Intel promises firmware fix for 320 SSD 8MB bug

VulcaN

[H]ard|Gawd
Joined
Aug 22, 2001
Messages
1,911
haha sorry for the new thread it never crossed my mind this would have been a wide spread issue or I would have searched first.

When I decided to get a new intel 320 series drive it was because of posts I read on this forum suggesting that it was faster AND safer.

I've had countless unsafe shutdowns over the years due to lockups when overclocking, but the last time I remember it being an issue was a long long time ago, usually only when data was still being written to a mechanical harddrive.

Welp heres to the next 8 hours of reinstalling, patching, installing apps and more patching only to find out I still missed something later this week when I go to use it!
 

john4200

[H]ard|Gawd
Joined
Oct 16, 2009
Messages
1,537
When I decided to get a new intel 320 series drive it was because of posts I read on this forum suggesting that it was faster AND safer.
Yes, I thought the same about the 320, based on the X25-M being very reliable. It is disappointing that Intel did not do more thorough unsafe shutdown tests on the 320, since they just added power-loss-protection capacitors and presumably modified the firmware accordingly. For a drive rated for 50,000 power cycles and touted for power-loss protection, it is ironic that Intel screwed that up.

At least Intel is promising a firmware fix for the problem, with an ETA of later this month.
 
Last edited:

VulcaN

[H]ard|Gawd
Joined
Aug 22, 2001
Messages
1,911
I was able to bandaid the drive until the fix comes by using the secure erase utility in parted magic.

Needless to say I installed my OS on the older G2 drive and will be using the 320 for games and applications - at least I can rest easy if theres an unexpected shutdown.
 

Happy Hopping

Supreme [H]ardness
Joined
Jul 1, 2004
Messages
6,730
I won't even touch vertex or any other brand name even if they are free. After reading about OCZ and consair, you can pay me and I won't use those brand names
 

Happy Hopping

Supreme [H]ardness
Joined
Jul 1, 2004
Messages
6,730
that is pure conjecture. i do like that you threw in the "unsafe shutdown" bit though, that does keep people from being too alarmist. thank you.

there have been what, seven individual users who have posted on this miniscule issue in that thread at intel? i believe the count is still at seven cases. at least now we can call them *confirmed* cases.
even if they have *only* sold 100,000 units, what does that put the failure rate at? a ridiculously low number.

glad they reproduced it. and it is easily fixable. if there was any merit for ANY panic of any form, they would have issued a "hot fix".
Which they did not.
they have agreed there is a issue, and they have a fix already. It is such a ridiculously absolutely low infinitesimal number of failures that they haven even bothered to release the fix until they roll it into their next firmware.
will you buy a Toyota, in which they said their brake will only fail once every 1 million times? This kind of management from Intel reminds me of the way they handle their Pentium bug

A better question is, I am paying for this, I work hard for my $$, why would I or any1 else for that matter, put up w/ this?

I use this computer at least 12 hr. a day, I can't afford the possibility of this type of absurd failure. Also, every time I soft boot, the hard drive goes to power cycle, so to say there is a low no. failure, I am having a hard time believing it.
 

Red Falcon

[H]F Junkie
Joined
May 7, 2007
Messages
10,259
Good find, will make things much easier for systems without optical drives.
 

Nocturnal

Gawd
Joined
Jul 20, 2006
Messages
790
Awesome. Just sent my 320 back via FedEx. I hope when I get my RMA it'll already have it.
 

lil layzie

Gawd
Joined
Jan 7, 2004
Messages
755
I've been running my 320 drive for nearly 3 months now and was really afraid of this bug. Glad to see the new firmware out.
Thanks for the update.

If you want to avoid burning a CD and boot the ISO from a USB drive, I found that the firmware update runs fine from a YUMI formatted multiboot USB:

http://www.pendrivelinux.com/yumi-multiboot-usb-creator/
Thank you for this and saving me from the extra work of googling how to do this.
 

Burner27

Supreme [H]ardness
Joined
Oct 23, 2000
Messages
6,386
Can this FW update be applied to those who currently experience the 8MB 'bug' to restore the full capacity of the drive?

Edit: Found Answer



What should I do if I have already experienced this issue?

