OCZ’s Octane 512GB solid-state drive

OCZ has been at the forefront of the desktop SSD market for several years now. Despite being something of a pioneer in the industry, OCZ has been more of a remix artist than a true original; the company’s consumer-oriented SSDs have long used controller and flash silicon sourced from other firms. Flash fabs are rather expensive to build, making it unlikely that OCZ will ever be able to challenge giants like Intel, Micron, and Samsung in that arena. Controller makers are considerably more affordable, however. Earlier this year, OCZ purchased one for $32 million.

Acquiring Indilinx, the brains behind the Barefoot controller in the first Vertex solid-state drive, put OCZ in a rather exclusive club. Only a handful of companies design the flash controllers and accompanying firmware that underpin modern SSDs. Precious few of those firms sell complete products based on their wares, especially when one considers only the 2.5″ SATA drives typically targeted at consumers.

Now, some nine months after Indilinx joined the family, OCZ is rolling out a line of Octane SSDs based on new controller silicon dubbed Everest. With a 6Gbps Serial ATA interface, eight memory channels, AES encryption, and support for the latest NAND flash, the Everest chip ticks all the boxes you’d expect from a modern SSD controller. The more important question, however, is whether the Octane SSD can keep up with its contemporary rivals. We’ve tested the drive against 16 other SSDs and a stack of mechanical disks to find out.

Climbing Everest

Before getting started, we should note that there are actually two lines of Octane SSDs. Our focus today is on the standard Octane, but there’s also a cheaper S2 series equipped with a 3Gbps SATA interface and slower NAND. Both Octane families use the same Everest controller, which still bears Indilinx’s name.

OCZ is being tight-lipped about many of the details surrounding the controller. However, we do know that it incorporates dual ARM cores to handle processing duties. The flash interface is spread over eight memory channels and supports 16-way interleaving, while the host interface supports the latest Serial ATA standard.

The Octane’s marketing materials are quick to hype nDurance, a feature that purportedly increases the endurance of the drive’s flash chips by a factor of two. OCZ would reveal only that nDurance uses a hardware engine combined with “proprietary algorithms in the FTL,” or flash translation layer. It seems unlikely that a twofold increase in NAND endurance could be achieved without some form of black-magic compression, so nDurance could employ similar techniques to those used in SandForce controllers. I suspect the hardware engine might be related to the controller’s support for 256-bit AES encryption. If an encryption engine is present, its algorithms could be tweaked to reduce the size of incoming writes from the host.

With an eye toward mainstream desktop workloads, OCZ has optimized the Octane for queue depths of three or less. The command queue associated with the current Serial ATA spec extends to a depth of 32, making OCZ’s focus look rather narrow. That said, our own DriveBench 2.0 workload, which includes two weeks of multitasking-heavy desktop usage, agrees with OCZ’s approach. More than 80% of DriveBench 2.0 requests have a queue depth of three or less.

OCZ says the Octane has been tuned to perform well with both 4KB and 8KB transfer sizes. The company also claims the drive sports “innovative latency reduction technology,” although it wouldn’t divulge specifics on what’s innovative about the approach. The Octane’s read and write latencies are listed as 0.06 and 0.09 ms, respectively.

The Everest controller will work with a range of different flash types, including ONFI 2.0 and Toggle DDR 1.0 NAND at speeds up to 200 MT/s. OCZ didn’t mention support for SLC memory when it announced the Everest, but it did reveal that the controller will work with TLC memory. TLC NAND packs three bits per cell, one more bit than the MLC flash commonly found in today’s SSDs.

We’ve yet to see TLC NAND incorporated in a solid-state drive, and the Octane doesn’t change that trend; the drive uses 25-nm MLC NAND chips manufactured by Intel. These synchronous chips appear to be similar to the ones that populate OCZ’s Vertex 3 SSD and other synchronous SandForce drives. The asynchronous memory typically used in cheaper SandForce SSDs will also appear alongside the Everest controller in the slower Octane S2 SSDs.

Our Octane sample came with 512GB of flash spread between 16 chips. SSD makers typically send out review units with capacities in the 256GB range, making OCZ’s decision to go with a nearly $900 drive seem a little odd. The performance specifications for each capacity point illustrate the likely motivation behind that move.

