Toshiba’s OCZ RD400 512GB SSD reviewed

Back when Samsung’s 950 Pro SSD debuted last year, it marked a turning point in the ongoing storage revolution. While Intel’s 750 Series SSDs paved the way for next-gen SSDs on the desktop, the 950 Pro made the M.2 2280 form factor popular for retail storage devices. It didn’t hurt that Intel’s Skylake platform brought an abundance of of PCIe lanes to the masses, and the NVMe protocol promised a major advance in performance versus the older AHCI, too. Ever since the 950 Pro hit the shelves, though, the market landscape for modern, high-end consumer SSDs really hasn’t changed all that much. The winds of change may finally be upon us, though, as Toshiba is joining the NVMe fray.

Feast your eyes on Toshiba’s OCZ RD400, née the OCZ RevoDrive 400. We first heard about this drive at IDF back in August, and we’ve been excited ever since at the prospect of a true competitor to Intel’s 750 Series SSDs and Samsung’s 950 Pro. Make no mistake, this is a full-fat PCIe 3.0 x4, NVMe, M.2 2280 drive. It comes in a few capacities, each of which can be optionally bundled with a PCIe-to-M.2 adapter.

Toshiba OCZ RD400
Capacity Max sequential (MB/s) Max random (IOps)
Read Write Read Write
128GB 2200 620 170k 110k
256GB 2600 1150 210k 140k
512GB 2600 1600 190k 120k
1TB 2600 1550 210k 130k

So what’s with the RD400 name? Recently, Toshiba rebranded many of its OCZ consumer products under the Toshiba OCZ umbrella. The Trion 150 became the TR150, the Vector 180 became the VT180, and the as-yet-unreleased RevoDrive 400 became the RD400. Toshiba has done some great work rescuing the OCZ brand from its association with unreliability and bankruptcy, so it makes sense that they’re keeping the name around. Under the Toshiba umbrella, OCZ released some great products—the Arc 100 and Trion 150 even earned TR Recommended awards.

The marketing materials and spec sheets for the RD400 are a bit light on technical details, so all we know officially is that the drive packs a Toshiba controller and Toshiba MLC NAND. An unfortunate side effect of the Toshiba acquisition has been a steadily increasing unwillingness on OCZ’s part to discuss the guts of its products, but we’ll forgive the company its reticence if its drives continue to impress.

We can infer that the MLC flash in the RD400 is fabbed on Toshiba’s 15-nm process, as 15-nm NAND is all the company has been shipping recently. And you can be sure that if the company had something newer, smaller, or faster, it would have been a prominent bullet point on the drive’s feature list. Inscribed on the controller chip are letters and numbers starting with “TC58.” TC58 was how Toshiba referred to the controller inside the Trion 100 and 150-series drives. too. We can guess that Toshiba took that controller as a baseline and added some NVMe support into it before slapping it on the RD400.

When it comes to endurance and warranty information, though, Toshiba is happy to share. The 512GB drive is rated for 296 terabytes written, and it’s guaranteed under a five-year advance-replacement warranty program. Toshiba’s suggested price for the 512GB unit is $309.99 for the bare drive or $329.99 with the bundled PCIe adapter card, and both Amazon and Newegg seem to be respecting those prices. Those tags position the RD400 just a tad below the price of a bare Samsung 950 Pro 512GB drive, but you’ll have to live without the Samsung drive’s encryption features if you want to reap those savings.

Before we begin our testing, it’s worth noting that we received an RD400 with Toshiba’s own adapter card. This card comes with a thermal pad that bonds the underside of the drive to the body of the adapter card. We carried out our testing using OCZ’s card, but we also double-checked our results with the Asus adapter card we usually test M.2 SSDs with to make sure the thermal pad wasn’t giving the RD400 any unfair advantages. The drive’s performance doesn’t change appreciably whether it’s mated to a thermal pad or to a regular adapter card, so we went ahead and used OCZ’s adapter for the RD400 for our tests.

On to testing. Let’s see if the first OCZ-branded NVMe drive has what it takes to roll with the heavyweights of the industry.