If you have already experienced a drive failure or encounter this problem before the firmware update was released, please contact your Intel representative or Intel customer support (via web: www.intel.com or phone: www.intel.com/p/en_US/support/contact/phone) for an SSD replacement. An alternative option is to use the Intel ® SSD Toolbox or similar tools to perform a secure erase in order to restore the SSD to an operational state; all data will be erased. After secure erase, update your SSD with the new firmware. The firmware update will not recover user data.
 
Last edited:

john4200

[H]ard|Gawd
Joined
Oct 16, 2009
Messages
1,537
Can this FW update be applied to those who currently experience the 8MB 'bug' to restore the full capacity of the drive?
As you said, the solution is apparently to secure erase the SSD before doing the firmware update.

However, I have seen some reports that the Intel SSD toolbox will not recognize the drive when it is in the 8MB state.

Has anyone succeeded in secure-erasing an 8MB-bug Intel 320 with the Intel toolbox?

If the Intel toolbox does not work, some people have had success using secure erase from parted-magic. There is a tutorial on using parted-magic for secure erase here:

http://www.ocztechnologyforum.com/f...d-Click-Method&p=578170&viewfull=1#post578170

Ironically, the tutorial is from OCZ, but parted-magic is free software and will work on any drive.
 

Happy Hopping

Supreme [H]ardness
Joined
Jul 1, 2004
Messages
6,730
Here's the kicker: say we download this firmware today, who's to say that this new ver. of firmware does not have some other problem, like another problem worse than the 8 MB bug, surfacing in a month or 2?

Meaning that we really have to wait it out 1 to 2 mth. to see if this web site is stable and no new problem
 

Red Falcon

[H]F Junkie
Joined
May 7, 2007
Messages
10,259
Here's the kicker: say we download this firmware today, who's to say that this new ver. of firmware does not have some other problem, like another problem worse than the 8 MB bug, surfacing in a month or 2?

Meaning that we really have to wait it out 1 to 2 mth. to see if this web site is stable and no new problem
That could be said for every single firmware release ever made for all hardware in existence.
Nothing is perfect, we just do what we can.
 

PC_User

Supreme [H]ardness
Joined
Sep 15, 2007
Messages
5,022
I just tried to update the firmware and got the following message: "No Intel SSD attached or installed."

Anyone know what this can mean or how to resolve?
 

Burner27

Supreme [H]ardness
Joined
Oct 23, 2000
Messages
6,386
Did you put the controller on your mobo back to IDE or compatibility mode instead of AHCI mode?
 

PC_User

Supreme [H]ardness
Joined
Sep 15, 2007
Messages
5,022
I just tried to update the firmware and got the following message: "No Intel SSD attached or installed."

Anyone know what this can mean or how to resolve?
Problem resolved. For anyone who experiences this issue, the drive must be set as a "master" drive (master sata port) rather than a "slave" drive.

Plugging it into a master sata port fixed the issue.
 

Red Falcon

[H]F Junkie
Joined
May 7, 2007
Messages
10,259
SATA is only Master, there is no Slave with SATA as it is a single-channel connection only, not dual like PATA.
Not sure why you would put the jumpers to Slave on your SSD.
 

PC_User

Supreme [H]ardness
Joined
Sep 15, 2007
Messages
5,022
SATA is only Master, there is no Slave with SATA as it is a single-channel connection only, not dual like PATA.
Not sure why you would put the jumpers to Slave on your SSD.
I didn't do anything with the jumpers. When I went into BIOS, it showed 2 of my 3 hard drives as "slave" and one as "master." All 3 are plugged into SATA ports (IDE config in BIOS).

I switched plugged one of the hard drive SATA cables into a different port, and then plugged the SSD into that SATA port. Now the system recognizes the SSD as "master" channel 1 and the others as "slave" channel 0.
 

VulcaN

[H]ard|Gawd
Joined
Aug 22, 2001
Messages
1,911
You should take it out of IDE/PATA mode and use AHCI if your motherboard supports it.
 

kalston

[H]ard|Gawd
Joined
Mar 10, 2011
Messages
1,044
Well, updated mine just fine.

If this firmware doesn't add any new issues, then my next SSD shall be an Intel one again.
 

Red Falcon

[H]F Junkie
Joined
May 7, 2007
Messages
10,259
Issue no longer persists since switching SATA ports.

Is there a real-world advantage to using AHCI?
As Dan_D said, about 3%. It's more for the functionality and compatibility, rather than for the minute performance gain.
 

