't remember reading too much about the problems that plagued SF drives or any articles that gave appropriately detailed reports on which drives are known or reported to cause stability issues.
I need to discuss with Anand but I do like the idea of "SSD Buyer's Guide" or something alike. I "run" an SSD sticky over at MacRumors and I've seen what people want is an SSD that won't cause troubles, which sometimes isn't easy due to the dozens of SandForce OEMs and whatnot.
just that.. I almost bought the crucial M4 due to its slightly better price.. but the samsung has a better rep and better performance... so, here's hoping it deserves the rep *Knocks on wood*
Really think this is on the same level as the Intel 8mb bug, a non-issue unless you're one of the few who is experiencing it. I have a crucial M4, I'll put on the new firmware, and not worry. Good to see a quick effective response on the issue, overall I would prefer knowing that the company will proactively correct issuws .
Looking at the time to fail though, 5200 hours roughly, when multiplied into seconds that's pretty close to 2^24, wonder if a debug counter set to 24bit integer to monitor up time is not being properly disabled...
When the drive first arrived it suffered from the 'black screen' issue where a laptops/desktops bios initiated too fast for the M4 drive which ended up with a black screen. Soft reset and you could boot. (I experienced this on a HP 8460p).
It took MONTHS for them to fix this issue... but they did in the end.
I'm currently using a 128Gb M4 in my 2560p and NOW this new issue appears. Come on Crucial you're messing with people's DATA!
Damn, I just installed my new 128GB M4 today. I bought on the recommendations from Anandtech forums too. Oh well hopefully this is fixed within 6 months as that is roughly 5000 hours of 24x7 use. I leave my comp on all the time so I will reach 5000 hours in 6 months easily.
Intel, OCZ, Corsair, Samsung, Crucial, Plextor, etc. have all had SSD issues of one sort or another. As Anandtech stated only a few months ago, consumer grade SSDs are "immature tech" and as such people should think twice about jumping in with both feet until the issues get sorted out.
In IT space the only way to have reliability is to buy enterprise products. Intel 510 had zero issues and so any other professional ssd. It's almost impossible for server HDD to die in a reasonable amount of time. In consumer space people will simply buy cheaper product. But if you save money on speed or features you product will not be popular. Testing is the easiest thing to cut, also helps time to market.
It seems like to even experience this issue you would have had to been a first month buyer after release and have never shut down your system since install.
5000 hours is about 7 months. They only came out in March to early April last year. Seems just by the numbers that the vast majority of owners won't experience this if the patch is coming this month. So far I've been very happy with the 512 GB and 64 GB m4 I bought late last year.
I purchased a 64GB Crucial M4 in November and had problems with it right out of the box. I went through the troubleshooting in several posts on different forums, but could never get it working quite right with my AMD system on a Gigabyte 990FXA-UD3 motherboard. I was about to send it back when I got a 2600K for Christmas.
I installed it as the cache drive on a Gigabyte Z68MA-D2H-B3 motherboard and it hasn't given me any more problems. It is overkill for a cache drive, but Windows 7 Professional 64-bit boots on the 2600K system in 18 seconds now, which is down from just over a minute with a 500GB SATA III Seagate Barracuda.
Thanks for letting people know about the firmware update. I will keep it in mind if I ever go back to trying to use the drive on my AMD system. Now if that one only had the caching feature...
I don't get it, why is the firmware counting the hours and why does it have to start doing anything special after 5000? It really worries me that it might have been some "break-on-purpose" thing so the drives last just enough for the warranty to expire and then you'd need a new one. But they messed up the interval, and their "fix" is going to fix just that part.
along with crucial, the intel 510 also uses the marvell controller, plextor and corsair included, in their non sandforce offerings. the sammies are too new to tell re issues. the 510 is not an enterprise drive.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
40 Comments
Back to Article
apexjr - Friday, January 6, 2012 - link
I have had the 240gb M4 for nearly 6 months and I have never had one problem with it. I don't doubt that it happens to people, but I think it is more directly related to the motherboard chipset then to the SSD. My MB isn't extremely new and doesn't get me 'max' throughput, but I think that has kept it 100% stable.Marlin1975 - Friday, January 6, 2012 - link
Do you have over 5000 hours on yours? D/L something to read the time on your drive. I bet you are no where near the 5k mark that is why.My M4 has only 1100ish hours on it. This seems to only affect those with 5200+ hrs on their drive. I saw a lot of people saying they had intel for their controller but I don't think that is the problem.
