Intel Pulls SSD Firmware Day After Release 125
CWmike writes "Intel has pulled a firmware upgrade it released on Monday for its X25-M consumer solid-state drives after users complained that the software caused crashes. The company on Monday made available a software package called SSD Toolbox to monitor and manage the performance and health of X25-M SSDs on systems running Windows 7. The package included a firmware upgrade and software called SSD Optimizer that included diagnostic tools to help keep the Intel SSD running at high performance. 'We have been contacted by users with issues with the 34-nanometer Intel SSD firmware upgrade and are investigating. We take all sightings and issues seriously and are working toward resolution. We have temporarily taken down the firmware link while we investigate,' an Intel spokesman said in an e-mail. The spokesman declined to comment on when the company would issue updated firmware."
Re: (Score:2, Funny)
A Saturn car??? they still make those?? = )
No, a Saturn V. They haven't got all the bugs out of the Ares yet.
Development process is flawed (Score:5, Insightful)
I'm starting to think that the whole SSD market is a prime example of the modern corporate development mentality of pawning off beta testing to the general public. It's clear that SSDs are not ready for general release, but companies do not want to spend the time or money to validate them against specifications or ensure that they work properly for their particular purpose. Let the public pay for your beta test program. It's a lot cheaper.
Re:Development process is flawed (Score:5, Informative)
Re: (Score:2)
Sounds familiar [google.com].
Re: (Score:1, Insightful)
So what you're saying is you bought a used product from a company that did not offer you a warranty, and expected the manufacturer to offer you, someone who they did not sell a product to, a warranty on a product that you did not buy from them?
Re: (Score:3, Insightful)
Re: (Score:3, Informative)
Moral of the story... (Score:2)
Don't go all googly-eyed and install an update two minutes after it's released. Wait a few days until the idiots have tested it. ...but you knew that, right?
Re: (Score:2, Funny)
I'm planning on getting an SSD drive in around 10 years time once the technology is mature!
Re: (Score:2)
I considered CD writers before that point not fit for general use. I'm actually amazed people bought those drives. No other choice I guess.
Interestingly my old LiteOn with BURNProof (one of the first drives with that tech) is still working after all these years. But the amount I paid for it can buy me 10 or more DVD writers today (but these modern DVD burners so far only seem to last 1.5 to 2 years
Re: (Score:1)
Anyone remember how long it took before PC CDROM drives had buffer underrun protection?
I considered CD writers before that point not fit for general use. I'm actually amazed people bought those drives. No other choice I guess.
Lessee, in those good ol' days i worked the helpdesk for 6 different brands of cd-burners at the same time(as well as a whole lot of other pc peripheral related junk) ;-)
In all fairness the early IDE burners were fine, you just had to pay attention to what you were doing. SCSI was a lo
Re: (Score:2)
Supposedly the first 1x CD-R drives had blanks that ran $20 a piece and had 32K of buffer memory, current DVD-RW drives have 2MB. A single buffer underrun ruined the entire disk. On Win 3.11 or MacOS 7.1, I imagine the entire affair was a bit terrifying. It didn't help that the drives themselves ran about $10,000.
Re:Development process is flawed (Score:5, Insightful)
When there isn't enough competition around, companies don't have to worry about Quality - the people will buy whats available, and if no one is offering a higher quality product, the low quality product will still sell.
If this market is to mature they need a company to step in with the emphasis on quality.
Re:Development process is flawed (Score:5, Interesting)
"If this market is to mature they need a company to step in with the emphasis on quality."
Funny, most people would think that company could be Intel. I would be very surprised if this issue was in any way expected by Intel. There were a few articles on the thorough testing performed on the G1 (firmware). With the G2 Intel seems to have lost some of that.
Re: (Score:1, Funny)
Perhaps they missed a "customer story" and the engineers who saw the problem were probably rebuffed with "that's not one of the customer stories" or "the customer hasn't asked for that". Or, perhaps there were no design docs (because, "customers don't buy design docs") so nothing to review except code (which, in any project, is a horribly inefficient and ineffective way to catch architecture or design related problems).
/rant_really_hea
Re: (Score:2)
Re:Development process is flawed (Score:4, Informative)
Re: (Score:2)
Re: (Score:2)
Re: (Score:3)
Re: (Score:1)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
http://www.storagesearch.com/ssd-2p5.html [storagesearch.com]
Just a few for your desktop.