 

IOMeter — Sequential and random performance

IOMeter fuels much of our latest storage test suite, including our sequential and random I/O tests. These tests are run across the full capacity of the drive at two queue depths. The QD1 tests simulate a single thread, while the QD4 results emulate a more demanding desktop workload. For perspective, 87% of the requests in our old DriveBench 2.0 trace of real-world desktop activity have a queue depth of four or less. Clicking the buttons below the graphs switches between results charted at the different queue depths.

Our sequential tests use a relatively large 128KB block size.



The RD400 instantly stakes its claim as part of the new breed, posting the kind of four-digit speeds we haven’t seen since we reviewed Samsung’s 950 Pro. The sequential read rates at QD4 are a hair’s breadth away from 2000 MB/s, a barrier no drive we’ve tested yet has broken. The RD400’s sequential write speeds are no less impressive, beating out Samsung’s V-NAND-equipped 950 Pro by at least a 30% margin at both queue depths.



The RD400’s random response times don’t blow us away in the same way its sequential speeds do, but they’re still quite respectable. Intel’s 750 Series drive fares better in our tests, but the version in our result dataset is the 1.2TB drive, which has a natural advantage over the RD400 512GB thanks to its extra capacity.

 

Sustained and scaling I/O rates

Our sustained IOMeter test hammers drives with 4KB random writes for 30 minutes straight. It uses a queue depth of 32, a setting which should result in higher speeds that saturate each drive’s overprovisioned area more quickly. This lengthy—and heavy—workload isn’t indicative of typical PC use, but it provides a sense of how the drives react when they’re pushed to the brink.

We’re reporting IOps rather than response times for these tests. Click the buttons below the graph to switch between SSDs.


To show the data in a slightly different light, we’ve graphed the peak random-write rate and the average, steady-state speed over the last minute of the test.

The RD400’s top random speed is rather lackluster, falling into the bottom half of our results. Even Crucial’s budget BX100 500GB posts a higher peak rate. On the other hand, the RD400’s steady-state rate is very good, almost doubling that of the 950 Pro. The 750 Series drive remains as untouchable as ever, doubling the speed of the RD400.

Our final IOMeter test examines performance scaling across a broad range of queue depths. We ramp all the way up to a queue depth of 128. Don’t expect AHCI-based drives to scale past 32, though—that’s the maximum depth of their native command queues.

For this test, we use a database access pattern comprising 66% reads and 33% writes, all of which are random. The test runs after 30 minutes of continuous random writes that put the drives in a simulated used state. Click the buttons below the graph to switch between the different drives. And note that the P3700 plot uses a much larger scale.


The RD400 scales well ’til QD8, after which its curve is more or less flat. That’s a better showing than most SATA drives, but it leaves something to be desired when pitched against Intel’s lineup. Just take a look at the next set of graphs.


The P3700 is miles ahead of the other drives, but we’d expect nothing less from this incredibly expensive datacenter drive. Still, even Intel’s consumer 750 Series SSD exhibits much better scaling capabilities than the RD400. The Toshiba drive does manage to pull ahead of the 950 Pro, but not by a particularly significant margin.

Next, we dispense with IOMeter traces in favor of honest-to-goodness file I/O.

 

TR RoboBench — Real-world transfers

RoboBench trades synthetic tests with random data for real-world transfers with a range of file types. Developed by our in-house coder, Bruno “morphine” Ferreira, this benchmark relies on the multi-threaded robocopy command build into Windows. We copy files to and from a wicked-fast RAM disk to measure read and write performance. We also cut the RAM disk out of the loop for a copy test that transfers the files to a different location on the SSD.

Robocopy uses eight threads by default, and we’ve also run it with a single thread. Our results are split between two file sets, whose vital statistics are detailed below. The compressibility percentage is based on the size of the file set after it’s been crunched by 7-Zip.

  Number of files Average file size Total size Compressibility
Media 459 21.4MB 9.58GB 0.8%
Work 84,652 48.0KB 3.87GB 59%

The media set is made up of large movie files, high-bitrate MP3s, and 18-megapixel RAW and JPG images. There are only a few hundred files in total, and the data set isn’t amenable to compression. The work set comprises loads of TR files, including documents, spreadsheets, and web-optimized images. It also includes a stack of programming-related files associated with our old Mozilla compiling test and the Visual Studio test on the next page. The average file size is measured in kilobytes rather than megabytes, and the files are mostly compressible.

