For months, we’ve heard reports of folks experiencing blue-screen-of-death errors when running solid-state drives based on SandForce’s latest SF-2281 controller. This so-called BSOD bug appears to affect all drives based on the controller, although SandForce claims that only “isolated” hardware configurations pose problems. We discussed the issue in the 120-128GB SSD round-up we published back in September, and at the time, SandForce was optimistic about a new firmware revision undergoing testing in its labs.
About a month later, that revision materialized in an OCZ 2.15 firmware update released to the general public. Tailored for OCZ’s Vertex 3, Agility 3, and Solid 3 SSDs, the 2.15 firmware pledges to address BSOD errors while also eliminating one cause of drive stuttering. Here’s the skinny direct from the release notes (PDF):
- Fixed a rare condition that may cause Windows Blue Screen error when the primary configured drive woke up from either a SATA slumber mode or S3/S4 mode
- Fixed a rare condition that may cause Windows Blue Screen error when the drive was configured as primary with OS installed
- Fixed a corner-case issue that may cause the drive to stutter or Windows freezing screen when a media read error occurred
OCZ got first dibs on SandForce’s new controller, so it’s no surprise the firm was the first to release a firmware update with the supposed BSOD fixes. SandForce’s other drive partners aren’t far behind, though. Corsair and Kingston have both released new firmware updates promising to address BSOD problems associated with their SF-2281-based SSDs.
Curious to try this BSOD-proof firmware for ourselves, we downloaded the 2.15 updates for our Agility 3 and Vertex 3 SSDs. These two models represent the most popular SandForce configurations on the market: the Agility pairs the SF-2281 controller with asynchronous memory, while the Vertex combines the chip with pricier (and faster) synchronous NAND. We haven’t observed substantial performance differences between similar SandForce configs from different drive makers, so we’re confident our results with OCZ drives based on 2.15 firmware will mirror what’s available from the competition.
Rather than rehashing our test methods here, we’ll point you to the appropriate page of our 120-128GB SSD round-up. The same systems and methods were used to test the 2.15 firmware. For reference, we’ve included our original Agility 3 and Vertex 3 scores, which were obtained with older 2.11 firmware. Let’s start with an overall score, which nicely summarizes the results of the most important tests in our benchmark suite.
So far, so good. The 2.15 firmware slows the OCZ drives a little, but the differences are small at best. On the Vertex 3, our overall performance score drops by less than 2% (that’s percent, not percentage points). The results for the Agility 3 are even closer.
Many of our tests showed no difference in performance between the old 2.11 firmware and the latest 2.15 release. Since you probably don’t want to scroll through several pages of graphs with little to tell, we’ll just pull out a few highlights, starting with sustained transfer rates in HD Tune.
The 2.15 firmware doesn’t really change the performance of either drive with reads. However, average write speeds are 19 and 16MB/s slower on the Vertex 3 and Agility 3, respectively. Those deltas don’t translate to slower performance for the 2.15 firmware in our real-world file copy tests. There’s no difference in random access times, either.
We did, however, notice some changes in the performance of the Agility 3 and Vertex 3 in our load-time tests.
Although the Agility and Vertex SSDs boot Windows about a second faster with the 2.15 firmware, they’re slower to load game levels. The new firmware adds fractions of a second to the load times of both OCZ drives in Duke Nukem Forever and Portal 2. Odds are you probably won’t notice the difference.
You might squeeze out a few more minutes of battery life in a notebook running either SSD, though. The Agility 3 and Vertex 3 consume a little more power under load with the 2.15 firmware, but their idle power consumption is much lower.
About those BSODs
We’ve yet to encounter a BSOD error or stuttering with any of our SandForce SSDs, so it’s difficult to determine whether the 2.15 firmware has been successful in resolving those issues. To get a better sense of how the firmware is affecting end users, we spent hours combing through hundreds of posts on the subject in OCZ’s own forums.