At the consumer end its a cartel. Nobody is going to drop the $$$$ flow too fast.
Re: (Score:2)
If this market is to mature they need a company to step in with the emphasis on quality.
FusionIO?
Oh wait - look at the price!
Re:Development process is flawed (Score:5, Funny)
I guess it's the "cheaper" "cheaper" alternative to shipping your whole QA department over to India.
Re: (Score:2)
Is there a VAR that does good validation? Seems like a market opportunity - I'm not buying SSD's myself out of good old-fashioned distrust of new-fangled things I can't hear spinning. ;) My slow, archaic, RAID-1 disk sets are at least free from surprising bugs. That is, at least after a few rounds of Seagate fixing the drive firmware.
Re:Development process is flawed (Score:5, Insightful)
Paranoid much? There may be companies out there that haven't got a lot to loose and can play that testing game. Intel is certainly not one of them. Anyway, SSD's have been on the market quite a while, although market penetration was always low. And do you think that OS support for TRIM would be there if we had to wait for another year?
Anyway, let's wait and see what causes the (alleged) problems and we'll know what to think of it. It's a bit early to put this to corporate greed. These are complex products.
Re: (Score:2)
Re:Development process is flawed (Score:5, Insightful)
I think it's more that nobody is taking seriously the fundamental differences between hard drives and flash. Nobody has really stopped to do a comprehensive assessment of what existing assumptions embodied in our software and users will be broken by flash memory that is asymmetric in both access speed and access granularity. As a result, the pre-Intel flash SSD controllers made really stupid trade-offs, and they ended up with drives that were less suitable for the consumer market than ordinary hard drives. Once Intel made everybody realize that latency and IOPS mattered a lot more to consumers than throughput, people moved on to the next difference, and started complaining about the lower write performance of a nearly full SSD. Even today, I still see people referring to it as a "bug", when it is nothing more than an inherent difference from the spinning platters of hard drives. Smart garbage collection (which requires smart OS support) is a way of hiding the limitation, but the lack of it isn't a bug any more than a hard drive with a small cache is faulty. It just has obvious room for improvement.
Re: (Score:2)
Remember kids - SSD garbage collection works the same way as Java. You ideally you want ~50% of the heap to be empty. ;)
Re: (Score:2)
Whatever, i've never had a pro
Re: (Score:3, Interesting)
It's not flawed so much as inevitable. A portion of the market will jump at the first example of a promising technology that ships. Being the first mover in a particular space holds special significance and advantages for companies competing for market share. The thinking goes that quality can be worked on iteratively through generations of product and there will never be a time when you reach perfect quality anyway.
Moral of the story: If you don't want to beta test products for corporations, then don't buy
Re: (Score:2)
Later on they figure out what corners they can cut, so the stuff doesn't last way past warranty
Truth is, there's not much point building some stuff to last that long. For example if most of your CD writers built
Re: (Score:1)
Re: (Score:1)
Of course. At my job, we get target release dates shoved down our throats all the time. We're constantly de-scoping products in order to meet release dates. It is more important to release by the target release date than it is to produce a product that meets our customers' needs.
The product I am working on right now has been de-scoped to the point that it will be completely useless to our customers. They've said so, and the engineers on the project understand why it won't work for them. But, that doesn't ma
Re: (Score:2)
Re: (Score:2)
I tried raising this on the OCZ forums, it wasn't even specific to OCZ products - I merely stated that "SSD's are finicky" which they absoloutely are and have been, it's not close to a simple, plug, play, forget - technology.
They are getting better and overall OCZ have done a good job with support but the vehement fans of the products absolutely not only blindly defended SSD's overall but took it as some kind of a personal attack, it's ridiculous.
FWIW I'm an early adopter, I've got 2 OCZ drives and 2 Intel
Re: (Score:1)
I have seen this, too. The SSD astroturfing club has a large membership. I seriously wouldn't be surprised to find out that OCZ was keeping these folks in the latest drives in exchange for their support on web forums.
I recently purchased a few of the OCZ Solid drives, based on the widely ill-reputed JMicron controller. I really haven't seen any problems with them at all yet. They're really fast compared to the platter drives in the systems I have put them in. Meh... we'll see..