Capacity Max sequential Max 4KB random Price
Read Write Read Write
128GB 535MB/s 170MB/s 37,000 IOps 7,7000 IOps $200
256GB 535MB/s 270MB/s 37,000 IOps 12,000 IOps $370
512GB 535MB/s 400MB/s 37,000 IOps 16,000 IOps $880
1TB 560MB/s 400MB/s 45,000 IOps 19,500 IOps TBA

Until the 1TB version of the Octane arrives next month, the 512GB model has the fastest write speeds of the bunch—and by quite a lot. The 256GB Octane’s 270MB/s peak sequential write speed is 130MB/s shy of the rate purportedly achieved by the 512GB model. There’s also a 4,000-IOps delta between the two for random 4KB writes. The 128GB variant’s write performance is another step down the ladder in both categories, although there’s no difference between the three in the read department.

The Octane 128GB has the sort of middling performance ratings one might expect from the 64GB member of an SSD family. SSDs of that size typically use only a handful of flash chips, preventing them from exploiting all of the parallelism available in the controller. Given the Octane’s performance ratings and the fact that our 512GB drive uses 32GB NAND chips, I have a sneaking suspicion the 128GB model has only four flash chips under the hood. OCZ won’t be making a 64GB version of the Octane, nor does it plan to offer that capacity in the S2 line.

Apart from the terabyte model, all of the 6Gbps Octane SSDs feature 512MB of cache memory. Interestingly, this cache is split between two DRAM chips: one on each side of the circuit board. The DRAM is accessed through a single channel that’s 16 bits wide. Micron provides the DDR3 chips, which can push bits as fast as 1600 MT/s.

OCZ’s projected prices for the Octane look pretty reasonable; the 128GB model is set to sell for $200, which puts it right in the sweet spot. If we’ve learned one thing over the past year, though, it’s that street prices can change on a daily basis. Octane SSDs were supposed to start shipping to retailers last week, so they should be a part of the raging solid-state price war shortly. At the moment, I only see one Octane listing online—a 256GB model at Newegg that’s out of stock.

Like just about every other consumer-grade SSD on the market, the Octane is covered by a three-year warranty. We’d be remiss not to point out that OCZ ‘s reliability reputation has a few blemishes, though. Some of the blood on OCZ’s shirt comes from its position on the cutting edge; the firm has aggressively adopted new controller technology, often releasing drives long before its rivals—and with much earlier, buggier firmware. OCZ’s partners must also shoulder some of the blame. SandForce is responsible for the BSOD bug that afflicted not only OCZ’s Agility 3 and Vertex 3, but also similar SSDs from Corsair, Kingston, and others.

We asked OCZ to explain how it performs validation testing on the Everest controller and Octane SSD, and the company told us it’s been testing Everest since before the Indilinx acquisition. In addition to performing “a significant amount” of interoperability testing with different platform configurations, OCZ says it uses automated test tools from the likes of Flexstar, Oakgate Software, Lecroy, and Ulink—gear that’s also being used by “many of [its] enterprise and OEM customers.” Only time will tell if OCZ’s efforts have rid the the Octane and its Everest controller of any niggling flaws, and we’ll be keeping an eye on the forums to see if end users experience any issues with the drive. If problems do pop up, the onus will be on OCZ alone to fix them.

Our testing methods

Unfortunately, we don’t have any other 512GB SSDs in the Benchmarking Sweatshop. We have tested a number of 240-300GB drives based on SandForce, Marvell, and Intel controllers. Those SSDs all have higher-capacity counterparts, but they aren’t tagged with substantially superior performance ratings for reads or writes. In most cases, the performance specifications are identical, so we should have a good look at how each family of drives performs in its optimal configuration.