That and crucial admited they have a fix for it so I am guessing its a simple code flip somewhere.
Shadowmaster625 - Friday, January 6, 2012 - link
I wonder how all these people got to 5000 hours. If you subtract the amount of time the drive spends asleep or powered off totally, you're not left with much more than 8 hours a day. And at 8 hours a day it would take two years for this issue to show up! Even if it was somehow running 24/7 without ever sleeping it would still take more than 6 months.Kristian Vättö - Friday, January 6, 2012 - link
It's not impossible. If it's your boot drive, it won't enter sleep mode unless you sleep the OS too. Some people like to leave their computer on 24/7 (e.g. for downloads, rendering etc) - let alone if you use it in a some kind of a server.Definitely a lot hours and for many users, it will take years to achieve.
DaFox - Saturday, January 7, 2012 - link
7 months isn't that long..MrSpadge - Saturday, January 7, 2012 - link
Running BOINC ;)rochlin - Friday, January 6, 2012 - link
I don't think Plextor, Samsung or Intel SSDs w. Intel controllers have had such problems. A website that has taken a leadership in SSD tech overviews and reviews SHOULD REPORT WHICH SSDs WE CAN COUNT ON instead of making not-very-useful comments like "Pretty much every SATA 6Gb/s SSD offering has experienced some issues"PRETTY MUCH: Not a useful term in a tech article. Ever.
FATCamaro - Friday, January 6, 2012 - link
Yeah, they don't want to piss of their advertisers haha. Samsung & intel have better records. Although to be fair until this issue so did Corsair. I love my intel 320 even tho its only 3Gb/s.Kristian Vättö - Friday, January 6, 2012 - link
This has nothing to do with our advertisers. Our ads are provided by a third party so we don't even know who the advertisers are ;-) I use adblock on top of that so I'm 100% clueless about our ads as I have never seen them.Like I said in my reply to rochlin, we have no reliable data about what is the most reliable and trouble-free SSD. If we had, you would know about it. Heck, I would love to know about it too! However, I don't think it would be good journalism to say what is good and what is bad SSD based on forum posts.
doobydoo - Saturday, January 7, 2012 - link
The only bias on Anandtech is the links to the Dailytech website.Seriously, Anand, Kristian, you can't try to claim Dailytech is impartial and unbiased, and because of that they really shouldn't be linked on your website.
Kristian Vättö - Saturday, January 7, 2012 - link
I agree. While I think I have never read a whole DT article, I don't think they deserve the visibility they get. I guess it was different when DT launched but nowadays, they are totally different. We have no impact on what they post and vice versa. If they post crap, it's linked to our site automatically, even if we strongly disagree with the article.We have some plans on improving pipeline visibility which may (and hopefully will) reduce or even remove the visibility of DT on AT. I'll be sure to bring this up in our next chat.
ddarko - Saturday, January 7, 2012 - link
Oh thank god.Mr Perfect - Monday, January 9, 2012 - link
DT does have interesting things once in a while, but the users are so caustic I just stopped reading it. The article just kind of provides the topic for the ensuing flame war.sprockkets - Sunday, January 8, 2012 - link
They seem to run on the same damn server. Jason Mick pretty much has to add "EDITORIAL" to every article it seems.It is fun to watch the fanbois eliminate/troll each other though.
damianrobertjones - Saturday, January 7, 2012 - link
There were a WHOLE load of posts on the Crucial forum tall those months ago and although I tried to inform people it obviously got ignored.Shame really as the proof was there for all to see.