RoboBench’s write and copy tests run after the drives have been put into a simulated used state with 30 minutes of 4KB random writes. The pre-conditioning process is scripted, as is the rest of the test, ensuring that drives have the same amount of time to recover.

Let’s take a look at the media set first. The buttons switch between read, write, and copy results.



The RD400 lives up to the promising performance it delivered in our sequential IOMeter batches. It slaughters our media set, posting greater-than-1000 MB/s speeds in both our read and write tests. Admittedly, its write speeds are the “worst” of our consumer NVMe trio, but they’re still far enough ahead of any of the SATA drives that we’re not too upset about it.

Next up, let’s see how the drive does with our work set.



OCZ’s fledgling NVMe drive aces our work set as far as read speeds are concerned, but its write speeds miss the mark. The work set write test is certainly the toughest hurdle among our “real-world” tests, typically yielding a very small absolute difference between the ritziest and dinkiest drives we bench. Nonetheless, the fact that the RD400 is handily beaten by OCZ’s own TLC-based Trion 150 in the 8T write test is cause for consternation.

That result aside, we’re still largely dazzled by the RD400’s numbers in our file I/O tests. Let’s see what it can do as a primary boot drive before we start delivering overall judgements.

 

Boot times

Until now, all of our tests have been conducted with the SSDs connected as secondary storage. This next batch uses them as system drives.

We’ll start with boot times measured two ways. The bare test depicts the time between hitting the power button and reaching the Windows desktop, while the loaded test adds the time needed to load four applications—Avidemux, LibreOffice, GIMP, and Visual Studio Express—automatically from the startup folder. Our old boot tests focused on the time required to load the OS, but these new ones cover the entire process, including drive initialization.

In these tests, the RD400 far more closely resembles the 950 Pro than the 750 Series. This is without question a Good Thing, since the 750’s boot times have long been its Achilles’ heel (despite Intel targeting that weakness specifically with firmware updates).

Load times

Next, we’ll tackle load times with two sets of tests. The first group focuses on the time required to load larger files in a collection of desktop applications. We open a 790MB 4K video in Avidemux, a 30MB spreadsheet in LibreOffice, and a 523MB image file in the GIMP. In the Visual Studio Express test, we open a 159MB project containing source code for the LLVM toolchain. Thanks to Rui Figueira for providing the project code.

The RD400 performs admirably here. It doesn’t blow our socks off, but neither does any other SSD, given how closely our real-world tests tend to cluster drives. If your already-SSD-equipped work computer is keeping you up at night because of how slowly it spins up productivity applications, the RD400 is not the answer to your prayers. Let’s see what it does with games.

The RD400 sets a record for how quickly it gets you into Middle-Earth, but it’s still only beating the rest of the field by a handful of percentage points. It may not be a cost-effective game library drive, but it does the job well.

We’re all out of performance tests. Flip the page for a breakdown of our test methods.

 

Test notes and methods

Here are the essential details for all the drives we tested:

