OCZ Vertex 3 Pro SandForce SF-2000 Based SSD Preview

9 thumbs up

At the heart of the upcoming Vertex 3 Pro SSD is a brand new controller from SandForce, the SF-2582. The SF-2582 is a member of the upcoming SandForce SF-2000 family of SSD processors, which was announced a few months back. The new SF-2000 SSD processor family builds upon the success of the current SF-1500 / SF-1200 series by adding 6Gb/s SATA III interface support, integrating a more powerful encryption engine, and adding support for new flash memory types. While new, the SF-2000 family does, however, leverage technologies from the previous generation.  As such, we'd recommend checking out some of our previous SandForce coverage for some back-story on SandForce and what makes their SSD controllers so sought after. In our look at the OCZ Vertex Limited Edition, we go into detail on DuraWrite and RAISE and some of SandForce's other proprietary technologies. We highly recommend checking that story out.

Back to the issue at hand. Along with virtually all of the features of the previous-generation, the SandForce SF-2000 SSD processor family will offer SAS-bridge support for non-512 byte sectors, enhanced ECC and BCH capabilities, and it features new Power / Performance throttling technology for more “green” computing environments. All of the features culminate in a new controller that’s capable of over 500MB/s sequential reads and writes, with roughly 60,000 random 4K IOPS according to SandForce. And just to give you a little taste before moving on to the next few pages, even at this early beta stage, the Vertex 3 Pro is already hitting those performance levels.


SandForce SF-2000 Block Diagram

With this new line of SSD processors, SandForce will first be introducing products into enterprise space, but new products will eventually trickle down into the consumer space as well. Due to OCZ's close working relationship with SandForce, it comes as no surprise that the first SF-2000 based SSD to land in the labs comes by way of OCZ, but expect many of SandForce's current partners (and perhaps a few more) to jump on board and offer products based on the SF-2000 family as well.

 

Additionally, we've got some more specifics regarding the new SF-2000 SSD processors direct from the source. The breakdown of new features, direct from SandForce is as follows:

  • Support for advanced 30nm- and 20nm-class Flash with Asynch/ONFi2/Toggle interfaces with data rates up to 166 Mega Transfers per second
     
  • Enhanced dual-ported SAS bridge support, including non-512-byte sector sizes, e.g., 520, 524, 528, 4K, etc., with Data Integrity Field (DIF) for true Enterprise-class SAS drive behavior and performance
     
  • TCG Enterprise security with selectable multi-banded 256/128-bit AES encryption with line-rate double encryption for data written to the drive
     
  • Advanced ECC engine correcting up to 55 bits per 512-byte sector to assure high data integrity and support for future generations of Flash memory
     
  • Power and performance throttling options to support green computing initiatives
     
  • Industrial temperature support (-40 to +85 degrees Celsius)

 

Article Index:

1 2 Next
0
+ -

I think thing's like these results is why over time the SSD is going to have the impact it will. I know it is expensive compared period across the board. The thing is SSD technology here seems to be stretching it's legs in this race. From a performance perspective looked at from a time end point such as in say a PIII top of the line to a current 6core top of the line extreme intel processor, or even an X2 versus a top of the line current Opteron 6/8/or 12 core and you will see my illustration. The PC equipped with it makes the other one not only look like it is standing still, in comparison it makes it useless from a now stand point.

So if this is where this platform segment is going in a few years it will be not only a necessity, but a general component in a PC I would think. How could you as even a general user use a PC that if now is at an almost constant 500 rating across the board use anything else ( I am talking from a future perspective here). If your main data drive doubles the performance ratio of most of these drives, think about what it looks like to a standard (this think makes SCSI Ultra320 look like 3 year old technology here) mechanical HD. Not to mention the price to size comparison is better than last year for sure, and twice as much for two years ago.

So I see this as an eventual standard, because anything that beats this compared to any other storage method, which I am sure these will do next year even if at a lesser degree than these point levels is almost a must have. Of course I woulod love to have a Vertex 2, a Vertex 3 would be a dream come true really. The operational level of anything with a drive of this caliber across the PC just has to be astounding!

0
+ -

The leading edge is usually the bleeding edge. Yeah, these are exfrigginspensive, but the performance is outstanding as well. Maybe I can afford them in four years or so.

Nice review Marco.

+1
+ -

I wont complain about the price and I'll give props to OCZ for the amazing work, effort and dedication that resulted in this accomplishment . Excellent job!!

But what will be great is that all this ongoing research and development is going to benefit us all in a few years because this type of performance will be standard in all SSD's plus the benefits of lower cost , greater affordability and larger capacities. Plus the reliability, MTBF :10 Million, that's insane!!

0
+ -

No, it won't.

