review oczs vector ssd reviewed

OCZ’s Vector SSD reviewed

OCZ deserves a lot of credit for helping to popularize solid-state drives among PC enthusiasts. So does Indilinx, whose Barefoot controller powered the original OCZ Vertex SSD. The two firms became one in 2011, when OCZ bought Indilinx for $32 million. That acquisition gave OCZ in-house controller expertise, a resource many of its SSD rivals lack, and we eagerly awaited the inevitable offspring.

The first products spawned by the marriage weren’t quite what we were expecting. OCZ introduced Octane, Vertex 4, and Agility 4 SSDs tagged with “Indilinx Infused” badges that referred not to the controller silicon, but to the accompanying firmware. Those drives all used off-the-shelf Marvell controller chips combined with custom firmware developed by OCZ. And so we waited some more.

The Indilinx badge has popped up again on OCZ’s new Vector SSD, but this time it has a deeper meaning. Yes, the Vector boasts custom firmware like its similarly infused predecessors. Unlike those other drives, though, the Vector has controller silicon all its own: the Indilinx Barefoot 3. This long-anticipated chip makes its debut in the Vector, and we’ve run the drive through its paces to see whether it’s been worth the wait.

Inside the Vector
Even at first glance, the Vector is clearly different from OCZ’s existing drives. There’s the slimmer 7-mm case, which has rounder corners and a bigger sticker than the firm’s previous 9.5-mm designs. Despite being thinner, the Vector is noticeably heavier, tipping the scales at 116 grams. For reference, the equivalent Vertex 4 model weighs 92 grams, while the Intel 335 Series and Samsung 840 Series are just 81 and 53 grams, respectively. Those differences are essentially meaningless for desktop use, but weight-obsessed notebook users may want to take note. For what it’s worth, the thick metal walls responsible for the Vector’s heft also make the drive feel practically bombproof.

A Phillips screwdriver is all that’s required to crack open the case and reveal the circuit board inside. On it, we find the Barefoot 3 controller surrounded by a collection of OCZ-branded memory chips. We’ll get to the flash in a moment, but first, let’s take a closer look at the controller. OCZ has supplied a helpful block diagram identifying the most interesting parts:

There are two cores inside the Barefoot 3 chip: an ARM-based Cortex processor and an Aragon co-processor of OCZ’s own creation. Earlier this year, former OCZ CEO Ryan Petersen described Aragon as a 32-bit processor with an SSD-specific RISC instruction set. This computational sidekick is tasked with efficiently managing the flash interface, OCZ says, but we couldn’t pry additional details from the company. OCZ wouldn’t get into more specifics on the ARM-based core, either, noting only that it’s based on the “Cortex” architecture. I suspect this is a Cortex-R variant, which is what ARM recommends for use in storage controllers.

A hardware randomizer appears in the Barefoot 3’s block diagram, suggesting that the Vector has full-disk encryption support. Except it doesn’t. Encryption is supported by the controller, but that functionality has been disabled in the Vector; OCZ doesn’t think consumers are all that interested in scrambling data stored on the drive. Having been forced to downgrade the encryption support of its SandForce-based drives due to a problem with the associated controller, OCZ may be in a good position to judge the importance of full-disk encryption among its customers. The company is “continuing to evaluate enabling this functionality,” suggesting it might be possible to unlock the feature via firmware update. At the very least, I’d expect OCZ to release an enterprise-oriented SSD that taps into the Barefoot 3’s encryption support.

OCZ says the Vector and its associated Barefoot 3 controller have been tuned to offer high performance and endurance without compression or lost capacity. If you’ve been following the solid-state storage business over the past couple years, you might interpret that assertion as a thinly veiled dig at SandForce—a company that supplied the controllers for two generations of multiple OCZ SSD families before the infamous BSOD bug reared its ugly head.

To accelerate performance and reduce NAND wear, SandForce controllers compress incoming data when writing it to the flash. The benefits of this approach hinge on the compressibility of the data, so the results can vary based on the file type. Since the Vector’s Barefoot 3 controller doesn’t use compression, it should be equally adept at shuffling everything from plain text files to highly compressed movies.

The bit about lost capacity likely refers to RAISE, a RAID-like flash redundancy scheme used by most SandForce-based SSDs. This feature guards against data loss due to physical flash failures, but you lose some capacity to store the necessary parity data. That’s why typical SandForce drives offer 120, 240, and 480GB of storage despite having 128, 256, and 512GB of flash memory onboard. The Vector doesn’t dedicate any of its flash capacity to RAID-like schemes, freeing up extra gigabytes for user storage.

Direction and magnitude
Like most contemporary SSD controllers, the Barefoot 3 has eight NAND channels. Each one can address up to eight individual flash dies, bringing the controller’s total capacity up to 64 dies. Those flash chips can conform to the ONFI or Toggle DDR standards, giving OCZ the flexibility to offer multiple products based on different flavors of NAND. The Vector uses synchronous ONFI 2.2 chips similar to those found in high-end SandForce SSDs.

The 25-nm MLC NAND comes from Intel, but OCZ’s name is silk-screened onto the packages. What gives? OCZ is buying entire flash wafers and then slicing and packaging the individual dies itself. This approach cuts costs, OCZ says, and the savings are purportedly passed along to the customer.

Packaging its own NAND should allow OCZ to sort individual dies based on their performance and endurance characteristics, reserving only the best chips for higher-end drives like the Vector. OCZ wouldn’t confirm whether it’s engaged in this process, which is commonly referred to as binning. We do know that OCZ isn’t the only SSD vendor buying NAND by the wafer, though. Kingston also does its own flash packaging for some drives.

