The OCZ Vertex 3 Review (120GB)
by Anand Lal Shimpi on April 6, 2011 6:32 PM ESTThe NAND Matrix
It's not common for SSD manufacturers to give you a full list of all of the different NAND configurations they ship. Regardless how much we appreciate transparency, it's rarely offered in this industry. Manufacturers love to package all information into nice marketable nuggets and the truth doesn't always have the right PR tone to it. Despite what I just said, below is a table of every NAND device OCZ ships in its Vertex 2 and Vertex 3 products:
| OCZ Vertex 2 & Vertex 3 NAND Usage | ||||
| Process Node | Capacities | |||
| Intel L63B | 34nm | Up to 240GB | ||
| Micron L63B | 34nm | Up to 480GB | ||
| Spectek L63B | 34nm | 240GB to 360GB | ||
| Hynix | 32nm | Up to 120GB | ||
| Micron L73A | 25nm | Up to 120GB | ||
| Micron L74A | 25nm | 160GB to 480GB | ||
| Intel L74A | 25nm | 160GB to 480GB | ||
The data came from OCZ and I didn't have to sneak around to get it, it was given to me by Alex Mei, Executive Vice President of OCZ.
You've seen the end result, now let me explain how we got here.
OCZ accidentally sent me a 120GB Vertex 2 built with 32nm Hynix NAND. I say it was an accident because the drive was supposed to be one of the new 25nm Vertex 2s, but there was a screwup in ordering and I ended up with this one. Here's a shot of its internals:
You'll see that there are a ton of NAND devices on the board. Thirty two to be exact. That's four per channel. Do the math and you'll see we've got 32 x 4GB 32nm MLC NAND die on the PCB. This drive has the same number of NAND die per package as the new 25nm 120GB Vertex 2 so in theory performance should be the same. It isn't however:
| Vertex 2 NAND Performance Comparison | ||||
| AT Storage Bench Heavy 2011 | AT Storage Bench Light 2011 | |||
| 34nm IMFT | 120.1 MB/s | 155.9 MB/s | ||
| 25nm IMFT | 110.9 MB/s | 145.8 MB/s | ||
| 32nm Hynix | 92.1 MB/s | 125.6 MB/s | ||
Performance is measurably worse. You'll notice that I also threw in some 34nm IMFT numbers to show just how far performance has fallen since the old launch NAND.
Why not just keep using 34nm IMFT NAND? Ultimately that product won't be available. It's like asking for 90nm CPUs today, the whole point to Moore's Law is to transition to smaller manufacturing processes as quickly as possible.
Why is the Hynix 32nm NAND so much slower? That part is a little less clear to me. For starters we're only dealing with one die per package, we've established can have a negative performance impact. On top of that, SandForce's firmware may only be optimized for a couple of NAND devices. OCZ admitted that around 90% of all Vertex 2 shipments use Intel or Micron NAND and as a result SandForce's firmware optimization focus is likely targeted at those NAND types first and foremost. There are differences in NAND interfaces as well as signaling speeds which could contribute to performance differences unless a controller takes these things into account.
25nm Micron NAND
The 25nm NAND is slower than the 34nm offerings for a number of reasons. For starters page size increased from 4KB to 8KB with the transition to 25nm. Intel used this transition as a way to extract more performance out of the SSD 320, however that may have actually impeded SF-1200 performance as the firmware architecture wasn't designed around 8KB page sizes. I suspect SandForce just focused on compatibility here and not performance.
Secondly, 25nm NAND is physically slower than 34nm NAND:
| NAND Performance Comparison | ||||
| Intel 34nm NAND | Intel 25nm NAND | |||
| Read | 50 µs | 50 µs | ||
| Program | 900 µs | 1200 µs | ||
| Block Erase | 2 µs | 3 µs | ||
Program and erase latency are both higher, although admittedly you're working with much larger page sizes (it's unclear whether Intel's 1200 µs figure is for a full page program or a partial program).