Now we have a second bug
MrSpadge - Saturday, January 7, 2012 - link
It's Crucial, not Corsair.Namey - Friday, January 6, 2012 - link
I can feeeeel your anger. It gives you focus, makes you STRONGER.Araemo - Friday, January 6, 2012 - link
They have when it has come up.. But as a website that relies primarily on actual hands-on testing.. even running one of these drives 24/7 would take 200 days to bring this issue up.. so they probably have no first hand evidence to report on.Kristian Vättö - Friday, January 6, 2012 - link
The problem is that there are no studies on this. Google "[enter the SSD here] issue" and you will find something, regardless of the SSD. It's always hard to come up with an objective article without any concrete figures. How can we prove that one SSD is more reliable than the other without having any backup?Anand tried to prove and find the SandForce bug for months. This Crucial issue is even harder because it requires 208 days of 24/7 running. Besides, how do we know that never SSDs, e.g. Samsung 830 and OCZ Octane, don't suffer from a similar issue? Only time will tell, I guess.
If you want my opinion on what SSD to get, it's Samsung 830. From what I have seen around the forums, it has no widespread issues. Then again, I thought the same about Crucial M4 a few days ago, and Intel 320 before the 8MB bug.
piroroadkill - Sunday, January 8, 2012 - link
http://www.xtremesystems.org/forums/showthread.php...Sacrificing SSDs for science!
Kristian Vättö - Monday, January 9, 2012 - link
I don't want to sound like a partypooper but it's still missing on-time tests. This Crucial issue isn't about the amount of data that is written, it's purely on-time related (at least according to Crucial).Also, it would be better that all tests were done on the same machine and the SSDs were brand new. I know I sound boring again but benchmarks and other tests are just nitpicking. You can question all the results if they are not done under the same circumstances. The last thing we want is an egg on our face for reporting questionable results. We don't like rumors and sensationalism like some other sites do. If we aren't sure about something, we prefer leaving it out :-)
bryanW1995 - Saturday, January 7, 2012 - link
Well, that statement is "pretty much" right. Intel 320 series had the 8mb bug, sandforce 2200 has had a constant stream of issues, and now crucial m4 as well. The only ones left standing are samsung 830 (only 2 months out so far) and intel 510 series.piroroadkill - Sunday, January 8, 2012 - link
Intel 320 isn't a 6Gbps drive, and since it has no other outstanding issues, and had its warranty upped to 5 years, I would safely bet that Intel 320 is currently the most reliable SSD available.TerdFerguson - Saturday, January 7, 2012 - link
This is good criticism, and I hope the author takes note. Anandtech pushes SSDs /hard/, but I don't remember reading too much about the problems that plagued SF drives or any articles that gave appropriately detailed reports on which drives are known or reported to cause stability issues.Kristian Vättö - Saturday, January 7, 2012 - link
I need to discuss with Anand but I do like the idea of "SSD Buyer's Guide" or something alike. I "run" an SSD sticky over at MacRumors and I've seen what people want is an SSD that won't cause troubles, which sometimes isn't easy due to the dozens of SandForce OEMs and whatnot.piroroadkill - Sunday, January 8, 2012 - link
I don't think are Intel SSDs with Intel controllers at 6Gbps.webmastir - Friday, January 6, 2012 - link
just got done setting up one of these when i read about it yesterday.Araemo - Friday, January 6, 2012 - link
just that.. I almost bought the crucial M4 due to its slightly better price.. but the samsung has a better rep and better performance... so, here's hoping it deserves the rep *Knocks on wood*Belegost - Friday, January 6, 2012 - link
Really think this is on the same level as the Intel 8mb bug, a non-issue unless you're one of the few who is experiencing it. I have a crucial M4, I'll put on the new firmware, and not worry. Good to see a quick effective response on the issue, overall I would prefer knowing that the company will proactively correct issuws .Looking at the time to fail though, 5200 hours roughly, when multiplied into seconds that's pretty close to 2^24, wonder if a debug counter set to 24bit integer to monitor up time is not being properly disabled...