In addition to the NAND, our Vector 256GB sample has a pair of 256MB DRAM chips onboard. These low-power DDR3 chips are split between the two sides of the circuit board and combine to form the drive’s cache memory. The resulting 512MB cache size is identical to that of the Vertex 4 and quite a few other drives in the 256GB range

Capacity NAND
Max sequential (MB/s) 4KB random QD32 (IOps) Price
Read Write Read Write
128GB 16 x 64Gb 550 400 90,000 95,000 $150
256GB 32 x 64Gb 550 530 100,000 95,000 $270
512GB 64 x 64Gb 550 530 100,000 95,000 $560

At least to start, OCZ will limit the Vector line to 128, 256, and 512GB flavors. All three use the same 64Gb NAND dies, and it looks like the 128GB model doesn’t have enough of them to fully exploit the controller’s performance potential. The 128GB drive’s sequential write specification trails that of its higher-capacity siblings by 130MB/s. That model’s random 4KB read throughput is also lower by 10,000 IOps. There’s no difference between the performance ratings of the 256 and 512GB variants, though.

All three models are covered by the same five-year warranty, albeit with some small print attached. The warranty is good for five years or 36.5TB of total host writes, whichever comes first. 36.5TB might not seem like a lot, but it works out to 20GB of writes a day for five years. If those numbers sound familiar, that’s because they mirror the endurance specification attached to Intel’s 520 Series SSD.

The Vector employs an assortment of tricks to extend the lifespan of its flash memory, including a proprietary garbage collection algorithm, multiple levels of error correction, and the adaptive NAND management functions fueled by the Aragon co-processor. OCZ has also taken steps to improve reliability, noting that the Vector is its “most extensively and comprehensively tested consumer SSD line to date.” A “large network of beta testers” was part of the validation process, and all Vector drives will undergo a burn-in test before they ship. OCZ says future firmware updates will be subjected to more validation testing, as well.

All this extra testing sounds good, but we won’t know how effective it is until a large number of Vector drives are out in the wild. In the meantime, there is some room for optimism. The Vertex 4’s customer reviews on Amazon and Newegg are definitely more positive than similar reviews of the older, SandForce-based Vertex 3. Customer reviews aren’t completely reliable, of course, but we haven’t heard about any pesky problems with OCZ SSDs since the SandForce BSOD bug. With everything but the Vector’s NAND and cache memory developed in-house, OCZ will have only itself to blame if the drive exhibits any flakiness.

The Vector comes with a 3.5″ bay adapter designed for older cases that lack dedicated mounting hardware for 2.5″ drives. OCZ also throws in an activation key for Acronis’ True Image HD cloning software. Adding an SSD system drive is a great way to improve the performance of an existing machine, and I’m sure upgraders will appreciate having a free tool to migrate their existing Windows installations.

As one might expect, the Vector is compatible with OCZ’s own Toolbox utility. This app covers the basics, including updating firmware and securely erasing the drive. Despite recommending that the Vector be paired with Intel’s RST storage drivers for optimal performance, OCZ suggests using Microsoft’s AHCI drivers for firmware updates. Apparently, other storage drivers don’t consistently pass firmware update commands correctly. You’d think a portion of the Vector’s additional validation testing would be dedicated to ensuring that firmware updates work properly with the recommended performance driver, though. OCZ says RST users can attempt firmware updates to see if they work, but I’m not sure I’d be comfortable rolling the dice without a fresh backup image to protect my data.

If you want to watch your drive’s slow but inevitable decline to fatal flash failure, the Toolbox app can also be used to monitor SMART values. It’s a death by millions of needles as individual NAND cells slowly burn out due to the destructive nature of the electron tunneling process used to program flash memory. The data is presented in a text window, which will probably confuse mainstream users more than the simple graphs used by the equivalent Intel and Samsung utilities. Those apps are a little more polished than the OCZ Toolbox.

Our testing methods
If you’re familiar with our test methods and hardware, the rest of this page is filled with nerdy details you already know; feel free to skip ahead to the benchmark results. For the rest of you, we’ve summarized the essential characteristics of all the drives we’ve tested in the table below. Our collection of SSDs includes representatives based on the most popular SSD configurations on the market right now.

  Interface Cache Flash controller NAND
Corsair Force Series 3 240GB 6Gbps NA SandForce SF-2281 25nm Micron async MLC
Corsair Force Series GT 240GB 6GBps NA SandForce SF-2281 25nm Intel sync MLC
Corsair Neutron 240GB 6GBps 256MB LAMD LM87800 25nm Micron sync MLC
Corsair Neutron GTX 240GB 6GBps 256MB LAMD LM87800 26nm Toshiba Toggle DDR
Crucial m4 256GB 6Gbps 256MB Marvell 88SS9174 25nm Micron sync MLC
Intel 320 Series 300GB 3Gbps 64MB Intel PC29AS21BA0 25nm Intel MLC
Intel 335 Series 240GB 6Gbps NA SandForce SF-2281 20nm Intel sync MLC
Intel 520 Series 240GB 6Gbps NA SandForce SF-2281 25nm Intel sync MLC
OCZ Agility 4 256GB 6Gbps 512MB Indilinx Everest 2 25nm Micron async MLC
OCZ Vector 256GB 6Gbps 512MB Indilinx Barefoot 3 25nm Intel sync MLC
OCZ Vertex 4 256GB 6Gbps 512MB Indilinx Everest 2 25nm Intel sync MLC
Samsung 830 Series 256GB 6Gbps 256MB Samsung MCX 27nm Samsung Toggle MLC
Samsung 840 Series 250GB 6Gbps 512MB Samsung MDX 21nm Samsung Toggle TLC
WD Caviar Black 1TB 6Gbps 64MB NA NA

Yes, we know the Samsung 840 Pro isn’t on that list. We hope to have a review of the drive soon, but our sample appears to have been delayed to receive a firmware update. We’ve seen several reports of 840 Pro drives failing prematurely under heavy loads, and we suspect this update is meant to address the apparent issue.