The bad news is that eventually all of the 34nm IMFT drives will dry up. The worse news is that the 25nm IMFT drives, even with the same number of NAND devices on board, are lower in performance. And the worst news is that the drives that use 32nm Hynix NAND are the slowest of them all.
I have to mention here that this issue isn't exclusive to OCZ. All other SF drive manufacturers are faced with the same potential problem as they too must shop around for NAND and can't guarantee that they will always ship the same NAND in every single drive.
The Problem With Ratings
You'll notice that although the three NAND types I've tested perform differently in our Heavy 2011 workload, a quick run through Iometer reveals that they perform identically:
| Vertex 2 NAND Performance Comparison | ||||
| AT Storage Bench Heavy 2011 | Iometer 128KB Sequential Write | |||
| 34nm IMFT | 120.1 MB/s | 214.8 MB/s | ||
| 25nm IMFT | 110.9 MB/s | 221.8 MB/s | ||
| 32nm Hynix | 92.1 MB/s | 221.3 MB/s | ||
SandForce's architecture works by reducing the amount of data that actually has to be written to the NAND. When writing highly compressible data, not all NAND devices are active and we're not bound by the performance of the NAND itself since most of it is actually idle. SandForce is able to hide even significant performance differences between NAND implementations. This is likely why SandForce is more focused on NAND compatibility than performance across devices from all vendors.
Let's see what happens if we write incompressible data to these three drives however:
| Vertex 2 NAND Performance Comparison | ||||
| Iometer 128KB Sequential Write (Incompressible Data) | Iometer 128KB Sequential Write | |||
| 34nm IMFT | 136.6 MB/s | 214.8 MB/s | ||
| 25nm IMFT | 118.5 MB/s | 221.8 MB/s | ||
| 32nm Hynix | 95.8 MB/s | 221.3 MB/s | ||
It's only when you force SandForce's controller to write as much data in parallel as possible that you see the performance differences between NAND vendors. As a result, the label on the back of your Vertex 2 box isn't lying - whether you have 34nm IMFT, 25nm IMFT or 32nm Hynix the drive will actually hit the same peak performance numbers. The problem is that the metrics depicted on the spec sheets aren't adequate to be considered fully honest.
A quick survey of all SF-1200 based drives shows the same problem. Everyone rates according to maximum performance specifications and no one provides any hint of what you're actually getting inside the drive.
| SF-1200 Drive Rating Comparison | ||||
| 120GB Drive | Rated Sequential Read Speed | Rated Sequential Write Speed | ||
| Corsair Force F120 | 285 MB/s | 275 MB/s | ||
| G.Skill Phoenix Pro | 285 MB/s | 275 MB/s | ||
| OCZ Vertex 2 | Up to 280 MB/s | Up to 270 MB/s | ||
I should stop right here and mention that specs are rarely all that honest on the back of any box. Whether we're talking about battery life or SSD performance, if specs told the complete truth then I'd probably be out of a job. If one manufacturer is totally honest, its competitors will just capitalize on the aforementioned honesty by advertising better looking specs. And thus all companies are forced to bend the truth because if they don't, someone else will.
Facebook
Twitter
RSS
153 Comments
View All Comments
GrizzledYoungMan - Thursday, April 7, 2011 - link
Thank you Anand for your vigilance and consumer advocacy. OCZ's disorganization remains a problem for their customers (and I'm one of them, running OCZ SSDs in all my systems).Still, I am disappointed by the fact that your benchmarks continue to exaggerate the different between SSDs, instead of realistically portraying the difference between SSDs that a user might notice in daily operation. Follow my thinking:
1. The main goal of buy an SSD, or upgrading an SSD from another SSD, is to improve system responsiveness as it appears to the user.
2. No user particularly cares about the raw performance of their drives as much as how much performance is really available in real-world use.
