Garbage Collection & The TRIM Bug

Like any RAID array of SSDs, there's currently no way to pass TRIM along to member drives. Once data is written to the RAID array it remains tracked and accounted for until the addresses are overwritten. This is the performance over time degradation issue we talked about at the beginning of our SSD coverage a couple of years ago.

SSD controller manufacturers have avoided the TRIM issue by equipping drives with idle garbage collection. The principle behind idle garbage collection is simple. When the SSD controller detects a period of no activity, it can query the file system for available addresses, and then internally mark those addresses for cleaning.

The RevoDrive x2 features idle garbage collection, which for most users should be sufficient to keep the drive running at full speed. There is a problem however. Current SandForce drives have a bug that impacts TRIM (and idle garbage collection). If you fill all available NAND on a SandForce drive by writing incompressible data to all user accessible LBAs and all spare area, the drive will no longer be able to restore itself to full performance via TRIM or any idle garbage collection. This impacts the RevoDrive x2 as well as standard SandForce SSDs.

I'm still in the early stages of measuring the real world impact of this bug and I made SandForce aware of it two weeks ago. At this point I'd simply caution against operating a SandForce drive near capacity filled with primarily incompressible data (e.g. compressed videos, photos). The problem is even more sensitive on a card like the RevoDrive x2 since there's no way to secure erase the drive to fully recover from a reduced performance state.

Expect to hear more about this in the next major SSD article on AnandTech.

Final Words

The RevoDrive x2, like its predecessor, continues to be a very niche product. Both sequential and random performance are measured in the multiple hundreds of megabytes per second. Like all SandForce drives, the RevoDrive x2's performance is highly dependent on the type of data you're moving. Highly compressible data will see speeds as high as 800MB/s, while incompressible data may behave more like it would on a single SSD from Crucial. The beauty of SandForce's technology is you get great performance on traditional desktop workloads, but as always you need to be mindful of what you're storing on the drive when you decide to go with SF.

The reality for OCZ however is there’s no performance benefit to purchasing a RevoDrive x2 over four Vertex 2s and creating your own RAID array. We’re talking about a simple RAID-0 here, something any performance mainstream motherboard supports right out of the box. To make for an even tougher sell, the RevoDrive x2 will likely cost more than four Vertex 2s. The original Revo was supposed to be slightly more affordable than an array of Vertex 2s, but a quick look at Newegg shows us that you pay a $40 premium for the Revo.

 

What OCZ (and other companies) ultimately need to do is introduce a SSD controller with a native PCI Express interface (or something else other than SATA). SandForce’s recent SF-2000 announcement showed us that SATA is an interface that simply can’t keep up with SSD controller evolution. At peak read/write speed of 500MB/s, even 6Gbps SATA is barely enough. It took us years to get to 6Gbps SATA, yet in about one year SandForce will have gone from maxing out 3Gbps SATA on sequential reads to nearing the limits of 6Gbps SATA. 

On the surface, OCZ is offering us what we want - a high performance PCIe SSD. However the RevoDrive needs to offer something more than ease of installation. 

All of the major controller players are hard at work on PCIe based SSDs, but I haven’t been made aware of immediate plans to release any of them. The focus continues to be winning 2.5” SATA SSD market share at this point. 

Random & Sequential Read/Write Performance
Comments Locked

46 Comments

View All Comments

  • me&er - Thursday, November 4, 2010 - link

    Anand,

    I've Secure Erased both Revo x2 and IBIS a number of times using hdparm.
    Infact, it's my preferred method with all SSD.

    Here's the methodology:

    http://www.ocztechnologyforum.com/forum/showthread...