We used the following system configuration for testing:

Processor Intel Core i5-2500K 3.3GHz
Motherboard Asus P8P67 Deluxe
Bios revision 1850
Platform hub Intel P67 Express
Platform drivers INF update
Memory size 8GB (2 DIMMs)
Memory type Corsair Vengeance DDR3 SDRAM at 1333MHz
Memory timings 9-9-9-24-1T
Audio Realtek ALC892 with 2.62 drivers
Graphics Asus EAH6670/DIS/1GD5 1GB with Catalyst 11.7 drivers
Hard drives Corsair Force 3 Series 240GB with 1.3.2 firmware
Corsair Force Series GT 240GB with 1.3.2 firmware
Crucial m4 256GB with 010G firmware
Intel 320 Series 300GB with 4PC10362 firmware
WD Caviar Black 1TB with 05.01D05 firmware
OCZ Agility 4 256GB with 1.5.2 firmware
Samsung 830 Series 256GB with CXM03B1Q firmware
Intel 520 Series 240GB with 400i firmware
OCZ Vertex 4 256GB with 1.5 firmware
Corsair Neutron 240GB with M206 firmware
Corsair Neutron GTX 240GB with M206 firmware
Intel 335 Series 240GB with 335s firmware
Samsung 840 Series 250GB with DXT06B0Q firmware
OCZ Vector 256GB with 10200000 firmware
Power supply Corsair Professional Series Gold AX650W
OS Windows 7 Ultimate x64

Thanks to Asus for providing the systems’ motherboards and graphics cards, Intel for the CPUs, Corsair for the memory and PSUs, Thermaltake for the CPU coolers, and Western Digital for the Caviar Black 1TB system drives.

We used the following versions of our test applications:

Some further notes on our test methods:

  • To ensure consistent and repeatable results, the SSDs were secure-erased before almost every component of our test suite. Some of our tests then put the SSDs into a used state before the workload begins, which better exposes each drive’s long-term performance characteristics. In other tests, like DriveBench and FileBench, we induce a used state before testing. In all cases, the SSDs were in the same state before each test, ensuring an even playing field. The performance of mechanical hard drives is much more consistent between factory fresh and used states, so we skipped wiping the HDDs before each test—mechanical drives take forever to secure erase.

  • We run all our tests at least three times and report the median of the results. We’ve found IOMeter performance can fall off with SSDs after the first couple of runs, so we use five runs for solid-state drives and throw out the first two.

  • Steps have been taken to ensure that Sandy Bridge’s power-saving features don’t taint any of our results. All of the CPU’s low-power states have been disabled, effectively pegging the 2500K at 3.3GHz. Transitioning in and out of different power states can affect the performance of storage benchmarks, especially when dealing with short burst transfers.

The test systems’ Windows desktop was set at 1280×1024 in 32-bit color at a 75Hz screen refresh rate. Most of the tests and methods we employed are publicly available and reproducible. If you have questions about our methods, hit our forums to talk with us about them.

HD Tune — Transfer rates
HD Tune lets us present transfer rates in a couple of different ways. Using the benchmark’s “full test” setting gives us a good look at performance across the entire drive rather than extrapolating based on a handful of sample points. The full test gives us fodder for line graphs, which we’ve split up by drive maker. You can click the buttons below each line graph to see how the Vector and our mechanical hard drive compare to different SSDs.

The Vector gets off to a slow start in HD Tune’s sequential read speed test, and the other OCZ SSDs fare even worse. Since the Vertex 4 and Agility 4 use the same Marvell controller as the Crucial m4, which is much faster here, OCZ’s firmware seems the likely culprit behind the poor performances.

Whatever hurts the Vector’s performance in the read speed test doesn’t seem to slow it down in the write speed test. OCZ’s new hotness climbs to third place behind the Samsung 830 Series and OCZ’s own Vertex 4. The gap to the next clump of competitors is pretty wide, too.

HD Tune runs on unpartitioned drives, a setup that isn’t always ideal for SSDs. For another perspective, we ran CrystalDiskMark’s sequential transfer rate tests, which call for partitioned drives. We used the app’s default settings: a 1GB transfer size with randomized data.

The Vector fares much better in CrystalDiskMark’s read speed test, falling to only the Samsung SSDs. It also offers the highest write speed of the bunch.

HD Tune — Random access times
In addition to letting us test transfer rates, HD Tune can measure random access times. We’ve tested with four transfer sizes and presented all the results in a couple of line graphs. We’ve also busted out the 4KB and 1MB transfers sizes into bar graphs that should be easier to read without the presence of the mechanical drive.

The line graph does a poor job of showing us where the Vector lies relative to the other SSDs, but it nicely illustrates the substantial gap between the access times of solid-state and mechanical storage. At all but the largest transfer sizes, that delta is more than an order of magnitude.

When we narrow our focus to the SSDs in the bar charts, the Vector is easier to isolate. It has the quickest random read access time with 4KB transfers and is practically tied for the lead when reading in 1MB chunks.

The picture is largely the same in HD Tune’s random write tests. This time around, the Vector shares the lead in the 4KB test with the other OCZ drives. The Vertex 4 is the fastest of the OCZ posse in the 1MB test, pushing the Vector into exactly the middle of the pack.

TR FileBench — Real-world copy speeds
Concocted by resident developer Bruno “morphine” Ferreira, FileBench runs through a series of file copy operations using Windows 7’s xcopy command. Using xcopy produces nearly identical copy speeds to dragging and dropping files using the Windows GUI, so our results should be representative of typical real-world performance. We tested using the following five file sets—note the differences in average file sizes and their compressibility. We evaluated the compressibility of each file set by comparing its size before and after being run through 7-Zip’s “ultra” compression scheme.

  Number of files Average file size Total size Compressibility
Movie 6 701MB 4.1GB 0.5%
RAW 101 23.6MB 2.32GB 3.2%
MP3 549 6.48MB 3.47GB 0.5%
TR 26,767 64.6KB 1.7GB 53%
Mozilla 22,696 39.4KB 923MB 91%

The names of most of the file sets are self-explanatory. The Mozilla set is made up of all the files necessary to compile the browser, while the TR set includes years worth of the images, HTML files, and spreadsheets behind my reviews. Those two sets contain much larger numbers of smaller files than the other three. They’re also the most amenable to compression.

To get a sense of how aggressively each SSD reclaims flash pages tagged by the TRIM command, we run FileBench with the solid-state drives in two states. We first test the SSDs in a fresh state after a secure erase. They’re then subjected to a 30-minute IOMeter workload, generating a tortured used state ahead of another batch of copy tests. We haven’t found a substantial difference in the performance of mechanical drives between these two states. Let’s start with the fresh-state results.