3. Thus, tests should focus on timing and comparing common operations, in both solo tasking and multi tasking scenarios (like booting, application loading, large catalog/edit files/database loading and manipulation for heavy duty desktop content creation applications and so on).
4. In particular, Sandforce is a huge concern when comparing benchmarks to real world use. Sure, they kill in the benchmarks everyone uses, but many of the most resource intensive (and especially disk intensive) desktop tasks are content creation related (photo and video, primarily) which use incompressible files. How is it that no one has investigated the performance of Sandforce in these situations?
Users here have complained that if we did only #3, only a small difference between SSDs would be apparent. But to my eyes, THAT IS EXACTLY WHAT WE NEED TO KNOW. If the performance delta between generations of SSDs is not really significant, and the price isn't moving, then this is a problem for the industry and consumers alike.
However, creating the perception with unrealistically heavy trace programs that SSDs have significant performance differences (or that different flash types and processes have significant performance differences) when you haven't yet demonstrated that there are real world performance differences in terms of system responsiveness (if anything, you've admitted the opposite on a few occasions) strikes me as a well intentioned but ultimately irresponsible testing method.
I'm sure it's exciting to stick it to OCZ. But really, they are one manufacturer among many, and not the core issue. The core issue is this charade we're all participating in, in which we pretend to understand how SSDs really improve the user experience when we have barely scratched the surface of this issue (or are even heading in the wrong direction).
GrizzledYoungMan - Thursday, April 7, 2011 - link
Wow, typos galore there. Too early, too much going on, too little coffee. Sorry.kmmatney - Thursday, April 7, 2011 - link
The Anand Storage Bench 2010 "Typical workload" is about as close as you can get (IMO) to a real work test. Maybe its a heavier multitasking scenario that most of us would use, but I think its the best test out there to give a real-world assessment of SSDs. Just read the description of the test - I think it already has what you are asking for:"The first in our benchmark suite is a light/typical usage case. The Windows 7 system is loaded with Firefox, Office 2007 and Adobe Reader among other applications. With Firefox we browse web pages like Facebook, AnandTech, Digg and other sites. Outlook is also running and we use it to check emails, create and send a message with a PDF attachment. Adobe Reader is used to view some PDFs. Excel 2007 is used to create a spreadsheet, graphs and save the document. The same goes for Word 2007. We open and step through a presentation in PowerPoint 2007 received as an email attachment before saving it to the desktop. Finally we watch a bit of a Firefly episode in Windows Media Player 11."
GrizzledYoungMan - Friday, April 8, 2011 - link
Actually, the storage bench is the opposite of what I'm asking for. I've written about this a couple of times, but my complaint is basically that benchmarks exaggerate the difference between SSDs, that in real world use, it might be impossible to tell one apart from another.The Anand Storage Benches might be the worst offenders in this regard, since they dutifully exaggerate the difference between SSD generations while giving the appearance of a highly precise way to test "real world" workloads.
In particular, the Sandforce architecture is an area of concern. Sure, it blows away everyone in the benchmarks, but the fact that it becomes HDD-slow when given an incompressible workload really has to be explored further. After all, the most disk-intensive desktop workloads all involve manipulating highly compressed (ie, not compressible further) image files, video files and to a lesser degree audio files. One more than one occasion, I've seen people use Sandforce drives as scratch disks for this sort of thing (given their high sequential writes, it would seem ideal) and been deeply disappointed by the resulting performance.
No response yet from Anand on this. But I'll keep posting. It's nothing personal - if anything, I'm posting here out of respect for Anand's leadership in testing.
KenPC - Thursday, April 7, 2011 - link
Nice write up. And - excellent results getting OCZ to grow up a little bit more.As a consumer, the solution of SKU's based on NAND will be confusing and complicated. How the heck am I supposed to know if the xxx.34 or the xxx.25 or some future xxxx.Hyn34 or xxxx.IMFT25 is the one that will meet one of the many performance levels offered?