Interface Cache Spindle speed Areal density Flash controller NAND
Corsair Force Series 3 120GB 6GBps NA NA NA SandForce SF-2281 25-nm Micron async MLC
Corsair Force Series 3 240GB 6Gbps NA NA NA SandForce SF-2281 25-nm Micron async MLC
Corsair Force Series GT 120GB 6GBps NA NA NA SandForce SF-2281 25-nm Intel sync MLC
Corsair Force Series GT 240GB 6GBps NA NA NA SandForce SF-2281 25-nm Intel sync MLC
Corsair Performance 3 Series 128GB 6GBps 128MB NA NA Marvell 88SS9174 34-nm Toshiba MLC
Crucial m4 128GB 6GBps 128MB NA NA Marvell 88SS9174 25-nm Micron sync MLC
Crucial m4 256GB 6Gbps 256MB NA NA Marvell 88SS9174 25-nm Micron sync MLC
Intel 320 Series 120GB 3GBps 64MB NA NA Intel PC29AS21BA0 25-nm Intel MLC
Intel 320 Series 300GB 3Gbps 64MB NA NA Intel PC29AS21BA0 25-nm Intel MLC
Intel 510 Series 120GB 6GBps 128MB NA NA Marvell 88SS9174 34-nm Intel MLC
Intel 510 Series 250GB 6Gbps 128MB NA NA Marvell 88SS9174 34-nm Intel MLC
Kingston HyperX 120GB 6GBps NA NA NA SandForce SF-2281 25-nm Intel sync MLC
OCZ Agility 3 120GB 6GBps NA NA NA SandForce SF-2281 25-nm Micron async MLC
OCZ Octane 512GB 6Gbps 512MB NA NA Indilinx Everest 25-nm Intel sync MLC
OCZ Vertex 3 120GB 6GBps NA NA NA SandForce SF-2281 25-nm Intel sync MLC
Seagate Momentus 5400.4 25GB 3Gbps 8MB 5,400 RPM 204 Gb/in² NA NA
Seagate Momentus XT 500GB 3Gbps 32MB 7,200 RPM 394 Gb/in² NA* 4GB SLC
Seagate Momentus XT 750GB 6Gbps 32MB 7,200 RPM 541 Gb/in² NA* 8GB SLC
WD Caviar Black 1TB 6Gbps 64MB 7,200 RPM 400 Gb/in² NA NA
WD Scorpio Black 750GB 3Gbps 16MB 7,200 RPM 520 Gb/in² NA NA

Our performance data also includes a number of lower-capacity SSDs and a handful of more traditional hard drives. I’ve grayed out the latter in the chart and in the graphs on the following pages to focus our attention on how the Octane fares against its solid-state competition. Neither the mechanical drives nor the hybrids are in the same league, at least in terms of performance. If you want to see some real carnage, watch for how the Octane’s performance compares to that of our 3.5″ desktop reference, the Caviar Black 1TB.

We used the following system configuration for testing:

Processor Intel Core i7-2500K 3.3GHz
Motherboard Asus P8P67 Deluxe
Bios revision 1850
Platform hub Intel P67 Express
Platform drivers INF update 9.2.0.1030

RST 10.6.0.1022

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 Series 3 120GB with 1.3 firmware

Corsair Force 3 Series 240GB with 1.3.2 firmware

Corsair Force Series GT 120GB with 1.3 firmware

Corsair Force Series GT 240GB with 1.3.2 firmware

Crucial m4 128GB with 0009 firmware

Corsair m4 256GB with 0009 firmware

Intel 320 Series 120GB with 4PC10362 firmware

Intel 320 Series 300GB with 4PC10362 firmware

Intel 510 Series 120GB with PPG4 firmware

Intel 510 Series 250GB with PWG2 firmware

Kingston HyperX 120GB with 320ABBF0 firmware

Corsair Performance 3 Series 128GB with 1.1 firmware

OCZ Agility 3 120GB with 2.15 firmware

OCZ Vertex 3 120GB with 2.15 firmware

WD Caviar Black 1TB with 05.01D05 firmware

Seagate Momentus 5400.4 250GB with 3.AAB firmware

Seagate Momentus XT 500GB with SD22 firmware

WD Scorpio Black 750GB with 01.01A01 firmware

Seagate Momentus XT 750GB with SM12 firmware

OCZ Octane 512GB with 1313 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. The Hybrid drives have also been subjected to five runs, but only in tests that show their performance improving after the first one.

  • 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 data created by the full test also gives us fodder for a line graph.

To make the line graph more readable, we’ve excluded everything but the high-capacity SSDs. The mechanical drives have also been grayed out throughout our graphs. Otherwise, the results are colored by drive maker, with the Octane set apart from the other OCZ SSDs in a different shade.