The Vector turns in the most impressive all-around performance in FileBench; it boasts the highest copy speeds with our collection of movie and MP3 files and is only a few MB/s off the lead with RAW images. The SandForce-based drives are faster in the TR and Mozilla tests, which use highly compressible files, but the Vector beats everything else. Note, too, that the SandForce drives are only a little bit faster when they’re leading the Vector but quite a bit slower when trailing it.

Some of the SSDs, including the SandForce-based models and the Samsung 840 Series, have slower copy speeds in our used-state FileBench tests. The Vector doesn’t skip a beat. Although it can’t catch the fastest SandForce offerings in the Mozilla and TR tests, OCZ’s latest remains the top dog overall.

TR DriveBench 1.0 — Disk-intensive multitasking
TR DriveBench allows us to record the individual IO requests associated with a Windows session and then play those results back as fast as possible on different drives. We’ve used this app to create a set of multitasking workloads that combine common desktop tasks with disk-intensive background operations like compiling code, copying files, downloading via BitTorrent, transcoding video, and scanning for viruses. The individual workloads are explained in more detail here.

Below, you’ll find an overall average followed by scores for each of our individual workloads. The overall score is an average of the mean performance score for each multitasking workload.

Impressive. The Vector is the first SSD to reach 5000 IOps in DriveBench overall. More importantly, it’s ahead of all the other SSDs by a decent margin.

There’s nothing quite like complete dominance. The Vector tops the standings in each and every test, showing no weakness in our simulated multitasking workloads.

TR DriveBench 2.0 — More disk-intensive multitasking
As much as we like DriveBench 1.0’s individual workloads, the traces cover only slices of disk activity. Because we fire the recorded I/Os at the disks as fast as possible, solid-state drives also have no downtime during which to engage background garbage collection or other optimization algorithms. DriveBench 2.0 addresses both of those issues with a much larger trace that spans two weeks of typical desktop activity peppered with multitasking loads similar to those in DriveBench 1.0. We’ve also adjusted our testing methods to give solid-state drives enough idle time to tidy up after themselves. More details on DriveBench 2.0 are available on this page of our last major SSD round-up.

Instead of looking at a raw IOps rate, we’re going to switch gears and explore service times—the amount of time it takes drives to complete an I/O request. We’ll start with an overall mean service time before slicing and dicing the results.

If we consider all the I/O requests in our two-week DriveBench 2.0 trace, the Vector has the quickest mean service time overall. The Corsair Neutron GTX and Samsung 830 Series are 0.05 milliseconds behind, which is a relatively small margin considering the service times of some of the slower SSDs.

Splitting the read and write requests shows us why the Vector fares so well overall. It has the lowest mean service time with writes and is within striking distance of the lead with reads. None of the other drives is as competitive with both kinds of I/O.

There are millions of I/O requests in this trace, so we can’t easily graph service times to look at the variance. However, our analysis tools do report the standard deviation, which can give us a sense of how much service times vary from the mean.

Despite its low mean, the Vector does exhibit more variability in its write service times than some of the more consistent SSDs. That said, the standard deviation isn’t high enough to raise a red flag. The Vector’s read service times are more consistent than those of any of the other SSDs, as well.

We can’t easily graph all the service times recorded by DriveBench 2.0, but we can sort them. The graphs below plot the percentage of service times that fall below various thresholds. You can click the buttons below the graphs to see how the Samsung 840 Series compares to SSDs from other drive makers.

Yeah, I’m a total nerd for loving graphs like these. The write distribution results aren’t terribly interesting; there are relatively small gaps between the drives, and the curves all follow the same basic shape. Clicking through the different read distribution graphs nicely highlights the differences between the Vector and its competition, though.

The Vector doesn’t have the highest percentage of extremely quick read service times under 0.2 milliseconds. However, it’s much more competitive when the threshold rises to 0.3 milliseconds and higher. Only the Samsung 840 Series beats the Vector with the threshold between 0.3 and 0.6 ms. The results start to converge after that, with only the slowest drives setting themselves apart as we approach the 1-ms threshold.

As the distribution plots illustrate, service times over 100 milliseconds make up a tiny fraction of the overall results. Those extremely long service times have the potential to cause the sort of hitching that a user might notice, so we’ve graphed the individual percentages for each drive.

Unlike most of the other SSDs, the Vector has similar percentages of really long service times with both read and write requests. Those percentages are very low, but keep in mind that we’re talking about millions of I/Os spread over a two-week period. In three its runs through our test trace, the Vector’s service time eclipsed 100 milliseconds for 1,244 read requests and 645 writes.

Our IOMeter workloads feature a ramping number of concurrent I/O requests. Most desktop systems will only have a few requests in flight at any given time (87% of DriveBench 2.0 requests have a queue depth of four or less). We’ve extended our scaling up to 32 concurrent requests to reach the depth of the Native Command Queuing pipeline associated with the Serial ATA specification. Ramping up the number of requests also gives us a sense of how the drives might perform in more demanding enterprise environments.

We run our IOMeter tests using the fully randomized data pattern, which presents a particular challenge for SandForce’s write compression scheme. We’d rather measure SSD performance in this worst-case scenario than using easily compressible data.

There’s too much data to easily show on a single graph for each access pattern, so we’ve once again split the results by drive maker. You can compare the Vector’s performance to that of the competition by clicking the buttons below each graph.

The web server access pattern is a special case, since it’s made up exclusively of read requests. That makeup doesn’t faze the Vector, which achieves higher transaction rates than everything but the Samsung 840 Series. Unlike a lot of the other drives, the Vector doesn’t slow down as the load ramps up beyond eight concurrent I/O requests .

The Vector easily trumps the Samsung SSDs in the rest of the IOMeter tests. OCZ’s own Vertex 4 proves to be stiffer competition. The older OCZ drive offers slightly higher transaction rates overall, though the Vector bests the Vertex 4 in each workload with 32 concurrent requests. Still, the two OCZ drives are no match for Corsair’s Neutron SSDs, which are way ahead regardless of the load.

Boot duration
Before timing a couple of real-world applications, we first have to load the OS. We can measure how long that takes by checking the Windows 7 boot duration using the operating system’s performance-monitoring tools. This is actually the first test in which we’re booting Windows off each drive; up until this point, our testing has been hosted by an OS housed on a separate system drive.

