OCZ Confirms Octane and Vertex 4 use Marvell-based Silicon

By Doug Crowthers
published

OCZ Technology has confirmed the Vertex 4 is based on a Marvell Controller, which utilizes a custom Indilinx firmware.

In our recent review of OCZ's Vertex 4, we discussed how the Vertex 4 was based off an Indilinx Everest 2 controller that definitely impresses us with its performance compared to other SandForce or Marvell based controllers. This was OCZ's second Everest controller after the purchase of Indilinx, which resulted in Octane SSD series. In information leaked to AnandTech and later confirmed by OCZ Technology, the Vertex 4 is not based on an Everest 2 controller but utilizes a Marvell controller (similar to controllers used in Crucial's m4, Intel's SSD 510, etc...) with a custom Indilinx firmware. OCZ is working on non-Marvell based solutions, but the Everest 1 (Octane) and Everest 2 (Vertex 4) are based on Marvell hardware. The firmware is entirely Indilinx's own development, but with Marvell hardware. As Intel has previously stated with their own SSDs, "the value in delivering an SSD isn't always in controller hardware but rather the firmware and validation." It looks to be OCZ's Indilinx firmware is what helped deliver the performance seen with the Vertex 4.

The Vertex 4 looks to be more of a continual joint effort between OCZ and Marvell, as seen with the Z-Drive R5 developed "Kilimanjaro" platform. As for more of an official statement, OCZ states "Marvell is a valued partner with OCZ and we have co-worked together on a number of projects to create SSD platforms. The Everest platform utilizes a custom solution from Marvell running at higher clock speeds and delivers enhanced performance and features with Indilinx proprietary firmware. OCZ will continue to co-work with Marvell on current and future SSD technology including our next generation enterprise solutions including the already announced Z-Drive R5 which features the jointly developed Kilimanjaro platform."  

Does this truly change anything about the performance of the Vertex 4? No, it just means what initially was thought to be an Everest 2 controller is truly a Marvell controller (thought to be Marvell's new 88SS9187 controller) with an Indilinx firmware. It doesn't change the fact that the drive offers great performance, great value and a 5-year warranty, which won us over during our review.

Doug Crowthers
Topics
OCZ
SSD
Storage
12 Comments Comment from the forums
  • e56imfg
    Honestly, I don't really care since it still provides great performance.

    It's just when PSU companies lie that pisses me off...
    Reply
  • atikkur
    that would mean, it is reliable enough (for those who not convinced by ocz's reliability mindset).
    Reply
  • e56imfg
    EDIT: Double Post Error
    Reply
  • divya nanjappa
    A hard drive's MTB = 20,000 at best.

    SSD = 1,000,000

    It would be nice if corporations chose to embrace (not so new technologies.)

    Because it would lower ROI, and provide a better user experience.
    Reply
  • borisof007
    I would agree Divya except my OCZ Vertex LE died 18 months after purchase, WAYYY under the millions of hours MTBF advertised. Granted, OCZ sent me a replacement, but i still had to pay for shipping which was pretty lame.

    Pay $300 for a good SSD, you expect it to work.
    Reply
  • dalauder
    I don't think this news is getting NEARLY the interest it should get. Consider:

    Vertex 4's use essentially the same controller as a Crucial M4 or Intel 510 but get better performance than an M4 or a Vertex 3 (overall, since it doesn't hit uncompressible or other bottlenecks). So we're looking at an SSD that should both achieve the performance AND reliability crown! That's actually amazing! And nobody really sounds that impressed.
    Reply
  • siuol11
    This article is so poorly worded I have no idea what is what.
    Side note: I've gone through at least 4 SSD upgrades, and each new one was supposedly 2X the performance of the old, but there is little real-world difference. I'll start getting excited again when we have software (probably Windows 10) that can actually take advantage of that speed.
    Reply
  • boiler1990
    I wonder if Sandforce even stay in business since one of their biggest clients just shifted to a competitor.

    I also hope these drives correct the reliability issues OCZ has been having - it just seemed a bit silly that a company that dropped it's RAM division to focus on SSDs couldn't make a reliable SSD.
    Reply
  • matt_b
    I'm curious how Sandforce is fairing out mentally right about now. They were the hot ticket for performance for the past couple of years, perhaps firmware and reliability both proving to be hit and miss over that time period has left behind that poor of an impression in SSD manufacturers?

    Personally, I'll take as much performance as I can get, but something as crucial (no pun intended) as data, I cannot sacrifice reliability, even if a RAID 1/5/10 setup is used. This is where although the Crucial M4 was a bit slower than the blazing fast SF controller products, I respect that Crucial (more so the Marvell controller) took the higher reliability route, even if some speed was sacrificed and posted numbers didn't really "wow" people over.
    Reply
  • blazorthon
    siuol11This article is so poorly worded I have no idea what is what. Side note: I've gone through at least 4 SSD upgrades, and each new one was supposedly 2X the performance of the old, but there is little real-world difference. I'll start getting excited again when we have software (probably Windows 10) that can actually take advantage of that speed.
    Small speed improvements are all you get if you're not using faster SSDs for heavy multitasking... That is where they truly shine, whereas the opposite is true for hard drives. Have an anti-virus scan, game, and disk error checker running at the same time or something like that and notice how the SSD doesn't slow down at all. That is where the biggest performance gains between SSD generations can be found if the generations all use the same version of SATA.

    Reply