The Octane’s sequential read speeds aren’t fast enough to put it in the first tier, which is made up of SandForce-based drives and Crucial’s m4. Keeping up with the Intel 510 Series is no small feat for this first Everest-based SSD, though, and it’s hardly far behind the leaders.

If we look at the line graph, there’s a little bit of fluttering in the Octane’s read speed as the test progresses through the drive’s capacity. The amplitude of the noise is small enough that I wouldn’t worry about it.

The same can’t be said for the wildly oscillating write speeds of the Force SSDs. This behavior is typical of consumer SandForce drives with the usual 7-8% overprovisioning. Although the Octane’s write speeds aren’t quite as fast, they are considerably more consistent. That’s good enough to slot the Indilinx drive into fourth place in average write speeds.

HD Tune’s burst speed tests are meant to isolate a drive’s cache memory.

The Octane has the biggest cache of the bunch, but it doesn’t score well in HD Tune’s burst speed tests. Everything else is faster with the exception of Intel’s 320 Series, which is slower only in the burst write test.

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. Once again, we’ve dropped the smaller SSDs and mechanical drives from the line graphs.

The Octane has very low access latencies throughout HD Tune’s random read tests. At worst, the drive is only fractions of a millisecond slower than the top dogs.

The same trend largely holds true in the random write tests, but there’s a little more separation at the 1MB transfer size. The Octane is about a millisecond behind the lead SandForce SSDs when tasked with random one-megabyte writes.

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:

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

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.

To get a sense of how aggressively each SSD reclaims flash pages tagged by the TRIM command, we’ve run FileBench with the solid-state drives in two states. We first test them in a fresh state after a secure erase. The SSDs are 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 states.

Our real-world copy tests prove to be fertile ground for the Octane and its Indilinx controller. The drive rips through the larger files in our first three sets with ease and ends up locking horns with the Intel 510 Series 250GB at the top of the standings. Our Mozilla and TR file sets prove more challenging for the Octane, but the Indilinx SSD still finds itself within arm’s reach of the front of the pack.

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 with each multitasking workload.

The synchronous SandForce setup of the Force GT 240GB can’t be beaten in DriveBench. However, the Octane cranks through I/O requests fast enough to put it the second tier of drives. Breaking down DriveBench’s individual test results will give us a better sense of where the Octane excelled and where it faltered.

Given our FileBench results, it’s no surprise the Octane has a share of the lead in the copy component of our DriveBench suite. The Indilinx-powered SSD also does well when transcoding video, but it doesn’t rank as highly in the other tests.

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 that it takes drives to complete an I/O request. We’ll start with an overall mean service time before slicing and dicing the results.

That’s encouraging. The Octane has the second-lowest mean service time in our latest DriveBench test. Only the Force GT 240GB is quicker, although its lead over the Octane is considerably wider than the OCZ drive’s advantage over the 120GB version of the GT. The other high-capacity SSDs lag a little bit farther behind the leading pack of synchronous SandForce drives.

Breaking down DriveBench’s read and write service times reveals the balanced nature of the Octane’s performance. The drive is near the front of the pack in both cases, and its mean service times are pretty close between the two. The same can’t be said for the Crucial m4, Intel 510 Series, and Corsair Performance 3 SSDs anchored by Marvell’s latest controller. Those drives handle DriveBench’s write requests much slower than they complete its reads.

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.

This measure of variance shows the Octane in a good light, although its service times aren’t quite as consistent as those of the top SandForce drives. The Octane still manages to edge out the Marvell-based SSDs. There’s much more variance in how quickly those drives handle DriveBench’s write requests.

If I haven’t already scared you off with too many graphs and statistics, this next pair will do it. We’re going to close out our DriveBench analysis with a look at the distribution of service times. I’ve split the tally between I/O requests that complete in 0-1 milliseconds, 1-100 ms, and those that take longer than 100 ms to complete.

There are two things to look for here. The first is the percentage of service times that complete in less than a millisecond. The Octane looks good from that angle, nearly matching the Force GT 240GB.

