Yes, the new 115GB is built using 25nm NAND device technology. Like OWC, SandForce is extremely conservative in what Flash devices they approve for use in drives driven by their processor and the formula for over provisioning to ensure the reliable usage of the product over the long long haul. 25nm flash is new and related to initial write cycle ratings, the amount of over provisioning was increased. Currently only our 115GB and 480GB Pro models utilize 25nm flash.

The sustained peak sustained data performance of our 25nm based solutions is absolutely in line with the prior 34nm based options. There are some differences up and down depending on type of testing, but overall – the performance remains exceptional.

Since you raise some concern about OCZ units, they have had multiple differences which do not apply to us:

• While OCZ shipped a 120GB formatted capacity drive which was actually a 115GB formatted capacity drive, we were planning this move to 115GB capacity well in advance….which was reflected in the 115GB model being listed in our Macworld US and Macworld Australia print ads that published Feb/Early March and the ads were submitted to the magazines in late January weeks before the OCZ capacity fiasco even broke out.

• Concerning OCZ’s performance drop, the key fact to know is that the SandForce controller supports 16 channels. As we understand it, the way the OCZ 60GB/55GB 25nm drive was built, only half the channels were actually populated/operating. With half the data bus, one would have to realize performance would take a hit. Additionally, there is also a likely impact in individual units where OEM or other grade NAND was used instead of Tier 1/Grade A chips.

The bottom line – as we see it, is that OCZ’s actions put a black eye unnecessarily on 25nm NAND technology.

There is no mystery here. The industry is moving to 25nm technology. What makes all the difference in this move is how SSD manufacturers build and design to use this flash type. OWC could have been shipping 25nm based product weeks before we did…but the firmware supporting this was still in testing and we were still testing built options for this use.

So right you are, not all SSDs are created equal.

More telling on some underlying change is the $40 drop in price from the old 120GB model to this new one – not counting the current $10 discount!

Has (what is basically a 128GB drive) model been changed to 25nm NAND chips? And why has the over-provisioning been bumped up to 11%?

OCZ’s recent move to 25nm 64gB NAND chips for their Vertex 2 drives resulting in a major performance drop (as mentioned above) has left a sour taste in my and numerous others mouths as potential buyers of SSD this year.

From these newer Vertex 2 drives, it has given many people the impression that the move to the 25nm chip equates to a price drop which doesn’t justify the loss of speed.

Can I ask that you – OWC – clarify why this change to the 120GB/115GB model was made, and what the implications are in terms of performance hits or improvements? In fact, a separate blog entry with actual benchmark comparisons would be fantastic.

After all, not all SSD’s are created equal, right?

I first thought OCZ should never be used. Now I realize OWC is the one to NOT BE TRUSTED!!! OWC is digging their own grave. Bye-bye guys.

Hope you can understand, but we cannot disclose business information. We can appreciate your question, but can assure you the SandForce processor is highly reliable and in fact offers better reliability and longevity that others on the market. That’s why we chose it for our line. Plenty of reviewers have noted they have had our SSD now for over a year without any issue. With a three year warranty and a 2 million MTBF, you can buy OWC with confidence.

Intel (well trusted brand as far as SSDs) has just stated it’s 1.4% for their G2 drives (source: http://www.macworld.com/article/158868/2011/03/intelssd.html).

Reason for asking: I happen to own an OCZ SSD (mainly because they were the only affordable, Mac-friendly, efficient solutions in Poland when I bought it and getting an OWC here would have cost much more – and in case of any failures would mean MUCH more hassle to RMA the disk).

Having read this post I may never buy OCZ again and will reconsider OWC – but looking around OCZ forums it is scary how many drives went dead. Is it the low-cost memory or maybe SF firmware issues? I don’t know.

But it makes me wonder if there’s something specific about SF controllers that make SSDs based on them less reliable than others. I’d appreciate if you could put some light on it. Thanks

and yes…OWC SSDs are fine with XP.

Still I would think faster SSDs would be of some benefit. Sequential read/writes aren’t everything.

Also even though my 07 MacBook has SATA II ports, I would still expect a SATA III SSD to perform better on my machine, so I’m also hoping like Dennis that OWC releases new models with SATA III.

Also, I use Boot Camp. Would OWC SSDs work O.K. with Windows XP?

Thanks for stopping by!

thanks.

I want to point out that saying you asked “…a Micron product representative on their thoughts on the use of _off-spec flash memory_ in a Solid State Drive application” is different (at least to those of us reading it here) than asking them “…on the use of _SpecTec_ flash memory”.

I sincerely do hope you secretly have a Sata3 SSD in the pipeline and announce it soon. I may wait until the Intel drive comes back down to MSRP which I thought was $580… but frankly more than $2/GB is kinda my breaking point and I’m really looking forward to SSD’s getting down to $1/GB (which I know will be another couple years).

I’m eagerly waiting for OWC’s new SATA III SSDs. I see that you don’t release information on new products until they are ready to ship. Does that mean when new products are announced that they are ready to ship same day or some known/announced number of days?

Thanks.

OWC Larry summed this issue most eloquently:
“Just because a memory device fails to meet one specification doesn’t mean it can’t be re-graded to be useful for another. In this case, Micron parts which did not pass a given spec were then ‘treated’ with that ‘S’ Blast logo over the original module and the new SpecTek coding applied. If the part met original spec, then it would have been sold under the original p/n vs. moved through SpecTek – which typically offers such parts at a substantial discount.

About the SpecTek grading/decoder info. While it does appear that these parts are marked with SpecTek’s highest grade, there is nothing on the SpecTek site to suggest the testing criteria used for this grade are comparable to the testing done for SSD drive application grading. There are references on SpecTek’s site for them having NAND solutions aimed at a variety of applications including Flash memory cards and USB Flash Drives. This potentially would suggest that criteria applied are based on use for those kinds of applications – not the same criteria for a device being screened for a Solid State Drive application. It may be the highest grade SpecTek bins to, but there is no basis we have seen to assume how that relates to the original Micron p/n specification.”

Like any manufacturer, we have products in continuous development and in majority of cases, do not release information on such until they are ready to ship. Because if a delay occurs, all a pre-announcement does is disappoint customers.