Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Please help me choose a 256gb SSD
- Thread starter handheldgames
- Start date
- Sort by reaction score
Just to clarify the current issues with Sandforce based drives...
There appear to be two problems with Corsair and Sandforce drives in general which should not be confused.
Corsair's recall is limited to their 120GB drive and perhaps all SSD's using Sandforce's reference design for such a drive, indicating there is some very specific hardware issue with that particular model...
Link
There is also a broader Sandforce firmware issue as suggested by this OCZ announcement on their forums...
So far, I haven't seen anything related to OWC. Let's hope it's a firmware update at worst.
There appear to be two problems with Corsair and Sandforce drives in general which should not be confused.
Corsair's recall is limited to their 120GB drive and perhaps all SSD's using Sandforce's reference design for such a drive, indicating there is some very specific hardware issue with that particular model...
Link
There is also a broader Sandforce firmware issue as suggested by this OCZ announcement on their forums...
So far, I haven't seen anything related to OWC. Let's hope it's a firmware update at worst.
Just spent some time going through many reviews on drive after driver over at NewEgg... Every product I checked based on the Sandforce chipset (current or legacy), irrespective of the manufacturer is plagued with failed product reviews. Total drive failure, constant lockups, etc....
Products based on Intel, Marvel & Samsung chipsets seemed to be void of these reviews... Perhaps we looking at the next IBM DeathStar fiasco?
Products based on Intel, Marvel & Samsung chipsets seemed to be void of these reviews... Perhaps we looking at the next IBM DeathStar fiasco?
Last edited:
In contrast to the bad news regarding Sandforce driven SSDs, it is interesting that OWC is actually increasing the production rate of the Mercury Extreme Pro 6G and showing no signs of retreat, they just published it yesterday:
http://blog.macsales.com/10663-sand...ng-up-in-production-to-meet-volume-deliveries
So they seem pretty self-confident (add the fact of 5 years warranty), which were absolutely unwise should the Sandforce chipset cause substantial problems on the realibility of their drives. Well, I hope it is not so. Anyway I'm going to tell you if mine would fail.
http://blog.macsales.com/10663-sand...ng-up-in-production-to-meet-volume-deliveries
So they seem pretty self-confident (add the fact of 5 years warranty), which were absolutely unwise should the Sandforce chipset cause substantial problems on the realibility of their drives. Well, I hope it is not so. Anyway I'm going to tell you if mine would fail.
Very interesting thread here at MR regarding OWC ssd.
https://forums.macrumors.com/threads/1057025/
Hmmmm
https://forums.macrumors.com/threads/1057025/
Hmmmm
Firmware seems to be the biggest issue with the SF-2281, but I'm wondering the same thing about the 120GB reference design (not seen issues on OWC's specific to that particular capacity though - looks like they got it sorted with their firmware revision 360, and they may not be following the reference design, though I didn't get this impression from a review that's linked in this thread).
Assuming Sand Force did blow the 120GB Reference design, that would be a major issue as most other drive makers would use it (cheaper, and known to work - customization would still be possible via the firmware).
Another possibility may be the actual NAND Flash used. Even with ONFI (Open NAND Flash Interface Working Group), NAND Flash is still flaky (only sets standards to communications and pin-outs sorts of things (communications interfacing), not how the actual memory is made). This is where component testing/validation comes in, and where shortcuts have become common place in my experience.
But I'm not seeing enough information to conclude that Sand Force totally screwed up with the SF-2281 (Client SSD Processors = 2xxx series parts <as well as 1xxx series>, as there's Enterprise SSD Processors as well = 2500/2600 series parts <1500 series as well>).
.pdf datasheets if you're interested (direct links; only a couple of pages each).
Seems that's the biggest problem from what I'm getting from the articles and posts as well.
As to OCZ's statement, I find it laughable. As a general rule, in-house design and manufacturing usually creates better products (Intel, Samsung, and Toshiba for example if we stick to SSD's). But OCZ's history is the opposite.
Firmware is all I've noticed with their products, and there's indications they may have solved it with revision 360. I wouldn't go so far as to think/state it's a certainty, but they seem to be in better shape than OCZ or Corsair's SF-22xx controller based products.
But I'm under the impression that they do use the reference designs, and don't seem to have an inordinate number of issues with the 120GB versions than the others (admittedly, I wish there was more detailed information on this, as to not get the wrong impression). So I'm still not sure if the 120GB issues with other companies gear is an issue with the reference design, or something else.
I'm getting the impression that the OWC offerings are fine, but there is still a little uncertainty overall (those that have made complaints didn't always state the drive's capacity or firmware revision/s). So selecting a disk based on a different controller would make sense from a reliability POV if they're wanting to buy one now.
BTW, Toshiba based disks would be good as well (haven't seen issues on those either).
There are indications that the firmware revision 360 may have their previous issues sorted. Assuming this is the actual case (engineers can no longer see the previous issues when testing this particular revision; assumes they've been thorough), then that would boost OWC's confidence in their product quality. So it's not unreasonable that this increased confidence would cause them to increase production (they could actually see the current situation with other SF-2281 products that are still having issues as a means of picking up additional sales).
In light of the issues surrounding SandForce based products.. The OWC Mercury Extreme Pro was scratched from the list, leaving the Samsung 470 256 SSD as the last SSD Standing. I like the Intel option... It's just my wallet didn't.
BuyDig.com has a great deal on the Samsung 470 SSD 256 for $399.00. PLUS they have a 45 DAY return policy with NO restocking fee... SWEET. Which is much better than NewEgg's limited exchange only policy..
At this point, there is no way I would trust a Chipset + 3rd Party NAND solution from any "Trusted" manufacturing partner.
Will update on Tuesday after the install
BuyDig.com has a great deal on the Samsung 470 SSD 256 for $399.00. PLUS they have a 45 DAY return policy with NO restocking fee... SWEET. Which is much better than NewEgg's limited exchange only policy..
At this point, there is no way I would trust a Chipset + 3rd Party NAND solution from any "Trusted" manufacturing partner.
Will update on Tuesday after the install
The last thing I need is to wake the MacPro to a panic locked SSD and have to go thru the hassle of a time machine restore. SSD is a fresh tech with lots of unproven product and a much higher chance of data loss vs traditional hard disks. Just being cautious.