Now, look at the percentage of requests that take over 100 milliseconds to complete—a really long time within the confines of a modern PC. The Octane doesn’t have the lowest percentages of 100+ ms service times, but it does a better job of keeping pace with the standard-setting SandForce drives than the Crucial m4 or the Intel 510 Series SSDs do.

IOMeter

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.

As we’ve done with the other line graphs in this review, we’ve dropped the smaller SSDs and mechanical drives to make the results easier to interpret. IOMeter scores for the smaller SSDs can be found on this page of our last SSD round-up. The results from the mechanical drives can be viewed here.

The Octane runs out of gas in IOMeter (sorry, I couldn’t resist). Although the drive’s transaction rates increase as the number of concurrent I/O requests rises, the Octane fails to catch up with the Force SSDs or the m4. Given OCZ’s apparent focus on queue depths of three or less, I expected the Octane to fare better at the lower end of our ramping IOMeter workloads.

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 time we’re booting Windows 7 off each drive; up until this point, our testing has been hosted by an OS housed on a separate system drive.

For the most part, our higher-capacity SSDs load Windows slower than smaller variants of the same drives. Perhaps that’s why the 512GB Octane, by far the biggest of the bunch, sits at the back of the solid-state field. It’s barely more than a second off the lead, though.

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.

The Octane bounces back in our gaming tests, although the solid-state field remains tightly bunched overall. Only two seconds separate the SSDs in Portal 2, and that delta shrinks to about a second in Duke Nukem Forever.

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.

Despite featuring double the storage capacity of its biggest SSD rivals, the Octane’s power consumption is quite reasonable. The m4 256GB and 320 Series 300GB drives are more power-efficient overall, but the same can’t be said for the larger 510 Series and Force SSDs, which draw more power than the Octane under load.

The value perspective

Welcome to our famous value analysis, which adds capacity and pricing to the performance data we’ve explored over the preceding pages. We used Newegg prices to even the playing field for everything but the Octane, and we didn’t take mail-in rebates into account when performing our calculations. Since the Octane 512GB doesn’t appear to be selling online, we’ve had to use OCZ’s official price for the drive.

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.

On a per-gigabyte basis, OCZ isn’t charging much of a premium for the Octane’s 512GB capacity. I’m hesitant to draw too many conclusions until the drives are readily available, though. In case you hadn’t noticed, SSD makers have been engaged in a hotly contested price war for months. The prices of OCZ’s other SSDs change on an almost daily basis, and I’d expect the Octane’s price to be similarly fluid when the SSD hits vendors in volume.

Our remaining value calculations use 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. Some of the drives were actually slower than our baseline in a couple of the included tests, so we’ve fudged the numbers a little to prevent those results from messing up the overall picture.

Overall, the Octane’s results put it in second place behind the Force GT 240GB. The other high-capacity SSDs sit lower in the standings, with the Marvell-based 510 Series and m4 models more than 200 percentage points behind.

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. To cut down on some of the clutter, I’ve dropped the labels from the mechanical drives, which are cloaked in gerbilesque anonymity. Their data points have been included only to illustrate the overall trend.

The Octane has about the same cost per gigabyte as the Force GT 240GB, which is the only drive to score higher in our overall performance index. Some of the other SSDs cost less per gigabyte, but they’re slower overall, making OCZ’s pricing look reasonable at the very least.

Overall, the SSDs are much faster than the mechanical drives—and a lot more expensive per gigabyte. That dynamic is unlikely to change anytime soon.

Although this analysis is helpful when evaluating drives on their own, what happens when we consider their cost in the context of a complete system? To find out, we’ve divided our overall performance score by the sum of our test system’s components. Those parts total around $800, which also happens to be a reasonable price for a modern notebook.

With an asking price that’s higher than the cost of the rest of our complete system, the Octane 512GB is off on its own in the upper right-hand corner of the scatter plot. The lower capacity points in the Octane family are much cheaper, of course, but we don’t yet know how they perform.

Conclusions

Quite some time has passed since Indilinx rolled out the Barefoot controller that fueled OCZ’s original Vertex SSD. Based on the performance data we’ve collected, Indilinx’s new Everest controller appears to have been worth the wait. This follow-up has all the features we’d expect from a modern SSD controller, and its performance is very competitive. The Octane often found itself near the top of the heap in our performance tests, and it maintained that position through much of our benchmark suite. “Balanced” is perhaps the best way to describe the Octane’s performance, which is good enough for second place in our overall index. Only the fastest SandForce-based drives we’ve tested offer better all-around performance.