Poor, poor quality control (Score:2, Funny)
I installed this firmware and I've never had more problems. My machine reboots 5-6 times a day and there is no end in si
Re: (Score:1)
I've installed the firmware too (unfortunately, just before the first few bad reports appeared).
But i haven't encountered any issues so far. Running Windows 7 x64 Professional.
I didn't reboot or powercycle my machine, though.
Luckily, i also have a backup :)
Re: (Score:2)
Blame the bean counters and marketingdroids. Engineering will not have been behind the push to release anything they didn't consider fit for purpose (which I'd guess in the case of firmware means a minimum of not turning your device into a brick)....
Not just Intel (Score:4, Informative)
Crucial's M225 (I own the 128GB version) 1711 firmware had significant bugs and was quickly yanked. In order to upgrade to the latest 1819 you have to downgrade back to 1571.
http://www.crucial.com/support/firmware.aspx [crucial.com]
Seems as if most consumer SSD products are still a bit in the "beta" stage.
It's still new techonology (Score:2)
It's still new techonology. Compared to the evolution from the first hard-drives and the first 1x CD-ROMs on the market, the SSD technology is still somewhat more reliable. I myself find myself waiting for this technology to mature a bit more before introducing it on my company computers.
Re: (Score:2)
There are Winchester drives and 1x cdroms that still work fine today. I doubt that in 25 years ANY of the SSDs built today will work.
Re: (Score:2)
Last I heard, flash manufacturers were shooting for "5 9s at 10 years", which is to say that they were going to retain 99.999% of cell data after 10 years. Flash is made of very low leakage capacitors, and every capacitor leaks. If your device never refreshes a cell in 25 years, I'd expect more than one failure. I don't know what the distribution looks like, nor what their
Re: (Score:1)
Re: (Score:1)
And then on the other hand, neither of those devices is particularly worth using anymore (a bottom end 3.5 inch hdd will have more recent interfaces, use less power, have more storage and be faster, a craptactular DVD drive won't take forever and a day to do everything).
Smart Machines (Score:2)
Re: (Score:2)
Well, at least I know you are ok with a tooth brush. Beyond that....
Re: (Score:3, Interesting)
Doesn't that defeat the purpose of standards?
That's like saying that we should get rid of the x86 instruction set and just use the micro ops. A layer of abstraction helps and is required for a standard to work.
Re: (Score:2)
the only reason they are using controller wear-leveling in the first place is that the popular operating systems of the time did not do so themselves
really the drive knows nothing about the filesystem and how it wants to store it's data, let the os decide how to wear level, it knows more and can make more informed decisions.. seriously the only reason I don't have an SSD drive already is this hardware wear-leveling crap which is a completely unsuitable way of fixing things.
Re: (Score:1)
They do it internally for a couple reasons:
- The file system does not handle it directly.
- The file should not handle it directly. If you do, a reformat destroys all of the wear-levelling data.
Re: (Score:2)
The file should not handle it directly. If you do, a reformat destroys all of the wear-levelling data.
And something akin to SMART but with added retention of how many times each sector has been written to is an issue how? would stay persistent even with formats and would help the OS determine the best places to write also.
The file system does not handle it directly.
If you use windows then yes.. but more accurately what I should have said is that the drives should allow you to access them in raw mode, hell even default to wear leveling in the controller if you want.. but allow us to disable it when we are using something good enough to do better.
Re: (Score:2)
It needs a firmware because the technology that controls read/write on the physical medium is not open source, and is probably not common across all drives. For SSD's, you have several different kinds of Flash that are in use, along with several different read/write controllers. The firmware is simply a low-level driver that translates between these read/write machine code and the standardised SATA interface. Some of the more intelligent firmware will also do some kinds of optimisation and load balancing on
Re: (Score:1)
I don't know about anyone else, but I am getting damn sick and tired of devices that NEED firmware. Why does every little peripheral need to contain LOGIC!!! I want DUMB machines, damn it! Why can't an SSD simply be a "mass storage device"?? Let the OS worry about wear leveling, etc.
A firmware upgrade makes it possible to make improvements or fix any bug discovered after a product was shipped.
It saves money to manufacturers (reducing RMA numbers), and time, and possibly downtime for clients.
I own a 32GB SLC Samsung SSD that as far as I know doesn't support firmware updating. That is terrible for me because I'm stuck with an expensive SSD that will never have TRIM support. If I had bought a different brand I could just get the firmware when they implement TRIM and I be happy.