Computurd

Gawd
Joined
Dec 19, 2008
Messages
956
Is there a real-world advantage to using AHCI?
yes, much more than 3%. i have never heard that number.
It has advanced NCQ, which allows for superior write speed. Benches of drives configured without ACHI v with ACHI show a massive improvement in random write speed.
 

kalston

[H]ard|Gawd
Joined
Mar 10, 2011
Messages
1,044
I'm pretty sure that NCQ can affect performance both positively and negatively, although that was on HDDs and it's probably completely different on SSDs. In any case AHCI is a must these days and I noticed more than 3% improvement (with HDDs), even on my old XP rig. Not to mention functionalities and compatibility, yeah.
 

Computurd

Gawd
Joined
Dec 19, 2008
Messages
956
NCQ has a nice feature of uber write combining on an SSD. Also, it allows for much more efficient transfer of files with the many channels of an SSD.
many synthetic tests of real world usage will have only one app loading at a time, with no OS background activity, open media players, email, virus scanners, web browsers, and other stuff. like any monitoring programs you use (AIDA 64) setpoint,etc etc.
this can blur the true advantages of using something like ACHI with its enhanced NCQ. this is where it shines. multitasking in a user environment.
 

Red Falcon

[H]F Junkie
Joined
May 7, 2007
Messages
10,259
NCQ is better in a single/multi-user environment, where as TCQ would be better for a many-user environment.

Like I said, the performance gain from AHCI is very small, but the features gained are great and makes it worth it.

Also, in IDE-mode, SATA-II/III drives are not hot-swappable, it is an AHCI feature only.
 

Computurd

Gawd
Joined
Dec 19, 2008
Messages
956
thanks for the primer, its always nice to be reminded of common facts :)

ACHI is great in that it write combines fantastically. this results in less degradation and less file fragmentation, not to mention faster write speeds.
So, the advantages of NCQ faroutweigh the advantages of hot-swap. it is far more than a 3% increase.
in the long run, the performance being held at a steadier rate will be much more beneficial.
if you want to test the true benefits you would have to do long term testing. Use an OS for an extended period of time with and without ACHI, then compare.

from the Wiki
NCQ is also used in newer solid-state drives where the drive encounters latency on the host, rather than the other way around. For example, Intel's X25-E Extreme solid-state drive uses NCQ to ensure that the drive has commands to process while the host system is busy processing CPU tasks. [3]

NCQ also enables the SSD controller to complete commands concurrently (or partly concurrently, for example using pipelines) where the internal organisation of the device enables such processing.

For example, the SandForce 1200[4] based OCZ Vertex II 50GB drive running on a Dell Perc 5i (which doesn't support SATA NCQ) delivers about 7,000 4k IOPS (50% write) at a controller queue depth of 32 IOs. Moving the drive to the similar Dell Perc 6i (which does support SATA NCQ) increases this to over 14,000 IOPS on the same basis.
wow, just under normal circumstances, 2x performance with 4k IOPS. i have seen this evidenced though many times.

and one more tidbit

This can reduce the amount of unnecessary drive head movement, resulting in increased performance (and slightly decreased wear of the drive) for workloads where multiple simultaneous read/write requests are outstanding, most often occurring in server-type applications.
 

Red Falcon

[H]F Junkie
Joined
May 7, 2007
Messages
10,259
thanks for the primer, its always nice to be reminded of common facts
You're welcome. ;)

As for the AHCI info, for SSDs you are correct.
For HDDs though, the performance boost is very tiny.

For some reason, I thought we were talking about HDDs, not SSDs with AHCI, wow. *head explodes* :D
 

PC_User

Supreme [H]ardness
Joined
Sep 15, 2007
Messages
5,022
NCQ is better in a single/multi-user environment, where as TCQ would be better for a many-user environment.

Like I said, the performance gain from AHCI is very small, but the features gained are great and makes it worth it.

Also, in IDE-mode, SATA-II/III drives are not hot-swappable, it is an AHCI feature only.
Changed it to AHCI mode. Same read/write speeds, so the performance increase (if any) is minimal. I also do not hot-swap my drives so that doesn't affect me.
 

Computurd

Gawd
Joined
Dec 19, 2008
Messages
956
hmm, your raid drivers current, write back selected>?

you should start a thread, we are on a serious derail here :)
 
Top