The good news is that most folks seem to be happy with the firmware. There are roughly twice as many users who report no issues with the new release as there are those complaining of problems. Most of those outstanding issues seem to be related to momentary freezing, often during the Windows 7 boot process. The number of users reporting persistent BSOD errors with the 2.15 firmware is relatively small in comparison.
Interestingly, Forum Support Manager RyderOCZ suggests there could be unresolved bugs creating BSODs. In this post, he states that the 2.15 firmware only addresses “a BSOD case that SandForce was able to replicate in their lab.” OCZ forum staffer Tony adds that “platform issues” may continue contributing to the problems some users are seeing.
We’ve heard from more than one source that some of the issues related to the SF-2281 can traced back to certain SATA controllers, including those from Intel. OCZ recommends updating motherboards to the latest BIOS to ensure that you’ve got the most recent Option ROM for your system’s storage controller. You’ll want to be running the latest drivers, of course, and there seems to be some confusion in the forums over whether it’s best to be using the standard AHCI drivers built into Windows 7 or Intel’s Rapid Storage Technology drivers (provided that you have a system with an Intel chipset, of course). Some folks report success with the AHCI drivers, but OCZ recommends the latest RSTs. Interestingly, a couple of users say their problems were only solved when upgrading to alpha RST drivers intended for Intel’s upcoming X79 chipset.
For its part, OCZ seems to be confident that the 2.15 firmware addresses “the main BSOD issue.” Director of Global Marketing Jessica Luken told us that the company’s tech- and forum-support teams both agree with that assessment.
There are still users complaining of issues, of course. Some of their problems may in fact be caused by incompatibilities between the SandForce controller and specific system configurations. However, after poring over countless forum posts, I’m inclined to believe at least some problems are being incorrectly attributed to SandForce SSDs. OCZ at least seems to be doing a reasonably good job of following up with users who continue to have issues.
Where do we go from here?
Given their impressive overall performance, it’s hard not to recommend SandForce-based SSDs. The asynchronous and synchronous configs that make up the Agility 3 and Vertex 3 have few peers, especially when one considers current prices. An apparent price war has broken out between SandForce drive partners, and we’ve been treated to some spectacular deals as a result. Corsair’s asynchronous Force 3 120GB, for example, was last listed at only $140 before selling out at Newegg.
SSD prices seem to be changing on an almost daily basis, so we’ve combed Newegg for the latest prices on the drives we’ve tested. Here’s how the landscape looks with one of our famous scatter plots, which tracks overall performance on one axis and the cost per gigabyte on the other. For simplicity’s sake, we’ve only plotted the performance of the OCZ drives with the 2.15 firmware.
We’re essentially looking at two tiers. The fastest drives all rely on synchronous SandForce setups, and as of this moment (noonish on November 2), the Vertex 3 is the cheapest of the bunch at just $190.
Tier two has a little more flavor, with the asynchronous SandForce SSDs joined by Crucial’s m4 and Intel’s 510 Series at roughly the same overall performance level. The 510 Series is far too expensive to be worthy of consideration, and the m4 is pricier than the Vertex 3, let alone its cheaper Agility cousin. With a $175 asking price, the Agility 3 looks like the cheapest of the asynchronous SandForce offerings you can actually buy. The Force 3 is still listed at $140 at Newegg, but it’s out of stock there and more expensive elsewhere.
Obviously, SandForce SSDs continue to offer compelling overall value versus their competition. Given the apparent success of OCZ’s 2.15 firmware, coupled with the fact that plenty of users haven’t experienced any issues with the latest SandForce SSDs, we’re confident in recommending them for desktop systems. However, I wouldn’t be as keen on popping one into a notebook unless you can grab a recent BIOS update for it. While the makers of enthusiast-oriented motherboards can probably be expected to keep their BIOS code updated with the latest storage-controller firmware, I don’t have the same faith in notebook vendors—especially for older systems. The stability of SandForce SSDs seems to at least in part depend on up-to-date SATA controllers.