High capacities appear to be a focus for the Octane family, which won’t extend below 128GB. The 128 and 256GB models carry much lower write performance ratings than the 512GB drive we tested, making me somewhat hesitant to pass judgment. We certainly know how the Octane 512GB performs, but it costs nearly $900, a price few folks can afford. To be fair, though, that price isn’t out of step with the cost of other 512GB SSDs.

We’d really like to see how smaller versions of this drive stack up against their competition. The $200 and $370 price tags attached to the 128 and 256GB variants certainly look reasonable, but those drives also have much lower write performance ratings than the 512GB model. We’re looking at getting our hands on the other members of the Octane family to see how the drive performs at more attractive price points.

We’re also going to be keeping our ears to the ground to see if end users have more luck with the Octane than they’ve had with SSDs affected by the SandForce BSOD bug. I’m cautiously optimistic that developing the controller and firmware in-house will lead to fewer problems with the Octane than OCZ has experienced with some of its other SSDs. OCZ may well have a winner on its hands, but without additional performance data and a sense of how the Octane is working for the masses, it’s simply too early to tell.

Comments closed
    • Xenolith
    • 8 years ago

    What firmware did you use for the Crucial m4? … doesn’t look like you used the most recent.

      • Dissonance
      • 8 years ago

      We detail the firmware revisions for all the SSDs in our test methods section on page 2. The m4 drives were running the 0009 firmware, which is the most recent.

    • gamoniac
    • 8 years ago

    TR and Geoff, do you plan to review the new Kingston SSDNow V200?
    Thanks.

      • Dissonance
      • 8 years ago

      We don’t have any plans to look at the V200 at the moment. The drive’s performance ratings are pretty underwhelming given the cost, and the JMicron controller doesn’t inspire a whole lot of confidence.

    • Firestarter
    • 8 years ago

    The number of comments is telling: This is a resounding meh.

      • FuturePastNow
      • 8 years ago

      It’s Saturday, and most people don’t have to read the review to know that a 512GB SSD will be priced out of their reach.

        • Chrispy_
        • 8 years ago

        Not just that, the on-paper performance of the affordable sizes doesn’t look very good:

        Capacity = 128GB
        Read = 535MB/s
        Write = 170MB/s
        4K Read = 37,000 IOps
        4KB Write = 7,7000 IOps
        Price = $200

        [i<]edit for table fail[/i<]

    • Chrispy_
    • 8 years ago

    This looks good. I’ve never had any problems with the Barefoot controller and I’d happily buy one of these for my next SSD upgrade (not that I need one).

    I’m still running barefoot drives from years ago, they’ve never missed a beat and IMO they were the best, and by that I mean best-balanced of the early SSDs (they didn’t have the IOPS of the X25-M, but the X25-M had pitiful sequential write speeds).

    Are there any Indinilinx horror stories of firmware bugs, mass drive failure, bluescreens? I’ve not known of any bad press coverage – which the other big players like Intel, Sandforce and Marvell (C300) have been highlighted for in the past.

    • Buzzard44
    • 8 years ago

    Allow me to re-iterate: m4.

    Also, I have a problem with the system value scatter plot. While these marginalize the cost of the device in question (whether it be a hard drive/graphics card/CPU) by reducing their cost to a small percentage of the total costs, the performance of the device in question accounts for 100% of the system performance. The net result is the same graph as the regular value graph very highly skewed.

    Don’t get me wrong, I love what y’all do and visit the site everyday. I love the value graphs, but I figure I might as well throw my opinion out there, as that’s what comments are all about.

    • Ryhadar
    • 8 years ago

    A lot of people are stilling running drives with the original Barefoot and Barefoot Eco controllers, along with the people that are running the early Intel SSD drives. Yet all the major tech sites I’ve seen still seem to concentrate on current hard drives and contemporary SSDs in SSD reviews.

    I’d really like to see the older SSDs in the comparison with the newer drives. Especially since these older drives can still be purchased, and usually significantly lower than current SSDs.

