Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us money per month for bills, and since we don't believe in showing ads to our users, we try to make the money back through forum registrations.
 
  • Locked thread
japtor
Oct 28, 2005

Scionix posted:

I have a crucial c300 drive that came before TRIM support, and I tend to wipe it with HDparm in ubuntu pretty drat often (usually every month or so).

This won't cause any performance degredation, right? I'm so anal about the write performance I'll nuke the thing as soon I start to feel like the drive is getting cluttered :downs:
:stare: Every month? How many extra writes are you doing writing everything back to the drive? :v:

Adbot
ADBOT LOVES YOU

japtor
Oct 28, 2005

every posted:

Also, what are people's thoughts about enabling noatime in Unix systems?
I've looked into it for myself for OS X and couldn't find much reason not to enable it, there seemed to be very few things that needed it, and on other *NIXs you can replace atime with relatime.

I just found this on Wikipedia: http://en.wikipedia.org/wiki/Stat_(Unix)#Criticism_of_atime

quote:

However, turning off atime updating breaks POSIX compliance, and some applications, notably the mutt mail reader (in some configurations), and some file usage watching utilities, notably tmpwatch. In the worst case, not updating atime can cause some backup programs to fail to backup a file.
When I was looking into it I kept seeing mutt and tmpwatch mentioned, as for anything else, :iiam:

japtor
Oct 28, 2005
I know you want RAID for the performance but if you're willing to go with one drive you can get the 512GB Samsung 830 for $700...well whenever they get in stock.

japtor
Oct 28, 2005
Exhausting the writes that quickly would be pretty impressive, seems like you'd have to be writing constantly to get to that point that fast (some people testing that out here). Is it still fully readable?

I know OWC recommended against the hack for their SF drives cause some issues they found, but I don't know if they ever specified what exactly.

japtor
Oct 28, 2005

HalloKitty posted:

Welcome to the club! Still, at least when I eventually got through the RMA process, after providing an extremely detailed explanation of the issue, which meant they didn't ask any further questions, they sent me a brand new Vertex 2E straight from Taiwan, not a refurb. Although my original was 34nm and the new is probably 25, eh, it's going in my laptop, so it's not going to be such a big deal if it falls over.

Some people say the problems started with the Vertex 3, but they clearly never had a Vertex 2.
No they've had the reputation with the older SF drives too...they just managed to get worse with the newer drives/controllers.

japtor
Oct 28, 2005
Re: retention, I posted this link before where people are just writing continuously to see how much the drives can take before dying/locking themselves up:
http://www.xtremesystems.org/forums/showthread.php?271063-SSD-Write-Endurance-25nm-Vs-34nm

Last time I checked it was just a Samsung that died but now a Corsair is dead too apparently, going by the chart the Corsair stopped retaining after nine days, while the Samsung couldn't be recovered after being stored away for a week iirc. Of course that's an incredibly small sample size but I don't know what else we have besides datasheets until a large firm comes out with their own tests.

japtor
Oct 28, 2005

Goose Halo posted:

Do you guys recommend the 120GB because you install a lot of games? I'm trying to get the best bang for the buck for my needs.
I have 40GB on my work machine but for my normal one I had 80 and now 256. In my case I have a Windows VM sucking up however many gigs, along with my user Library being pretty big (having that on a platter drive has some negative performance implications), and I'm sure just a bunch of random crap building up over time. I felt a little cramped with the 80, like I'm pretty sure I deleted my VM to reclaim the space for something else, and I was getting a bit annoyed with moving stuff around so I just went big and got the 256 for my current machine.

I'm using around 100 gigs now and being pretty loose with the usage (it's nice not having to give a crap), so I probably would've been fine with 120. If you're going to watch the usage a lot 60 will work out, but I'd suggest going a bit bigger for wiggle room cause managing the space can get annoying. It's still a bit annoying with other drives, but in that case at least you can move stuff around if necessary.

japtor
Oct 28, 2005

Bob Morales posted:

Are the OEM Samsung drives in Dell and Apple laptops the same as the 470? Does anyone re-brand the Toshiba OEM drives?
Yes on the 470 afaik, albeit with different hardware IDs or whatever I guess (like how the Apple ones have TRIM enabled while the retail 470 doesn't). I'm not sure but it seems like Kingston is the only one to do the Toshiba ones these days. They might be around as JMicron branded controllers otherwise if they're still using their parts.

japtor
Oct 28, 2005

Civil posted:

I just picked up my first SSD, an 80GB intel 320-series. It's going to go into my laptop, I'll just do a fresh win7 install on it. Prior to that, I'd like to make sure that it has the latest firmware.

I have a SATA to USB 3 dock on my desktop PC. Should that work fine with intel's SSD toolbox software? Or should I just stop being lazy and plug it directly in via SATA?
I think you need to use SATA. Well at least you don't have a Mac mini, I should do the eSATA mod on one of mine these days just for drive maintenance crap.