I really
Re: (Score:1)
Humm... wear leveling in software ... wait until an OS misconfiguration, an OS bug or a virus ruins your SSD by writing thousand of times the same sector. I am not talking about losing your precious data which should be backuped but losing the whole disk. And don't even think about using the warranty. SSD makers are not suicidal. The first thing they would had to their product is a counter of writes per block in order to prove that the failure was not their fault.
Re: (Score:2)
Well, it used to work like that back in the MFM/RLL days for hard drives. The drive was a box of motors and steppers, and the logic to drive it all was on the controller card. Not sure if I want to go back to those days again. Especially since a controller that only knows how to control the bits of a hard drive would have absolutely no idea on how to talk to a SSD.
vs Pricing (Score:1)
My X-25M had this problem (Score:1)
"Closed Source Filesystem" (Score:5, Interesting)
This post really needs a link to:
http://lwn.net/Articles/355149/ [lwn.net]
"Do you want to trust your data to a closed source file system implementation which you can't debug, can't improve and — most scarily — can't even fsck when it goes wrong, because you don't have direct access to the underlying medium?"
This is what you get with a flash drive at the moment unfortunately - a closed source filesystem that presents a single "file" as a block device over sata. And this firmware update is a filesystem driver change. Ouch.
Re: (Score:2)
And so many are keen on cloud computing.
I have no access to the source code of my HDD, and I don't care as long as it works as expected and keeps working for long enough. I have no access to the "source code" of my CPU either. So what? I wouldn't even be able to debug it even if I had it, or fix my CPU.
Same goes for an SSD. For now I don't think they're cheap and reliable enough for me. But maybe next year or the year after.
Re: (Score:2)
As alluded to in the link, modern hard drives transparently remap sectors when they go bad, so that when you ask to read the known bad one you actually get something from a completely different area altogether (wherever the spare sector space is at). All that happens via a closed-source firmware implementation you can't debug, improve, or fix if it goes awry. We're already being presented with a logical view of our drive hardware that maps to underlying physical hardware in a way we have no control over,
Hey Intel, how about unlocking the read speed (Score:2)
So the first X-25M were blazing fast reads and pretty amazing at small file writes but somehow ground to a halt at 75 MB/s - about the speed
of a really good hard disk 3 years ago.
I hear the G2 drives can do 100 MB/s with a firmware update. Now, I know that you want to save the really fast writes for your SLC version
cause that much more moo-lay but c'mon - 100 MB/s for a $400 drive, that's artificially limited.
No thanks, until you let the drive perform to its proper capability ( i'm guessing 160 MB/s sequent
Re: (Score:2)
There are extremely few read-world apps that require sequential writes at speeds even reaching 80MB/s, much less 100MB/s. Sure, you can do it with a file copy, but how often does that come up as something you're waiting for? You certainly can't reach that speed with anything network-based. The fact that Intel optimized for smaller writes instead was absolutely the right thing. Sure, the Patriot or OCZ drives manage fast write speeds [anandtech.com] just as you describe. But they're mundane on random reads and writes [anandtech.com],
Re: (Score:2)
Tradeoff my ass. It's an artificial limitation.
Re:Hey Intel, how about unlocking the WRITE speed (Score:2)
I'd made a typo in the original post - it should have been WRITE speed not READ.
Anyway, if you still disagree with me, have a look at http://hothardware.com/Articles/Intel-34nm-X25M-Gen-2-SSD-Performance-Update/?page=1 [hothardware.com]
Amazing what Intel managed with a firmware update - up to THIRTY PERCENT improvement in write speed.
I'm betting their design still has significant headroom left in the sequential writes department but , as I said before, they're protecting their premium SLC, the X-25E
Re: (Score:2)
Yes, I was replying to what you meant, not what you said. A week ago, I'd have been completely with you; the 80MB/s cap was obviously a locked speed rather than a design limitation, as pointed out back in April [pcper.com]. I'm not so sure there's as much headroom in the current 100MB/s limit though. While we know that it's possible to hit closer to 160MB/s in raw speed here based on similar units, the approach Intel is using to deliver its higher random performance has to hurt sequential performance too; that's the
Re:Hey Intel, how about unlocking the write speed (Score:2)
I took a look at the AnandTech link you posted and from the data shown, I wonder if you quite truly understand the meaning of "mundane".