Level load times
Modern games lack built-in timing tests to measure level loads, so we busted out a stopwatch with a couple of reasonably recent titles.

Like the other SSDs, the Vector has much quicker load times than our mechanical hard drive. The margin is big enough that you’ll notice, but you might have a hard time differentiating between the Vector and the other SSDs. Even though OCZ’s latest drive is never more than 0.3 milliseconds off the lead, most of the SSDs are stacked within a second of each other.

Power consumption
We tested power consumption under load with IOMeter’s workstation access pattern chewing through 32 concurrent I/O requests. Idle power consumption was probed one minute after processing Windows 7’s idle tasks on an empty desktop.

The Vector’s idle power consumption sneaks just under one watt, which is reasonable but unremarkable. The drive’s power consumption climbs by a factor of more than 4X when we apply our IOMeter load, making the Vector one of the most power-hungry SSDs we’ve measured.

The value perspective
Welcome to another one of our famous value analyses, which adds capacity and pricing to the performance data we’ve explored over the preceding pages. The Vector isn’t selling online just yet, so we’ve used the 256GB model’s $270 suggested retail price in our calculations. We used Newegg prices for the rest of the field, and we didn’t take mail-in rebates into account when performing our calculations. We also waited for Newegg’s Black Friday and Cyber Monday specials to expire to prevent those limited-time offers from tainting the results.

First, we’ll look at the all-important cost per gigabyte, which we’ve obtained using the amount of storage capacity accessible to users in Windows.

Ouch. The Vector is above the dollar-per-gigabyte threshold by five cents, making it one of the most expensive consumer-grade SSDs around. Among the drives we’ve tested, only the Intel 520 series and Corsair Neutron GTX have a comparable cost per gigabyte. The Intel 320 Series costs even more, but that drive is a bit of an outlier. Despite its sluggish 3Gbps interface, the 320 Series has become a popular choice in server circles, allowing it to maintain a price that’s inflated by consumer SSD standards.

Our remaining value calculation uses a single performance score that we’ve derived by comparing how each drive stacks up against a common baseline provided by the Momentus 5400.4, a 2.5″ notebook drive with a painfully slow 5,400-RPM spindle speed. This index uses a subset of our performance data described on this page of our last SSD round-up.

Yeah, that’s a pretty definitive result. The Vector scores higher than any of the other SSDs in our overall performance index—by a sizable margin. OCZ’s first Barefoot 3-infused drive is 79 percentage points ahead of the next-closest competitor. To put that gap into perspective, check out the tight spacing in the second-tier pack. The next seven SSDs are separated by just 39 points.

Now for the real magic. We can plot this overall score on one axis and each drive’s cost per gigabyte on the other to create a scatter plot of performance per dollar per gigabyte. The best place on the plot is the upper-left corner, which combines high performance with a low price.

The Vector combines its high per-gigabyte cost with the best overall performance, allowing it to rise above the fray on our scatter plot. Unfortunately, the drive’s price premium counterbalances its performance advantage.

Although the Samsung 830 Series seems to be ideally placed on the plot, finding the drive in stock online is becoming increasingly difficult. The 830 Series is being replaced by the 840 Series and the 840 Pro, and it won’t be long before the older model is gone for good. Newegg has run out of the 256GB model already, and Amazon’s stock seems to be dwindling.

Indilinx’s next-gen SSD controller makes its long-awaited debut in OCZ Vector SSD, and it looks like the chip was worth the wait. Simply put, the Vector is one of the fastest SSDs ever to land in our labs. It posted the best performance we’ve ever measured in a number of tests and easily claimed the top spot in our overall performance index. More impressive is the fact that the drive is competitive regardless of whether the I/O consists of sequential, random, read, or write requests. The Vector’s performance is very well balanced.

We’re less enthusiastic about the price, though. OCZ’s suggested retail for the 256GB model we tested is $270, which works out to $1.05 per gigabyte. That’s a little steep in today’s climate of deep SSD discounts, an environment that OCZ is largely responsible for fostering. Given the firm’s history, I’d expect the Vector’s street price to fall below the official MSRP. The only other consumer-grade SSD around $270 right now is Samsung’s top-of-the-line 840 Pro, which has both full-disk encryption support and a free copy of Assassin’s Creed III—perks the OCZ drive lacks.

The Vector’s five-year warranty is a nice perk in its own right, of course, but we all know longer warranties don’t guarantee better reliability. Only time will tell if OCZ’s investment in additional validation testing has produced a solid SSD free of troublesome issues. With its hands in everything from the controller to the firmware to the NAND packaging, OCZ will be the only one to blame if there are any problems with the Vector. It will also deserve all of the credit if the drive proves to be resilient.

OCZ plans to keep the Barefoot 3 controller to itself, giving the company an edge over rival SSD vendors that lack proprietary controller technology. Those without are forced to draw from the same shallow pool of controller silicon, while OCZ can expand its range of Barefoot-based offerings to include less expensive models with cheaper NAND and enterprise-oriented units that should command higher margins. As long as this latest Indilinx controller is as robust as previous designs, OCZ should be well-positioned for the next wave of SSDs.

Right now, the Vector looks like an especially good fit for high-performance desktop systems. I’d have a hard time recommending that folks pay full price, though. $270 is a lot to spend when so many quality drives are available around the $200 mark—and especially since all the ones we tested offered largely comparable performance in our load-time tests. The Vector certainly isn’t a raw deal, but its value proposition would be improved by a lower asking price.

0 responses to “OCZ’s Vector SSD reviewed

  1. You also seem to be pushing this review….like you are trying to sell the 840 off to people.

  2. Then we’ve found your problem. I’m sorry that common courtesy is not something you find appropriate to show when you present yourself online.

  3. Since that’s the case, it can safely be assumed that the current layout is “broken” and “Related articles” should be moved to the very bottom.

  4. Nope, the better Samsung SSD really is in my laptop. I just can’t be bothered with the hassle of swapping them over.

  5. “difference between [b<]properly done SSDs[/b<] is not very noticeable in real world." "I have a 240GB [b<]sandforce[/b<] drive in my 2500K" Explanation is in bolded parts.