  Interface Flash controller NAND
Crucial BX100 500GB SATA 6Gbps Silicon Motion SM2246EN 16-nm Micron MLC
Crucial BX200 480GB SATA 6Gbps Silicon Motion SM2256 16-nm Micron TLC
Crucial MX200 500GB SATA 6Gbps Marvell 88SS9189 16-nm Micron MLC
Intel X25-M G2 160GB SATA 3Gbps Intel PC29AS21BA0 34-nm Intel MLC
Intel 335 Series 240GB SATA 6Gbps SandForce SF-2281 20-nm Intel MLC
Intel 730 Series 480GB SATA 6Gbps Intel PC29AS21CA0 20-nm Intel MLC
Intel 750 Series 1.2TB PCIe Gen3 x4 Intel CH29AE41AB0 20-nm Intel MLC
Intel DC P3700 800GB PCIe Gen3 x4 Intel CH29AE41AB0 20-nm Intel MLC
Mushkin Reactor 1TB SATA 6Gbps Silicon Motion SM2246EN 16-nm Micron MLC
OCZ Arc 100 240GB SATA 6Gbps Indilinx Barefoot 3 M10 A19-nm Toshiba MLC
OCZ Trion 100 480GB SATA 6Gbps Toshiba TC58 A19-nm Toshiba TLC
OCZ Trion 150 480GB SATA 6Gbps Toshiba TC58 15-nm Toshiba TLC
OCZ Vector 180 240GB SATA 6Gbps Indilinx Barefoot 3 M10 A19-nm Toshiba MLC
OCZ Vector 180 960GB SATA 6Gbps Indilinx Barefoot 3 M10 A19-nm Toshiba MLC
Plextor M6e 256GB PCIe Gen2 x2 Marvell 88SS9183 19-nm Toshiba MLC
Samsung 850 EV0 250GB SATA 6Gbps Samsung MGX 32-layer Samsung TLC
Samsung 850 EV0 1TB SATA 6Gbps Samsung MEX 32-layer Samsung TLC
Samsung 850 Pro 500GB SATA 6Gbps Samsung MEX 32-layer Samsung MLC
Samsung 950 Pro 512GB PCIe Gen3 x4 Samsung UBX 32-layer Samsung MLC
Samsung SM951 512GB PCIe Gen3 x4 Samsung S4LN058A01X01 16-nm Samsung MLC
Samsung XP941 256GB PCIe Gen2 x4 Samsung S4LN053X01 19-nm Samsung MLC
Toshiba OCZ RD400 512GB PCIe Gen3 x4 Toshiba 15-nm Toshiba MLC
Transcend SSD370 256GB SATA 6Gpbs Transcend TS6500 Micron or SanDisk MLC
Transcend SSD370 1TB SATA 6Gpbs Transcend TS6500 Micron or SanDisk MLC

All the SATA SSDs were connected to the motherboard’s Z97 chipset. The M6e was connected to the Z97 via the motherboard’s M.2 slot, which is how we’d expect most folks to run that drive. Since the XP941 and 950 Pro requires more lanes, they were connected to the CPU via a PCIe adapter card. The 750 Series and DC P3700 were hooked up to the CPU via the same full-sized PCIe slot.

We used the following system for testing:

Processor Intel Core i5-4690K 3.5GHz
Motherboard Asus Z97-Pro
Firmware 2601
Platform hub Intel Z97
Platform drivers Chipset: 10.0.0.13

RST: 13.2.4.1000

Memory size 16GB (2 DIMMs)
Memory type Adata XPG V3 DDR3 at 1600 MT/s
Memory timings 11-11-11-28-1T
Audio Realtek ALC1150 with 6.0.1.7344 drivers
System drive Corsair Force LS 240GB with S8FM07.9 firmware
Storage Crucial BX100 500GB with MU01 firmware

Crucial BX200 480GB with MU01.4 firmware

Crucial MX200 500GB with MU01 firmware

Intel 335 Series 240GB with 335u firmware

Intel 730 Series 480GB with L2010400 firmware

Intel 750 Series 1.2GB with 8EV10171 firmware

Intel DC P3700 800GB with 8DV10043 firmware

Intel X25-M G2 160GB with 8820 firmware

Plextor M6e 256GB with 1.04 firmware

OCZ Trion 100 480GB with 11.2 firmware

OCZ Trion 150 480GB with 12.2 firmware

OCZ Vector 180 240GB with 1.0 firmware

OCZ Vector 180 960GB with 1.0 firmware

Samsung 850 EVO 250GB with EMT01B6Q firmware

Samsung 850 EVO 1TB with EMT01B6Q firmware

Samsung 850 Pro 500GB with EMXM01B6Q firmware

Samsung 950 Pro 512GB with 1B0QBXX7 firmware

Samsung XP941 256GB with UXM6501Q firmware

Transcend SSD370 256GB with O0918B firmware

Transcend SSD370 1TB with O0919A firmware

Power supply Corsair AX650 650W
Case Fractal Design Define R5
Operating system Windows 8.1 Pro x64

Thanks to Asus for providing the systems’ motherboards, to Intel for the CPUs, to Adata for the memory, to Fractal Design for the cases, and to Corsair for the system drives and PSUs. And thanks to the drive makers for supplying the rest of the SSDs.