Here's a definition from dictionary.com:
2. common; ordinary; banal; unimaginative.
The link
http://www.anandtech.com/storage/showdoc.aspx?i=3667&p=6 [anandtech.com]
shows that, for HDDs, the random write rate at 4kb is 0.8 MB/s ( Seagate Momentus 5400.6 ) and 1.5 MB/s ( Velociraptor 300 GB).
Now, since I use primarily a notebook, the Velociraptor is a non-starter.
I'm also excluding all the
Re: (Score:2)
I'm glad you're visiting dictionary.com to improve your reading comprehension. While you were occupied insulting me for no good reason, you seem to have missed that the discussion context here was comparing among the somewhat similar MLC drives on the market right now. The post I replied to might lead a reader to believe that the superior write speed of the OCZ and Patriot drives translated into them being better overall than Intel's design. I was pointing out that while they excelled at that category, n
Re: (Score:2)
My reading comprehension is just fine but thank you all the same. And, if you thought I was being insulting, then you're easily offended.
But, since some clarification appears to be in order, I'll give it a shot.
You wrote: "they're mundane on random reads and writes" and linked to (http://www.anandtech.com/storage/showdoc.aspx?i=3667&p=6), which compared 2 of the better performing HDDs with a number of SSDs, as you well know.
Since SSDs are still quite the expensive novelty, notwithstanding the growing po
Re: (Score:2)
Not so fast. SSDs don't write a page at a time but a BLOCK ( or several). Also, your calculations are for writing but you didn't factor in the erase time which is 3 times as long.
So, if your math is right, the max speed would be a lot lower - which is clearly wrong.
Also, you don't explain how the other controllers which have fewer channels can have such higher sequential write speeds.
Re: (Score:3, Insightful)
Re: (Score:1, Informative)
Re: (Score:3, Insightful)
It's a "feature" of the system. It means that there is no way to play favorites with official moderators. It does mean though that idiots occasionally get mod points and proceed to blow them all on modding someone down with whom they disagree.
But really, don't worry about it - idiots are still outnumbered by decent moderators without vendettas. Furthermore, this type of use is generally fixed by the metamod system: slashdot.org/metamod.pl.
I've been around for years, and this issue has always been around. Af
Re: (Score:2)
I'm just offended that someone is allowed to label several of my comments "troll" simply because they don't like me.
You must be n- oh, wait, never mind.
Re: (Score:1)
Re: (Score:2)
It's a right of passage. Once the 8-digit newbs sign up, all your posts will be modded +5 funny and virgins will suddenly find you desirable.
That's what happened to me when the 7-digit newbs started appearing around here.
Re: (Score:1)
That's "rite of passage".
Re: (Score:2)
Mmm.... you're right, I think. And sometimes you can pin down exactly who it is, too.... like, for example, if you suddenly have 5 posts modded as "troll" (that clearly aren't trolls) the day after you call somebody out for being a self-important twat with no clue what he's actually talking about... (look at my posting history for an example, you get a cookie if you can guess who it is)....
It happens. Some people are idiots. Theoretically the meta-mod system is in place to mitigate that kind of asshattery.
Re: (Score:1)
Re: (Score:1)
Re: (Score:1, Troll)
Don't sweat it; like oodaloop said, just keep posting informative and interesting comments. As to the why, I suspect I know who the culprit is, as you only have one freak [slashdot.org].
BTW, don't try for funny. Funny gains no karma, but attempted humor often gets modded troll or flamebait.
Re: (Score:2)
Have you tried directing it to /dev/null? They may be slow in offering up helpful suggestions, but boy are they great listeners!
Re: (Score:1)
Re: (Score:1, Informative)
He's just being a jerk. In Unix-like systems, /dev/null is a "null file" and disregards any data written to it.
Re: (Score:3, Insightful)
This is slashdot, not "ubuntuforums
Re: (Score:2)
Cool. I think many hard-core geeks have been told to RTFM at least once in their lives, maybe not in those exact words. I know I have (early college years, at least probably 2-3 times) and I think I'm better for it. I started realizing a lot of the issues that ESR describes in the article I linked to, especially as I realized that if I didn't start reading manuals I'd literally be taking up all the time of these people. For a lot of things, you just have to study the material, and you get more confident the
Re: (Score:2)
Or at least know how to Google.