  6. Still not clear on why they decided to stop working with us, but it appears some undelivered or unanswered emails were the initial cause of the disconnect with their hired PR firm. Really some silly, simple snafu that turned into a break in communication. Nothing exciting or dramatic, just unfortunate.

  7. This.

    When I picked up my $230 3770k at microcenter last weekend, they had 830 256gb for $150! This wasn’t in any of their ads or listed on the site, and there were 3 in the case at the time.

    I should bought every damn one of them!

  8. We should talk about something besides your lack of manners since the situation is intractable. What do you think about AsRock motherboards?

  9. ohhhhh, I found it now. If I had a Korean monitor I might have seen it with all that vertical space…lol

  10. It’s not just me bro. I’m just the only one dumb enough to still be talking to you about it.

  11. No one *wants* to have an emotional discussion with you. A couple of us are asking you to come across as less of a dick, but think about it, if you come across as reasonable and polite, there’s nothing for anyone to get emotional about! You have lived among humans for some time now, right? This concept is familiar to you? In other words, if you want not to be on the receiving end of so much emotional response, YOU can do something about it: quit making yourself such a target for it. Until you recognize you play a very pivotal role in shaping the conversations you have with others ( DUH! ), you’re going to continue to get some blowback, Mr. Obvious.

  12. A thrice-off anomaly (Anand had two fail, and one of the other site writers had one fail) but fair enough, it was prerelease firmware. As long as production drives are good, then great. Would you be willing to give the same benefit of the doubt to OCZ as long as their drive doesn’t show any wierd failure problems?

    I think this is the issue – all OCZ drives get painted with a wide brush. Some of it is not unique to OCZ, namely Sandforce issues which all Sandforce drives had, and since we like quoting behardware return rates who knows how many returns were due to that versus different defects. Their cheap drives (Petrol, Octane) were, well, cheap and got lots of returns. I feel a little bad for the people who didn’t research those drives first. And here’s a detail in the conclusions of those return statistics that’s worth noting [b<]"The most popular ranges, namely the Vertex 3s and Agility 3s, do relatively well with returns of 1.51% and 2.03% respectively."[/b<] So with their most recent non-lowend drives OCZ's last drive lineup did much better. So do we count that as a positive trend, at least for the non-lowend drives? The point is, OCZ seems to have improved their reliability. Perhaps the management shakeup bodes well too. Time will tell, and it's OK to be skeptical and wait and see, but as with any new drive including the 840 we ought to have the same attitude. Even mighty Intel had, and has, firmware bugs (8MB bug) plague them so we should apply the same skepticism to all new drives. "Past performance does not predict future returns" as stock market people like to say, and that saying cuts both ways.

  13. I’m curious why wasn’t it made clear that the 840 Pro review is delayed due to issues with Samsung? This entire debacle of a comments thread could’ve been avoided, since as abrasion says, other sites have had their reviews for near a month

  14. Like I said, you hand’t said it to me and I haven’t read all the trash you’ve posted to other people – I try to limit my exposure to your drivel. I’ll give you the benefit of the doubt that the SSD 840 is faster, but I’ll have to see it run on the same test bench to be sure. And even then, meaningfully faster? I’d not give a damn over a few precents. Meanwhile, Anandtech has experienced a 100% failure rate – 2 of 2 drives have failed. I said before that I’d guess Samsung will promptly fix the problem, but the problem seems to be pretty darn bad and you don’t even acknowledge it.

  15. LOL. That’s the first time you’ve said anything about that to me, so you can’t say I’m missing points that you haven’t before made. But, to humor you, yes I have heard of troubles with OCZ drives. I have four of them deployed around the office at work and have had no problems with any of them, though. For some reason, anecdotes of positive experiences with OCZ drives don’t get near as much traction as anecdotes of negative experiences. I don’t discount that others have had problems but I have to balance that with my own firsthand experiences. At the same time, I have personally had some trouble with a Crucial M4 drive, which only required a firmware flash to fix, and I see that Anandtech reports that both his initial SSD 840 Pro and it’s replacement have died on him. I’m confident Samsung will fix that, but at the same time it makes it dubious to criticize only OCZ while keeping mum about everyone else’s troubles. I’m certainly no fan of OCZ, but I’m not a hater either. I just know I have 4 OCZ drives with zero problems and 1 Crucial drive and it had a problem. I also have a few Intel drives deployed and no problems with any of them.

    If it means anything to you – although I can’t imagine it will, I’ll through a little SSD 840 Pro exception into my first post just to be a sporting fellow.

  16. Did you write that the wrong way around?

    [quote<]I have a 240GB sandforce drive in my 2500K and a 256GB 830 in 'cheaper-than-i3' B970 laptop[/quote<]

  17. I don’t know. I have a 240GB sandforce drive in my 2500K and a 256GB 830 in ‘cheaper-than-i3’ B970 laptop.

    For whatever reason (probably compressed dataz + Sandforce) the crummy 2.4GHz dual core is feeling snappier than my big, beefy 4GHz desktop with double the RAM.

  18. As a matter of fact, no I haven’t. Things can always break, having a supercap is just a layer of protection not a 100% guarantee because it could fail too right? But the cost seems too high to compete in consumer drives. And if worse comes to worse and a PSU blows up your computer: 1) a supercap might not save you if the SSD gets destroyed by the PSU 2) that’s what backups are for.

  19. my god. i thought abrasion was just his username. i didnt realize it referred to the painful butthurting feeling his posts inspire in the rectums of readers

  20. You missed my point. I don’t see how the price of the Vector is any issue since the SSD 840 Pro is retailing for just as much.

  21. As Geoff noted in the article, the drives have gotten better from a reliability standpoint since the 3 series and the BSoD bug. You really don’t have enough data points to make a argument like that. You also are missing “Quality, expensive, and speed vs speed”

    Samsung 840s aren’t the 830s.

  22. You need to work on phrasing your posts better. Your posts often have a pretty contemptuous and petty tone. If you were aiming for thoughtful yet respectful, you missed.