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 every component of our test suite. For the IOMeter database, RoboBench write, and RoboBench copy tests, the drives were put in a simulated used state that better exposes long-term performance characteristics. Those tests are all scripted, ensuring an even playing field that gives the drives the same amount of time to recover from the initial used state.

  • We run virtually all our tests three times and report the median of the results. Our sustained IOMeter test is run a second time to verify the results of the first test and additional times only if necessary. The sustained test runs for 30 minutes continuously, so it already samples performance over a long period.

  • Steps have been taken to ensure the CPU’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 frequency at 3.5GHz. Transitioning between power states can affect the performance of storage benchmarks, especially when dealing with short burst transfers.

The test systems’ Windows desktop was set at 1920×1080 at 60Hz. 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.

 

Conclusions

Before we deliver a verdict, let’s reduce the RD400’s performance characteristics to a single overall rating. We distill the overall performance rating using an older SATA SSD as a baseline. To compare each drive, we then take the geometric mean of a basket of results from our test suite. Only drives which have been through the entire current test suite on our current rig are represented.

Very impressive. Toshiba’s NVMe drive slots in ever so slightly higher than Samsung’s 950 Pro. The RD400’s incredible sequential speeds are likely what gives it the edge versus the 950 Pro. Intel’s 750 Series drive is still the one to beat, but it typically commands a much higher price than its competitors.

So let’s see where the RD400 falls in terms of performance per dollar. In the scatter plot below, the most compelling position is toward the upper left corner, where the price per gigabyte is low and performance is high.

The RD400 lands just a tiny bit higher and just a tiny bit left of the 950 Pro 512GB. Shrewd pricing by the bean counters at Toshiba, to be sure. Intel’s 750 Series may still wear the NVMe performance crown, but it costs almost $0.90 per gigabyte. Both the RD400 and 950 Pro sell at prices closer to $0.60 per gigabyte, a much more palatable figure. The Intel drive requires the buyer to dedicate a full PCIe slot to storage, too, unlike the M.2 2280 convenience offered by both OCZ and Samsung’s products.

Toshiba’s OCZ RD400 512GB wowed us with its hitherto-unseen sequential read rates, and it’s about as compelling a package as Samsung’s 950 Pro SSD. That’s no mean feat, and frankly, that kind of competition is just what the NVMe SSD market needs. If you’ve been saving your hard-earned cash for the 950 Pro, the RD400 deserves a long look, as well.

More than anything, we’re hoping the RD400 heralds a slew of similar drives from other manufacturers. More competition in this space could help to democratize NVMe storage for those of us with shallower pockets. In the meantime, though, Toshiba’s ongoing efforts to resurrect the OCZ brand have reached a spectacular pinnacle with the RD400.

Comments closed
    • tipoo
    • 3 years ago

    Since NVMe supports up to 64K queues with 64K commands each, would NVMe SSDs show any scaling beyond what AHCI SSDs are usually tested at, at just queue depths of 32-128?

    • SixIron
    • 3 years ago

    While I applaud Toshiba making it price competitive with the 950 Pro, I would remind them that it still bears the OCZ name, which still carries an enormous amount of negative baggage for many enthusiasts, from their SSD’s in the early days

    • atmartens
    • 3 years ago

    [dupe]

    • atmartens
    • 3 years ago

    [quote<]Intel's 750 Series may still wear the NVMe performance crown, but it costs almost $0.90 per gigabyte. Both the RD400 and 950 Pro sell at prices closer to $0.60 per gigabyte, a much more palatable figure.[/quote<] I paid $160 for an 80 gb Intel SSD years ago ($2 / GB). These prices are cheap!

      • Srsly_Bro
      • 3 years ago

      And that was when $2 GB wasn’t a bad price. I have a Crucial C300 64 GB is a small notebook.

        • DarkMikaru
        • 3 years ago

        Ha.. haven’t heard that mentioned in a while. My old C300 64GB is in my old laptop that I gifted to my mom. It’s still kicking.