A complicating factor that you mentioned in the article, is that for a specific manufacturer and process size, there can be varying levels of NAND performance.
I strongly urge you to consider working with OCZ to 'bin' the drives with establshed benchmarks that focus on BOTH random and TRUE non-conmpressible data rates. SKU suffixes then describe the binned performance.
You also have the opportunity to help set SSD 'industry standard benchmarks' here!
Then give OCZ the license to meet those binned performance levels with the best/lowest cost methods they can establish.
But until OCZ comes up with some 'assured performance level', OCZ is just off of my SSD map.
KenPC
KenPC - Thursday, April 7, 2011 - link
Yes, a reply to my own post......But how about a unique and novel idea?
What if.. a Vertex 2 is a Vertex 2 is a Vertex 2, as measured by ALL of the '4 pillars' of SSD performance?
Vertex 3's are Vertex 3's, and so on......
If different nand/fw/controller results in any of the parameters 'out of spec', then that version never ships as a 'Vertex 2'.
After all, varying levels of performance is why there is a vertex, a vertex 2, and an onyx and an agility, and an onyx2, and an agility2, and etc etc within the OCZ SSD line.
Why should the consumer need to have to look a second tier of detail to know the product performance?
KenPC
strikeback03 - Friday, April 8, 2011 - link
So any time Sandforce/OCZ upgrades the firmware you need a new product name? If something happened in the IMFT process and they had to buy up Samsung NAND instead, new product? And of course everyone wants to wait for reviews of the new drives before buying.I personally don't mind them changing stuff as necessary so long as they maintain some minimum performance that they advertise. The real-world benchmarks in the Storage Review articles showed a 2-5% difference, to me that is within margin of error and not a problem for anyone not benchmarking for fun. The Hynix NAND performing at only ~70% of the old ones are a problem, not so much the 25nm ones.
semo - Thursday, April 7, 2011 - link
You've done well. I hope you continue to do this kind of work as it benefits the general public and in this particular case, keeps the bad PR away from a very promising technology.The OCZ Core and other jmicron drives did plenty to slow down the progress of SSD adoption in to the mainstream. You caught the problem earlier than anyone else and fixed it. This time around it took you longer because of other high priority projects. I think your detective and lobbying work are what keeps us techies checking AT daily. In my opinion, the Vertex 2 section of this article deserves home page space and a catchy title!
Finally, let's not forget that OCZ have not yet fixed this issue. People may still have 25nm drives without knowing it or be capable of understanding the problems. OCZ must issue a recall of all mislabeled drives.
Shadowmaster625 - Thursday, April 7, 2011 - link
It is ridiculous to expect a company to release so many SKUs based on varying NAND types. It costs a company big money to release and keep track of all those SKUs. When you look at the actual real world differences between the different NAND types, it only comes down to a few percentage points of difference. It is like comparing different types of motherboard RAM. It is a waste of time and money to even bother looking at one vs another. OCZ should just tell you all to go pound sand. I suspect they will eventually, if you keep nitpicking like this. The 25nm Vertex 2 is virtually identical to the 34nm version. If you run a complete battery of real world and synthetic tests, you clearly see that they are within a few % of each other. There is no reason for OCZ to waste any more time or money trying to placate a nitpicking nerd mob.semo - Thursday, April 7, 2011 - link
The real issue was that it wasn't just a few % difference. Some V2 drives were nowhere near the rated capacity with the 25nm NAND. So if you bought 2 V2 drives and they happen to be different versions, RAID wouldn't work. There is still no way to confirm if the V2 you are trying to buy is one of the affected drives as OCZ haven't issued a recall or taken out affected drives from retail shelves. Best way to avoid unnecesary hassle is not to buy OCZ at all. Corsair did a much better job at informing the customer about the transition:http://www.corsair.com/blog/force25nm/
The performance difference was higher than a few % as well.