  23. [quote<]OCZ were first to market[/quote<] I have to admit, the Petrol/SATA-II Octanes were a mistake. This is the mismanagement I mentioned and yes, these should never have been on the market in the first place. Of the list that behardware brings up, only two items on it are still sold by OCZ: The Agility 3 and the Vertex 2, with behardware noting that as older products vanish from the market, OCZ's reliabilty is improving. Denying that OCZ was problematic is silly. They definitely had QC issues. I'm just saying that when stuff breaks you want simple, fast returns. I have been mostly buying Samsung 830's and Intel 330's since they've been available but this Vector is something that I will probably try in a few machines to see how it goes.

  24. Love how you try to make TR look like assholes every time you have a problem. It’s so endearing.

  25. What this review tells me is: BUY EVERY SINGLE Samsung 830 256GB before they are GAWN!

    Even looking at Anand’s review that includes the 840 pro… you still come to the same conclusion. The performance premium for the 50% markup is NOT worth it for the Vector OR the 840 Pro when the difference in performance is ~2-5%.

  26. 840 pro has a hefty price premium associated with it and it’s already on the market. OCZ drives generally are never around MSRP or if they are they’re usually reduced in price within a month.

  27. I think drive compression is a good thing, if done in the OS with a mind towards the drive or on the drive if uncompressed performance isn’t lacking.

    When we switch to SATA on PCIe, the next generation of Sandforce controllers may be in the lead again, as uncompressed performance may very well be limited by the aggregate speed of the NAND, and compression will be the only way to overcome that.

  28. Possibly, but I have yet to see or hear about drives dieing from the write ceiling. In the case like this the controller is made for a non-compressed workload and it does quite well. So that just means the system would have to deal with NTFS compression if enabled, which also works quite well now days.

    There is some talk around the net about this already and a few benchmarks scattered about. It’s never really been put into context when it comes to SSDs. I actually think on drive compression is sort of like cheating in a sense.

  29. Being the newly crowned fastest drive in it’s segment (well, maybe 2nd fastest, pending a review of the SSD 840 Pro once Samsung figures out how to keep them from dying after just a few days of use), of course it’s going to carry a price premium. That doesn’t hurt my feelings. The drive puts on an impressive show and if OCZ really has been as serious about reliability as they claim, then it’s really an excellent drive.

    As with all computer parts, prices will come down in time.

  30. Component return rates:

    [url<][/url<] Notice how OCZ's average return rates are about ten times that of Intel for example. Notice how OCZ absolutely dominates the worst return rates chart. How about something ridiculous: 40.00% for the OCZ Petrol 64 GB Really? FORTY percent return rates??? I'm gonna wait and see if they got their act together, a dozen RMAs are not acceptable in my book, regardless of how "nice" the RMA folks may be.

  31. I agree;

    But it also sounds like it might turn out to be a net negative, depending on how the controllers deal with it, and it might also increase write amplification.

  32. What was wrong with the relationship with Samsung?

    I wonder if you can tell us how company shakeups affect what you do. You guys always seemed to have a good, but no better than other sites, relationship with OCZ. Anand always seemed to have an ‘inside track’ with Ryan at OCZ, getting product early or posting in-depth previews well before other sites. Now TR has a rather big OCZ release fully reviewed before Anandtech. I am curious how the management changes at OCZ may have affected this (not that I expect you to speculate about Anandtech) or maybe just some general ‘insider views’ of the product review business.

    *Anand’s review is now posted.

  33. We have been working to repair relations with Samsung, and you saw an 840 review here recently. As Geoff stated in the text of this review, we are slated to receive an 840 Pro review unit, but Samsung delayed it while working on a firmware update. Our bet is that the update will address the early deaths apparently affecting many 840 Pro drives. Samsung has been reluctant to send the drive as-is, so… I woundn’t recommend buying one right now.

    I think one reason you’re being modded down is that folks perceive you as being a jerk for no reason. Life might go easier for you if you worked not to create that impression in public. 😉

  34. I agree, slightly annoying quirk I just dealt with instead of actually pointing out for years.

  35. Definitely looks like a nice drive. I’m guessing MSRP is going to be no where near what these drives will actually sell for or they wont sell at all. Considering the poop that Samsung dished out with the 840 and the the attached price which was like salting the wound, I think I may be recommending a different drive and it wont be because OCZ is the cheapest.

    Honestly though, the best part about this drive in my opinion is the lack of compression. This may not seem like a big thing to most users, but if there is no compression that means you can enable compression yourself! Like turning on NTFS compression. It increases the amount of space on your SSDs (which are usually small) and increases the speed since they have near instantaneous randoms. The only downside to compression is of course CPU overhead, but with modern day CPUs thats next to nill. I’m really surprised hardware websites don’t cover this aspect.

    If you enable NTFS compression on a drive with compression already on it, the results are usually a net negative. You increase your CPU overhead and don’t really get that much extra space, it usually results in a negative speed increase too.

    I would actually suggest TR take a look at this aspect for SSDs, it completely changes the ball game. It’d probably get more coverage if it wasn’t universally understood that compression sucks on mechanicals.

    Oh death by writes was hilarious. XD

  36. People are modding you down because you act as if TR determines which hardware manufacturers send for review.

  37. I believe it is where it is because it is at the bottom across all pages. The thing that is different about the last page is not the location of the drop-down, but the fact that there is additional content at the end of the article that is not on all the other pages of the article.

  38. Could we get real world battery life comparisons on these SSDs? Idle and a specific load might not be telling us the whole story.

  39. Thats surprisingly far ahead of the pack! Well done Indillix/OCZ. Will be interesting to see it against the 840 pro since the 2 are likely to occupy the higher range of prices. Waiting for Anand to bench the Vector or TR to bench the 840 pro, then we can get a real comparison. This will be tough to top though.

    Ive been putting off getting an SSD due to cost, and the fact that i loathe reinstalling windows since i always forget some programs that i need to reinstall or settings i need to change, and it takes forever to get things back to how i like it. But i think after xmas im going to have to count the pennies and see if i can grab a decent 256GB SSD, this will be on my comparison list no doubt.