      • Waco
      • 8 years ago

      I’m running a pair of them in RAID 0 and have yet to experience any issues. 🙂

      • MadManOriginal
      • 8 years ago

      You can compare them in Anandtech’s Bench database, although it might not have all the drives you want to compare. I hope a TR version is coming soon 🙂

    • odizzido
    • 8 years ago

    “proprietary algorithms in the FTL,”

    It’s those fracking cylons! They’re in the fracking ship!

    • Meadows
    • 8 years ago

    So, has anyone learnt how the drive features “innovative” “latency-reducing” technology?

      • Chrispy_
      • 8 years ago

      they made the SATA connector with slightly shorter pins so the bits don’t have to travel as far.

        • bobcostas22
        • 8 years ago

        This could make it much faster in the long run, particularly system wide.

    • Compton
    • 8 years ago

    I have no complaint with Indilinx, new or old.

    I suspect OCZ will make a “MaxIOPS” version with 32nm Toggle NAND, making the 256GB model perform closer to the 512GB review samples. As far as lower QD optimization goes, I’ve no problem with that either — that’s just smart. But I can’t escape the fact that these drives seem more like Samsung 830s — Samsung sent out a few 512GB review units as well.

    The move to smaller lithography flash is responsible for much of the price drop for SSDs, but at each stage you have to effectively step up to a higher capacity drive to get the same number of flash devices/channel. Flash is getting faster, as are the controllers and firmware, but 5x and 3x NAND are still “better”. And now we’re about to hurdle head long into a disaster in TLC. I wouldn’t store my music collection on a TLC drive.

      • OneArmedScissor
      • 8 years ago

      [quote<]at each stage you have to effectively step up to a higher capacity drive to get the same number of flash devices/channel.[/quote<] Not always. They can fill them with equal numbers of smaller chips, which is exactly what they tend to do to keep the low to mid capacity models roughly even. Most of the time, whatever the highest capacity drive happens to be, whether it's 256GB, 512GB, or 1TB, that's the one that suddenly sees a sizable write speed jump. If you will notice, those also universally cost more per GB, and that wasn't an accident. :p

      • Palek
      • 8 years ago

      [quote<]I wouldn't store my music collection on a TLC drive.[/quote<] Would you store your collection of TLC albums on a TLC drive? It has to be asked!

      • just brew it!
      • 8 years ago

      The MLC/TLC trend amuses me. For most of my life, things have been transitioning from analog to digital. Now we’re moving back to something that is effectively an analog technology for the core of our next-gen mass storage devices!

      [b<]Edit[/b<]: I'm also quite puzzled as to why the next gen beyond MLC is referred to as TLC (triple level cell). There are actually [u<]8[/u<] distinct voltage levels stored in a "TLC" cell... which corresponds to 3 digital bits... so they are really "triple bit" cells, not "triple level".

    • Dissonance
    • 8 years ago

    OCZ has answered a number of our initial questions about the Octane and its Everest controller, and the review has been updated accordingly. None of our performance data or conclusions have changed, however.

    • Bensam123
    • 8 years ago

    Meh… Krogoth STILL not impressed. While the controller is impressive compared to others on the market currently, this represents a product that just made it to market. The latest generation of Sandforce controllers have been on market for quite some time, so they’ll probably released their next version well before Indilinix does and more then likely with a much larger performance gap (if you look at historical jumps between chips).

    Has TR considered a retroactive price-chart? I’m sure it’s entirely possible to do with some website magic. You guys could add a spot on the front page as an example of moving price points compared to hardware performance. This could really be done with all the hardware categories the site covers and makes performance/price charts for. It would most definitely be a uniquely attractive quality on the site.

    This may also be out of the scope of articles included here, but have you guys also considered testing some PCI-E SSDs? FusioIO has a bunch of them, although they all look like they’d be out of the price range of most consumers except for the lowest end ones. PCI-E SSDs in general seem to be falling into the $300 or less price range as the older technology with smaller storage capacities gets replaced. For example OCZs Revodrive. OCZ also has their own version of a hybrid hard drive, one of their Revodrives.