    Regards,
  • ezinner - Thursday, November 4, 2010 - link

    Argh! Why do motherboard and peripheral manufacturer's take so long to release port upgrades? They churn out new processors and chipsets all the time, but lag behind on interconnecting high speed devices. I am so sick of the USB 2.0, serial and parallel ports still found many motherboards. Give me USB 3 and the new Intel optical interface. First you have to mass market them, then devices will come out that supprt them. Give us faster lanes. Music vendors are even worse. Keyboards still were using floppy drives and scsi ports during the last decade and use slower, lower capacity sd-ram.
  • Kevin G - Thursday, November 4, 2010 - link

    Chipsets are proprietary but high speed interconnects like SATA, USB ect. are industry standards which anyone can make devices for. Proprietary designs are easier to build as the designers know exactly the limited number of devices they interface with.

    The other issue with standards is that until chipsets include them, they'll incur an extra cost via an external chip for motherboard manufacturers. Thus even though the standard is complete and available, there will be some resistance for acceptance.
  • anon1234 - Thursday, November 4, 2010 - link

    Does this card offer management of the RAID controller, i.e. can the RAID be turned off (JBOD mode)? There are a couple of scenarios where having 4 separate 'drives' is preferable to a single stripe set.

    Does TRIM still work if you do RAID/striping in software?
  • Spazweasel - Thursday, November 4, 2010 - link

    Don't underestimate the value of ease-of-installation. That can add up pretty quickly to even a moderate sized organization. If it only takes a couple of minutes to install or swap a drive-on-a-card, versus drive-in-a-HD-format, you can make up the premium in price quickly.
  • wojtek240 - Friday, November 5, 2010 - link

    I somewhat liked the review but the ad and the review at the same time, its just a bit too much Ananad, credibility is still #1 for some folks.
  • JarredWalton - Friday, November 5, 2010 - link

    Advertisements are from a completely separate company, and they're basically "randomly" selected. I say randomly in quotes because I'm not even sure of all the details--some company could buy all of the storage ad space for a month if they were so inclined. I've read many a magazine where an ad appears opposite a review, and in some cases those are very negative reviews.

    Any good (and smart) company keeps editorial staff separate from the advertising, and AnandTech operates that way. I can say with 100% honesty that no one has ever asked me to deal with advertising, and that never once has anyone suggested I give a better review to a product because they're "a big advertiser." In fact, I know of a couple instances where companies have threatened to pull advertising because of negative (or at least not glowing) reviews, and Anand pretty much said see you later.

    Other than that, I'm not sure what in this article would undermine credibility based on the ads anyway, as my impression is more of a, "yeah, it works, and it's like RAID-0 with four SSDs only you use a single PCIe slot. It carries a price premium as well, secure erase can be iffy, TRIM doesn't work so you have to depend on idle garbage collection, and we encountered a bug on SandForce when writing lots of uncompressable data". That's definitely not enough for me to want to jump at the Revo X2, especially considering the price.
  • adonn78 - Friday, November 5, 2010 - link

    How much faster will it boot my PC over other SSD's or RAided drives? how long will it load a level of the 5 most popular games? Five us real world tangible statistics... IO's mean nothing to a gamer.
  • krazyderek - Friday, November 5, 2010 - link

    the OCZ vertex 2 120gb x 2 seems to be lacking QD32 benchmarks to compare it at the same performance metric, also shouldn't OCZ vertex 2 120gb x 4 also been included since that's the apples to apples comparison to a revo drive x2 in terms of controller count??
    And for that matter, why isn't the original revodrive shown for reference aswell?

    i'm blown away by the revodrive x2 unfortunately that's cause there aren't any other numbers to compare it too...
  • AnnonymousCoward - Friday, November 5, 2010 - link

    I couldn't find the PCIe gen in this article or the original RevoDrive one. Another site says it's PCIe 1. With all the talk about bandwidth, this is relevant. So PCIe 1 x4 = 1GB/s, and you used 804MB of that.

    Also, the charts are sort of hard to read with all the text and common colors; here's a way to improve it: http://i53.tinypic.com/3480bw4.png

    Otherwise, good review. I think motherboards need to start including non-graphics multi-lane PCIe to accommodate high bandwidth drives.

Log in

Don't have an account? Sign up now