  40. You’re the only one who didn’t know the Vector was coming (it was annoucned as far back as September).

    Also, you’ve got a lot of balls to demand TechReport’s review of the Samsung 840 Pro. You’re not in any kind of position to tell them anything. They’ve said they’ll review it when they receive it.

  41. Have a Kingston using the Barefoot Eco in a Sibling’s machine; it’s noticeably slower than anything else, yet faster than a spinner, and definitely more reliable than either Sandforce controller.

  42. The use of Marvell and now an Indilinx controller in their latest offerings brings significant hope, where if they’d continued to rely on Sandforce they might already be gone; without looking at the books, of course. But a solid performing controller developed in-house is a solid step forward, and what it’s going to take to compete with Intel and Samsung.

  43. Funny, and also true- OCZ bet pretty big and seemed to take much of the public loss, and yet I don’t know of a manufacturer that hasn’t caught hell for their SSDs.

  44. lol! Ooooh, it’s below all the other junk, er, article links that aren’t on any other review page, and much farther down than my usual windows viewing size (or even if I go full screen with 1200 vertical pixels). Since I read every single TR article I’ve never bothered scrolling down through the article links. I guess the article links on the final page are included as part of the equivalent to the ‘text block’ on the other pages. It should really be directly below the text of the article just like every other page in the review.

    Basically, it should look like this: [url<][/url<] (the extra comments link above 'email author' could be removed.) instead of this: [url<][/url<] in order to have consistency between article pages. It doesn't seem like I was the only one missing it either, unless SecretMaster was being secretly condescending 😉

  45. Most newer SSDs track total host writes with a SMART value. You should be able to read it with any SMART app, although I’ve only used the manufacturer’s utility software.

  46. Quick Question for the TR staff or knowledgeable Gerbils: You mention that the warranty is contingent on either 5 years or the 36.5 TB write quota. Is there any easy way to find out the lifetime writes on an SSD? For example, will a SMART utility give you that information, or is it proprietary data that you can’t get access to?

    I have a couple of SSDs from different manufacturers that have been in use for some time, and I’d be curious to see what my lifetime write totals have been.

  47. I thought at first it was funny. Then I realized it was probably true, and so I became sad.

    +1 for tugging at my heartstrings.

  48. Sandforce have been very quiet for a long while now. 12 months ago they were discussing progress with Anand and were throwing around numbers like 2x performance on small files. That’s 170K IOPS, if you do the simple math.

    I’m also curious to see how long it will take Intel to bring their new controller (as in the DC S3700) to the consumer market.

  49. I’m hoping OCZ does well, comes down in price, and turns out to be reliable. You can’t blame them for selling a good product at a premium since it [b<]*IS*[/b<] good and they sorely need some revenue to recover from some past mismanagement. All the OCZ haters can carry on hating, but I actually have a soft spot for the company. Having experienced their RMA process a dozen times or so I have never had any major complaints. (I was never really counting but I would estimate that I've upgraded-to-SSD or built-with-SSD ~400 times now, and that doesn't include all the machines that came pre-built with SSD's in them) Yes, some of the cheaper models had higher failure rates but statistically no worse than other 'value' brands (Intel and Samsung being notably more expensive and reliable). OCZ has always handled my RMA's with a simple, no-quibble system that kept me buying them. Other brands (Patriot, ADATA, SanDisk) I have been actively avoiding after cursing and fighting my way through their RMA processes. My guess is that OCZ were first to market with Sandforce, represented the largest portion of Sandforce users, and therefore received the worst of the vocal minority when the SF issues outside their control really hit the fan. Certainly a lot of the butthurt was caused by people messing around in the user forums when they should have been submitting RMA's.... [i<]Edit: Oh, and nice review Geoff. Not only have you beaten Anand to the punch this time, but it was thorough enough that I probably won't bother doing more than skimming his review when it turns up.[/i<]

  50. It seems I can once again call OCZ a company who’s products are worth buying (even if its just this lineup). Doesn’t mean ill stop using an old vertex 2 that bricked my computer for a short while as a miniature snack tray.

  51. I don’t understand how anyone can release a drive without supercap functionality 2012, even for non enterprise. As Intel demonstrated with the 320, this can be done cheaply with a gang of ordinary caps. Nice numbers numbers and all, but for this reason I’ll never consider this drive.

  52. The performance is impressive. Sadly the price is, too.
    SSD’s are still far above my “price tolerance” when it comes to storage, the Samsung 830 came closest so far.

  53. Wow, colour me impressed. Glad to see Indilinx pull a fast’n’clever one out of the hat – I’m still a proud owner of two original vertex drives that are still trundling along nicely, although have been a steadfast crucial/samsung convert in recent years after what I’ll still refer to as the sandforce debacle.

    Hopefully OCZ can prove to all us cynical types that they’ve turned over a new leaf as far as reliability and support goes… time will tell.

  54. I second this opinion on MadMan’s second point. It’s been a slightly annoying feature for a long time!

  55. Geoff – two things:

    1) For these new SSDs, could some of the TR staff use them in their daily work rigs to see if any problems crop up, or to test any issues that users report later? Keeping a backup with a reliable SSD would be necessary of course but I assume you guys do that anyway.

    2) Could TR please add the article navigation drop-down bar to the last page of reviews? I often get to the last page of a review and want to return to a previous page, but I have to go back to another page then find the page I want. It would help page layout consistency and navigation. Thanks!

  56. Well here’s to hoping that this drive proves to be reliable, that would most certainly make their next big release a bit more exciting.

  57. [quote<]noting that the Vector is its "most extensively and comprehensively tested consumer SSD line to date."[/quote<] read: "We actually tested this SSD once or twice instead of just releasing it." [quote<]A "large network of beta testers" was part of the validation process[/quote<] Wait, did they test it inhouse? I always thought the early adopter retail customers were the OCZ beta testers. 😀 #OCZhatertroll

  58. Excellent review guys, particularly with the inside-the-millisecond benchmarks. My one request: Do some sort of head to head between the Vector and the Samsung 840 Pro when you get the chance. Those two drives look like the high-end contenders in the SSD space for the next cycle.