    I’m still keeping my hopes up for a company to make a volatile memory PCI-E SSD. Something like the I-RAM only with DDR3 and a controller that can actually keep up with it. Possibly with an intuitive software component that allows people to choose what to always have available on board it. All of the Ramdisk software I’ve tried is pretty terrible (buggy, unintuitive, and unstable) and you’re always limited by the space available on your motherboard.

      • Dissonance
      • 8 years ago

      Retroactive price charts sound like a lot of work 😉 You’ll have better luck with the PCIe SSDs, which are becoming more interesting as they become more affordable. As attractive as the Fusion-io stuff is, it’s hard to justify spending a lot of time reviewing something in that price range. It’s likely that we’ll be spending some time with PCIe options once the queue of SATA drives lined up for The Twins is cleared. More storage goodness is on the way 😉

        • Compton
        • 8 years ago

        None of these storage products start with “Cherry” and end with “ville” do they??

        • Bensam123
        • 8 years ago

        Sounds good. It’d be interesting to see how these drives play in raid arrays too. If they scale linearly as one would suspect or if they have other factors that interfere with it, such as TRIM or the read/write bouncing some of the drives show.

        • MadManOriginal
        • 8 years ago

        It might seem like a lot of work but all you’d have to do is set up dynamic links to websites to determine their prices (I am assuming this isn’t terribly hard since price search engines seem to have it down pat) then write some code to insert them into a graph relative to the performance. The performance data would remain static from TR’s review, only the price variable would change. Not sure how you’d want to factor in MIRs.

        A database of review info like Anandtech has would be awesome and adding price/performance to that would put it a step above what any other site does.

          • Firestarter
          • 8 years ago

          [quote<]I am assuming this isn't terribly hard since price search engines seem to have it down pat[/quote<] That's because it's their core business, and they invest a lot of time and money into it. You can't expect TR to do the same. I guess the best TR can do would be to partner up with such a price search engine and display historical data, for example the median over 14 days of the lowest price for the shipped product. Even then, a snapshot of the price at the time of writing would have to be added, because any recommendation in the conclusion would be based on that.

            • Bensam123
            • 8 years ago

            They don’t need to use this particular graph in the article itself. Traditional graphs could be used in the article itself. I can’t speak about the amount of work this would take, but TR already has a dynamic price link on the front of the articles (or used to).

            • Firestarter
            • 8 years ago

            Well the amount of work implied in doing it all yourself is pretty huge. But, if they could arrange a deal with a pricewatcher, they could pull the relevant data from them. Depending on the quality of the data, it could be a pretty straightforward from there on out.

          • Bensam123
          • 8 years ago

          Yeah that’s what I was thinking mad as TR links to pricegrabber for relevant prices at the front of the article (or used to). I don’t know enough about website development to know how much work is involved with this though so I didn’t pursue it further. :l

          I do think it would add an awesome sort of ‘stock ticker’ look to the site. Especially if the graphs are use on the front page somewhere. You can make graphs look pretty sweet if you take the effort to do so.

    • indeego
    • 8 years ago

    The 256GB drive retails for $380. You could RAID0 them (Hey it’s OCZ, people buy this brand to live on the edge!) for $140 less than one 512G drive.

      • eitje
      • 8 years ago

      TRIM?

        • derFunkenstein
        • 8 years ago

        Intel says it’s coming to the rapid storage drivers.

    • ew
    • 8 years ago

    My guess is the 128GB and 256GB drives are not competitive against other drivers in their price range.

      • HisDivineOrder
      • 8 years ago

      My guess is the 128GB and 256GB drives are not competitive price-wise against other drives in their performance class.

    • phez
    • 8 years ago

    [quote<]We're also going to be keeping our ears to the ground to see if end users have more luck with the Octane than they've had with SSDs affected by the SandForce BSOD bug.[/quote<] I can't imagine many people would rush back to OCZ-camp to test these after experiencing their drive failures ...

    • dpaus
    • 8 years ago

    Love the ‘Value’ scatter plot! But it’d be nicer if you labeled the four rogue Klingons off in their own quadrant….

      • tay
      • 8 years ago

      Haha awesome, had to go back and check

        • derFunkenstein
        • 8 years ago

        So did I. Pretty amusing.

      • cygnus1
      • 8 years ago

      Damn Klingons!

      • stabgotham
      • 8 years ago

      Epic!

Pin It on Pinterest

Share This