Bob Morales posted:

Are those the same as the M3 or M4? The 510 uses a Marvell chipset, I think they were trying to show one drive of each controller.
Pretty sure Intel has their own firmware, no clue about the actual chips though.

japtor
Oct 28, 2005
Anyone have trouble with the Intel 320 after the firmware update? Just curious cause I came across this thread today:
http://communities.intel.com/thread/24339?tstart=0

Star War Sex Parrot posted:

For my money, the Samsung 830 is still the drive to buy. I'm really tempted to grab one, but I'm curious to see if Apple ends up with one for their iMacs this spring.
The only thing that'd stop me is if you can't upgrade the firmware from a Mac. I haven't looked at the newer ones but the older updater for the 470 seemed to just be a Windows app that downloaded the firmware and made a bootable USB drive which ran the update. So I guess as long as you could get that USB image it could work.

I also kept a copy of the PDF cause it looked like this (and used Hobo Sans for section headers inside!) :barf:

neckbeard posted:

I got a 7.9 with my Crucial M4 on a fresh install :)
My pretty old used G1 X25M got 7.3 :v:

japtor
Oct 28, 2005

dietcokefiend posted:

I got bored one day and really just wanted to see what happened. Lighting on fire hasn't happened yet.... yet
Join in on that forum doing endurance tests :v:. Are you watching the write speeds too? I think the SF drives people were testing just slowed down/throttled if they were writing enough so they wouldn't kill themselves...well the SF drives that lasted that long in the first place.

japtor
Oct 28, 2005

Podothehobbit posted:

Yeah unfortunately people are continuing to have issues after the update. Very strange.

http://communities.intel.com/message/149281
Maybe I should've gotten the Samsung instead :ohdear:. I remember seeing bad reviews for the 300 and 600GB drives on Newegg, but there were so few to begin with I didn't know how to take them.

I'm not too concerned about this since I have a UPS and power stuff in order along with daily backups...but ultimately poo poo happens and I'm afraid I might have to force power off after a crash or something.

japtor
Oct 28, 2005

Shadowhand00 posted:

Have a question - I've recently started using a Samsung 830 128GB SSD - the numbers I'm getting in terms of benchmarks range to the 248MB/s for both Read and Write. From what I read though, it seems like my numbers should be higher. I looked through the SMART info and it currently looks like its running in SATA3 Mode.

Should I be expecting higher numbers? I guess I wouldn't notice a huge difference in speed, but I like having all my stuff running well or correctly configured.

To give a little more information, I have 1 RAID 0 Array set up, along with the SSD. BIOS is set to RAID mode because of this but from what I've read, any drive not in RAID mode runs in AHCI.
That speed seems like it's being capped at SATA 3.0Gbps (SATA II), which may be what you saw? (Rather than SATA III which is 6.0Gbps and would give numbers around 500MB/s with that drive iirc)

japtor
Oct 28, 2005

Binary Badger posted:

Dont forget to check to see if the firmware is up to date; I got my Samsung 830 128 GB a week ago and it was behind one revision. Was having kernel panics watching streamed media, once I ran the update from CXM01B1Q to CXM03B1Q everything was rock solid.

Edit: running OS X Lion 10.7.3 on a 2009 MacBook. I backed up before I did the update.
What's the updater process on Macs with them? I remember some older updates with the 470 said they were Windows only, but it seemed like the Windows specific part was just an app that wrote out a bootable updater disk to a USB drive. I also remember it mentioning it'd wipe data, but I'm guessing they don't do that anymore (unless you left out mentioning restoring after the backup).

japtor
Oct 28, 2005

wanderlost posted:

Just pulled the trigger on 2 160GB Intel X25-M G2 drives. Will they have up-to-date firmware, or will I need to update them? How difficult is that going to be on a mac?
They're pretty old now so they should be up to date. If they're like the 320 I got the firmware version is printed on the label. Not sure about updating, I think you burn a bootable CD or write out a USB stick.

Why'd you get the G2 over the 320, just a good deal?

japtor
Oct 28, 2005

Intel posted:

If you are an Apple* or Linux* user, you will continue to use this tool for firmware updates

("this tool" referring to the boot CD, the Intel Toolbox software is still Windows only)

japtor
Oct 28, 2005

The foster posted:

I bought a Corsair Performance 3 128GB in september last year, and did a fresh install of Win 7 on it. Even though I didn't run any benchmarks on it I was relatively happy with the speed increase.
A week ago a friend asked me to test my SSD because he wasn't happy with his own. This is what I got:



What could be the reason for this abysmal performance? I'm guessing the percieved speed increase was because of the clean Win7 install.
My motherboard is an Asus P5KPL, wich apparently doesn't support AHCI because of some licensing issue.
The option doesn't exist in BIOS. But still, it supports SATA II so I should be getting much better results than this, right?
The perceived speed increase was probably cause the access time, if you were coming off a platter drive at least (which would be around 10ms). No clue what's up with those low speeds though.