damianrobertjones - Friday, January 6, 2012 - link
When the drive first arrived it suffered from the 'black screen' issue where a laptops/desktops bios initiated too fast for the M4 drive which ended up with a black screen. Soft reset and you could boot. (I experienced this on a HP 8460p).It took MONTHS for them to fix this issue... but they did in the end.
I'm currently using a 128Gb M4 in my 2560p and NOW this new issue appears. Come on Crucial you're messing with people's DATA!
damianrobertjones - Friday, January 6, 2012 - link
P.s. The Black screen issue was apparent on the Crucial support forum with NO acknowledgment at all.josephjpeters - Friday, January 6, 2012 - link
I thought all the OCZ haters said Crucial M4 was the way to go?Azsen - Friday, January 6, 2012 - link
Damn, I just installed my new 128GB M4 today. I bought on the recommendations from Anandtech forums too. Oh well hopefully this is fixed within 6 months as that is roughly 5000 hours of 24x7 use. I leave my comp on all the time so I will reach 5000 hours in 6 months easily.Beenthere - Saturday, January 7, 2012 - link
Intel, OCZ, Corsair, Samsung, Crucial, Plextor, etc. have all had SSD issues of one sort or another. As Anandtech stated only a few months ago, consumer grade SSDs are "immature tech" and as such people should think twice about jumping in with both feet until the issues get sorted out.MrSpadge - Saturday, January 7, 2012 - link
Yeah, just jump in with one foot: use the SSD as a cache, so you don't loose any data if it fails.Herp Derpson - Sunday, January 8, 2012 - link
In IT space the only way to have reliability is to buy enterprise products. Intel 510 had zero issues and so any other professional ssd. It's almost impossible for server HDD to die in a reasonable amount of time. In consumer space people will simply buy cheaper product. But if you save money on speed or features you product will not be popular. Testing is the easiest thing to cut, also helps time to market.THizzle7XU - Saturday, January 7, 2012 - link
It seems like to even experience this issue you would have had to been a first month buyer after release and have never shut down your system since install.5000 hours is about 7 months. They only came out in March to early April last year. Seems just by the numbers that the vast majority of owners won't experience this if the patch is coming this month. So far I've been very happy with the 512 GB and 64 GB m4 I bought late last year.
mrcaffeinex - Sunday, January 8, 2012 - link
I purchased a 64GB Crucial M4 in November and had problems with it right out of the box. I went through the troubleshooting in several posts on different forums, but could never get it working quite right with my AMD system on a Gigabyte 990FXA-UD3 motherboard. I was about to send it back when I got a 2600K for Christmas.I installed it as the cache drive on a Gigabyte Z68MA-D2H-B3 motherboard and it hasn't given me any more problems. It is overkill for a cache drive, but Windows 7 Professional 64-bit boots on the 2600K system in 18 seconds now, which is down from just over a minute with a 500GB SATA III Seagate Barracuda.
Thanks for letting people know about the firmware update. I will keep it in mind if I ever go back to trying to use the drive on my AMD system. Now if that one only had the caching feature...
Visual - Monday, January 9, 2012 - link
I don't get it, why is the firmware counting the hours and why does it have to start doing anything special after 5000?It really worries me that it might have been some "break-on-purpose" thing so the drives last just enough for the warranty to expire and then you'd need a new one. But they messed up the interval, and their "fix" is going to fix just that part.
m.oreilly - Monday, January 9, 2012 - link
along with crucial, the intel 510 also uses the marvell controller, plextor and corsair included, in their non sandforce offerings. the sammies are too new to tell re issues. the 510 is not an enterprise drive.