    • tootercomputer
    • 3 years ago

    I built a new system around the i7 6700k and the z170 chipset/mobo a few months ago. I was not very knowledgeable about m.2 drives at the time (had been out of the “build” loop for a while), and as I put my system together and did some homework on these, I could not resist the 950 pro 256G. Yeah, I paid a premium, but geez, these puppies are the future: small, incredibly fast, no cables, and they connect directly to your mobo (if it’s up-do-date.) As these catch on, I just have to wonder how much chipsets and motherboard will change in design, cases. I can imagine mobos with multiple m.2 slots (even more than the two my gigabyte mob has). And I’m glad to see more choices like the Toshiba/OCZ here; it will drive prices down and push performance like competition always does.

    • mkk
    • 3 years ago

    I keep looking to TR reviews for the real world lowdown. Keep up the good work.

    A bonus with this product is that they have their own PCIe card for those who need/want that.

    • PrincipalSkinner
    • 3 years ago

    So it’s marginally faster where it matters. Why bother?

      • sweatshopking
      • 3 years ago

      cause it’s marginally faster and marginally cheaper

        • PrincipalSkinner
        • 3 years ago

        I meant marginally faster when compared to SATA SSDs in real world scenarios. Why would someone get this instead of cheaper or capacious SATA3 SSD?

    • Chrispy_
    • 3 years ago

    Impressive that it manages to compete with Samsung. That’s no mean feat these days, no matter how little I care for Samsung’s attitude towards their customers. Why so expensive, though?

    It’s not the controller, since that’s a variant of same controller their budget [s<]Trion[/s<] TR range uses. Is it the NAND? If it's the NAND, how soon until an NVMe product appears using normally-priced A19 MLC? I want NVMe but I don't like paying double the cost/GB for the privilege.

      • DPete27
      • 3 years ago

      None of the above. In fact, you’re even saving BOM by losing the enclosure vs SATA SSDs. I’m convinced they’re just pricing NVMe M.2 SSDs so high just because of the performance numbers. Quite the cash cow if you ask me.

      “Hey Intel, thanks for giving us enough IO to uncork our SSDs. Now we can charge double for the same product.”

    • Krogoth
    • 3 years ago

    Nothing new to see here.

    NVMe SSD media is certainly faster on paper but under a non-server and workstation loads. You end-up being CPU-bound (Mhz/IPC is still king) in almost everything.

    • Firestarter
    • 3 years ago

    So, what explains the price difference to SATA SSDs other than the usual R&D and volume arguments? Is the Toshiba MLC NAND that they use on these SSDs different from the stuff on your run of the mill SSD? Because if it isn’t and it actually costs roughly the same to manufacture, then that means they’re charging double pretty much just because they can. I wanted a 1TB NVMe SSD a while ago, but in my opinion the novelty and the pretty substantial performance improvement still don’t warrant a 100% price difference compared to the already very speedy 1TB SATA SSDs out right now.

      • derFunkenstein
      • 3 years ago

      That’s why we need more competitors in this space. They’ll come once Marvell or one of the other usual non-Samsung suspects has an NVMe controller in the wild, I think.

        • albundy
        • 3 years ago

        yup, just a matter of time.

    • derFunkenstein
    • 3 years ago

    Excellent and thorough as always, Tony.

    I keep eyeing a bigger drive than my 250GB BX100 but I have held off waiting for more NVMe drives to show up. I’m guessing that within 6 months it’ll be close enough to SATA M.2 pricing that I’ll bite. I’ve got an x4 M.2 slot on the back of my motherboard just waiting.

    • sweatshopking
    • 3 years ago

    this is the [i<] first [/i<] pci e ssd i'd consider

      • drfish
      • 3 years ago

      That gives me [url=http://goo.gl/CMII81<]an idea[/url<]...

        • anotherengineer
        • 3 years ago

        Solidworks………….the way it’s meant to be used 😉

        • sweatshopking
        • 3 years ago

        I LOVE YOU

      • floodo1
      • 3 years ago

      I’m buying a 1tb because I’ve been waiting for a stupid fast 1tb m2 drive for a year!

Pin It on Pinterest

Share This