Alereon posted:

What led you to believe the X25-M G2 would be more reliable than the SSD 320?
All I can think is write cycles for 34 vs 25nm, or the 8MB bug apparently still happening.

japtor
Oct 28, 2005

...! posted:

Would I be better off skipping those and just grabbing a regular SSD?
Yes.

quote:

if the cache drives are almost as good.
They're not.

quote:

Would there be an appreciable speed difference between running two SSDs in RAID 0 or just running one SSD by itself?
No unless you're benchmarking or just doing heavy I/O regularly in your workflow.

quote:

I learned the hard way a long time ago to never run mechanical HDDs in RAID 0, but given that SSDs are supposedly more reliable I wouldn't be as susceptible to one of the drives dying, right?
poo poo happens. Look at all the SandForce bullshit before, OCZ (...in general), Intel 8MB bug, funky Corsair one, etc.

japtor
Oct 28, 2005
Maybe wait and see how the Intel 330 is and how much it'll go for there, only up to 180GB but the announced prices for the US look really good.

Fake edit: £170 from Amazon unless the price changes:
http://www.amazon.co.uk/gp/offer-listing/B007P3RML0/ref=dp_olp_new?ie=UTF8&condition=new

japtor
Oct 28, 2005
You can get 180GB for $234 though! Well when the 330s start shipping at least.

japtor
Oct 28, 2005

Strike Anywhere posted:

I bought and deployed around ten 60GB Intel 520s into the wild at my work so far running Windows 7 x86. People love the speed.

For those of you with 60GBs in the enterprise, have you used any particular tricks to keep the 60GB drives from filling up? Besides urging users to keep docs on the network shares, I'm thinking about putting together some automatic scripts or settings pushed from group policy to do disk cleanup or whatever.

Do you have any other suggestions? I can't come up with any reasonable arguments to spend more money on 120s just so we can be lazier about keeping the drives neat and tidy.
Depending how much RAM the machines are running you could save a bit by disabling (or moving?) the hibernation file and limiting swap size if that won't gently caress anything up.

Otherwise maybe basic stuff like putting in as many links to servers in obvious places so they kind of save to there by default or something. Kind of did something similar with my mom's machine when I turned it into a VM, just moved everything out and put a shortcut on the desktop to the documents folder, took everything out of My Documents and put a shortcut there too...I'm sure there's much better ways to do this but it was sufficient in this case.

japtor
Oct 28, 2005
The old Samsungs were just slow and not worth the price, but they were reliable. I think OCZ has had the worst failure rate since they started iirc, the second gen managed to made it worse, I guess cause you had OCZ being OCZ compounded with the second gen SF issues.

japtor
Oct 28, 2005

negativeneil posted:

as someone that is about to purchase his very first SSD in the form of a 120gb Intel 330, is there any indication from intel's track record that my drive will go to poo poo like so many of you have experienced?
There's not much track record with Intel + SF other than the 530. In general Intel has been one of the more solid ones though, barring some bugs with early firmware updates for the G1 and G2, and that 8MB bug that still seems to be hitting some people even after the fix...but overall statistically they're still good afaik. Samsung's the only other one with a similar track record I think (other than slower performance with their old drives).

japtor
Oct 28, 2005

unpronounceable posted:

In terms of reliability, remember that Intel was the first company to offer a 5 year warranty on their SSDs, with their 320s. I don't know for certain, but I don't think any other company has matched that since.
On the other hand the warranty on the 330 is 3 years :v:

japtor
Oct 28, 2005
That reminds me of this from way back: http://www.anandtech.com/show/2899/4

quote:

By writing less to flash SandForce also believes its controllers allow SSD makers to use lower grade flash. Most MLC NAND flash on the market today is built for USB sticks or CF/SD cards. These applications have very minimal write cycle requirements. Toss some of this flash into an SSD and you’ll eventually start losing data.

Intel and other top tier SSD makers tackle this issue by using only the highest grade NAND available on the market. They take it seriously because most users don’t back up and losing your primary drive, especially when it’s supposed to be on more reliable storage, can be catastrophic.

SandForce attempts to internalize the problem in hardware, again driving up the cost/value of its controller. By simply writing less to the flash, a whole new category of cheaper MLC NAND flash can be used. In order to preserve data integrity the controller writes some redundant data to the flash. SandForce calls it similar to RAID-5, although the controller doesn’t generate parity data for every bit written. Instead there’s some element of redundancy, the extent of which SF isn’t interested in delving into at this point. The redundant data is striped across all of the flash in the SSD. SandForce believes it can correct errors at as large as the block level.
OCZ just decided to put SandForce's claims to the test :colbert:

...by shipping to customers instead of testing on their own.

Adbot
ADBOT LOVES YOU

japtor
Oct 28, 2005

Thermopyle posted:

How do we know they didn't test on their own? I mean, I know OCZ is shady as gently caress, but I'm just wondering if I missed something, or if you're just making a joke, or if I didn't miss anything, and you're making assumptions...
Yeah it was a joke...but that you couldn't tell whether it was a joke or not says enough about their reputation.

  • Locked thread