SSDs are not going to follow the same cost curve as HDDs. Smaller process geometries and increased fab costs related to the R&D required to build drives at smaller nodes is going to keep the miles-wide gap between HDD and SSD intact.,

In five years, SSDs will have more of the market. They will be cheaper. They will be faster. Hard drives will still dominate in terms of sales, capacity, and cost/GB. Even if we assume SSDs are 10x cheaper (down to about 22 cents / GB), that'd still be 5x more expensive than HDDs are *now.*

0
+ -

I do not agree with Joel here. Where there is a will, there is a way, especially with semiconductors. That said, the SSD may not be made of NAND Flash forever either, who knows but I can almost guarantee you we will go almost all solid state in the future.

+1
+ -

Joel H:

No, it won't.

Yes it Will.

Joel H:
Smaller process geometries and increased fab costs related to the R&D required to build drives at smaller nodes is going to keep the miles-wide gap between HDD and SSD intact.

So what about "Profit". You just cant concentrate R&D on just making things smaller, faster while keeping the cost of manufacturing high. R&D has to make ways on making the build process cheaper, and more profitable.  Why keep moving a moving to one expensive process after another when you can work with what you got already mastered because over the long run, the cost of production will go down, sales will go up and the profit margin will rise substantially.

Maybe a good example (or not) is the Playstation 2, at first it was expensive to built and Sony were selling it at a lower cost than the  production value, but because it was affordable to the consumer, it sold millions and millions over time, thus Sony was able to make huge profits on the PS2 units.

Now, I'm not saying that a company should lose money or subsidize, its just that, they don't have to be the fastest performing, but they do have to be reliable, affordable and accessible to all and at the same time super beneficial to the company with high profits which is what you'll get over a period of time.

A better example would be Intel. Their R&D teams have found ways on making their processors smaller, more powerful, cheap and easier to produce. That's why they are selling tons and had recent record profits.   But whats is the factor that make them affordable to acquire, "Competition". If it wasn't for AMD or ARM, Nvidia or any other company present, would the I7's be priced how they are?

Joel H:
SSDs are not going to follow the same cost curve as HDDs. In five years, SSDs will have more of the market. They will be cheaper. They will be faster. Hard drives will still dominate in terms of sales, capacity, and cost/GB. Even if we assume SSDs are 10x cheaper (down to about 22 cents / GB), that'd still be 5x more expensive than HDDs are *now.*

I'm not gonna argue that, but the prices  will come down.  My final answer is that "Competition" from current and new companies will drive the SSD market to be more affordable. Right now a  premium SSD is more suited for the OS and applications, and not for Storage purposes, which is also why companies are charging so much. Its like its a luxury to own one but an excellent investment for the benefits it offers.

0
+ -

According to its specs, this SATA 6G drive is rated for 550MB/s max read speeds with 500MB/s writes and up to 70K IOPS. Yes, you read that correctly. This drive’s performance could potentially double, or even triple, that of some current-gen products being sold today in some scenarios.

Damn!! Thats fast as anything!!

A few weeks ago, there were articles about WD and Seagate not going into SSD's... and i agreed with their decision, But already, we've touched 250gb's and read/write speeds that have no comparison to the hdd's....

Now i'm starting to worry a bit about wd and seagate.... i mean, sure that hdds are going to be with us for a long time and even cheaper to produce.... but these companies have no infrastructure to produce ssds.

You can only improve a technology soo much until it becomes obsolete.

+1
+ -

The thing with SSD's is this (not to mention this is a corporate drive which are usually twice as much if not more than a consumer SSD), the limitation was 22nm for this memory. When they reach that threshold which they may very well get around and I would not be surprised, but either way there will be a threshold met close to that maybe one measurement down and producible maybe even 4 to 18nm or even 8 down to 12nm. At the speed there moving now that will most likely be hit either this year or the next. When that happens it will be software development as well as the other existing components in the board. When this happens there will not be as much play as was available the last two years, and they will start coming down at a faster pace, because then they will only have volume for profit just like mechanical Hdd's. Which is also why the mechanical Hdd is as cheap as it is now.

The other thing you have to consider is the speed which is insane here, but will get better. Lay the threshold now at 500MB/s throughput, and put that against the current top of the line drive of the same main type with the same components. The system with the SSD will operate at least twice as fast period point blank, and most likely at least a 3-400% operational window. Your PC today with mechanical hdd's will look like a year 2000 PC compared at best. No matter the speed of your processor because then the drive and memory will be operating that much faster. So with a regular mechanical HD the same PC will operate at one speed, with one of these or better SSD feeding it the operational speed will be triple across the board. Your GPU cannot render any faster than the drive can load data to the memory to be processed.

So with twice the data at a minimum everything else moves faster as well.

0
+ -

I would love one of these! But could not justify the cost considering I already have a few SSD's.

Maybe in the future someone here at HH would be lucky enough to win one in a possible future Giveaway! :)

0
+ -

There is some competition out there but I think OCZ Vertex 3 is going to definatly be the drive to have! At the current price those who are better equipped finacially are going to be enjoying the benefits of this amazingly fast drive but then the other guy LOL like me are just going to be a lttle patient to have some fun! but I no doubt will be waiting! Good review!

1 2 Next
Login or Register to Comment
Post a Comment
Username:   Password: