Home Samsung’s 960 EVO SSD reviewed
Reviews

Samsung’s 960 EVO SSD reviewed

Tony Thomas
Disclosure
Disclosure
In our content, we occasionally include affiliate links. Should you click on these links, we may earn a commission, though this incurs no additional cost to you. Your use of this website signifies your acceptance of our terms and conditions as well as our privacy policy.

NVMe may still have that new-tech smell to it, but the reality is that SSDs using that protocol have been on the market for a couple of years now. Sadly, those drives have remained out of reach for all but the deepest pockets. The new protocol purported to unlock all of the performance of highly parallel flash storage, and manufacturers weren’t bashful about making the cost of entry similar to that of an on-Broadway showing of Hamilton.

Samsung’s 950 Pro was perhaps the first NVMe drive sold at a merely steep price point rather than a ludicrous one. Its recent follow-up, the 960 Pro, pushes the performance boundary even higher, but it does little to make NVMe more accessible to the masses. Builders can rejoice now, though, because last week Samsung launched the 960 Pro’s little brother, the 960 EVO. In a radical departure from previous EVO drives, the latest drives are M.2 gumsticks with NVMe support. We’ve been graced with 250GB and 1TB samples to test for you, and a 500GB drive rounds out the lineup.

Samsung 960 EVO
Capacity Max sequential (MB/s) Max random (IOps) Price
Read Write Read Write
250GB 3200 1500 330k 300k $129
500GB 3200 1800 330k 330k $249
1TB 3200 1900 380k 360k $479

The 960 EVO shares much of the same underlying technology that powers the 960 Pro. As it did with its top-end 960 SSD, Samsung pairs 3D V-NAND and its Polaris controller. To make the EVO more affordable, that V-NAND uses a TLC configuration this time around rather than MLC, and that move lowers potential performance and endurance. To at least offset the speed loss, Samsung equips these EVO drives with TurboWrite, the company’s proprietary pseudo-SLC caching scheme. This time around, they’re calling it “Intelligent TurboWrite,” and the technology does indeed sound like it’s gotten a little smarter.

Previous iterations of TurboWrite used a fixed slice of storage—usually a meager handful of gigabytes—to act as a high-speed SLC buffer for incoming writes. This “Intelligent” revision uses a similar dedicated SLC portion, but also allows the drive to seize additional buffer space if it’s available. The 960 EVO 250GB has a fixed 4GB TurboWrite cache that can be supplemented with another 9GB of the drive’s capacity. The 1TB has a fixed 6GB cache, and it can commandeer a whopping 36GB more for use as a buffer. Savvy users already avoid filling their SSDs to capacity, so Samsung is shrewdly exploiting this habit to eke out even more performance than the TLC flash could attain on its own. In a worst-case scenario, a 960 EVO filled to the brim will still have its dedicated TurboWrite partition available to accelerate writes.

With their stickers peeled off, the 250GB and 1TB drives are difficult to tell apart. In each drive, the Polaris controller, DRAM cache, and two NAND packages all lie on a single side of the PCB. The EVO drives don’t use the “package-on-package” controller design Samsung previously used in the 960 Pro 2TB, so this time around we can actually see the Polaris chip. Both drives use Samsung’s latest 48-layer, 256Gb V-NAND packages, but the 250GB drive uses a specially-packaged version of that flash to maximize performance in spite of its lower capacity.

Samsung’s price sheet indicates that the 960 EVO 250GB will sell for $129, and the 1TB version will go for $479. If you’re willing to cough up that much dough, you get 256-bit AES hardware encryption and TCG Opal support for your trouble. Samsung warrants the drives for three years, par for the course with the EVO series. The company expects the 250GB drive to last 100 terabytes written, while the 1TB drive should take a 400-terabyte beating.

Now that we’ve met the 960 EVO, it’s time for some action. Samsung dazzled us with the 960 Pro, so we hope the 960 EVO produces a similar impression at a lower cost.

 

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 960 EVO’s sequential read speeds are nothing short of phenomenal. Both capacities break 2000 MB/s at both queue depths, something that not even the 960 Pro achieved. Sequential writes are more of a mixed bag. The 1TB drive puts up a very strong performance at over 1000 MB/s, but the 250GB version only manages to squeeze out about 350 MB/s. That’s an acceptable speed, but unremarkable—there are older, cheaper SATA drives like the Arc 100 240GB and Vector 180 240GB that write good deal faster.



Random response times exhibit largely the same pattern. Read times are snappy, just a hair slower than the 960 Pro. Write latencies are blazingly fast for the 1TB, but far more pedestrian for the 250GB.

Thus far, the 960 EVO 1TB shows great promise. It can’t catch up to the 960 Pro 2TB’s sequential write speeds, but its read speeds are top-notch. The 960 EVO 250GB’s reads just as quickly, but its writes are a bit lackluster. But the good news is that across the board, both drives beat out the 850 EVO of the same 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 that 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.


From crest to trough, the 960 EVO 250GB’s plot doesn’t look all that different from the 850 EVO 250GB’s. Clearly IOMeter write tests are not the drive’s strong suit. The 1TB version, on the other hand, hits an insane peak that even the 960 Pro can’t match. Even after it’s used up all its caching tricks, the 960 EVO 1TB writes at incredible speed. 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.

An insane peak indeed. The 960 EVO 1TB sets a new record for peak random write rate in our sustained test, beating even the datacenter-class DC P3700. The P3700 still holds the crown for steady-state performance, but the EVO 1TB snags second place. The 960 EVO 250GB again manages to beat the 850 EVO 250GB, but the margin is smaller than we might have expected.

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 960 EVO 250GB simply isn’t fast enough in IOMeter to convert its NVMe capabilities into any sort of meaningful scaling. But the 960 1TB does not disappoint, ramping up smoothly to QD32 before starting to taper off. Let’s look at the Samsung lineup together to get a sense of where the 960 EVOs fall.


As in all the prior tests, the 250GB EVO improves upon its predecessor, but only by a small margin. The 1TB EVO’s performance is only contested by the 960 Pro, whose scaling curve is noticeably steeper. All those dollars have to buy you something, after all.

The 960 EVO 1TB passed all of our IOMeter tests with flying colors. The 960 EVO 250GB offered lightning-fast reads, but its writes proved only a little better than the 850 EVO 250GB’s. Perhaps real-world workloads will better showcase the drive’s capabilities.

 

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.



Now we’re cooking with gas! With IOMeter, the 250GB drive only really excelled in read tests, but its write capabilities finally come to life in RoboBench. These result prove that it really does belong in the elite NVMe club. Reads and writes break 1000 MB/s for both drives. The 1TB drive actually claims the RoboBench write record in both single- and eight-threaded tests. Even the speed-demon 960 Pro can’t keep up with the 960 EVO 1TB’s gargantuan TurboWrite SLC cache.

Next up, the work set.



The EVOs do well here, just not as startlingly well as in the media set. The work set has always been a great equalizer, and these drives handle it just a little better than the 850 EVOs did. The 960s manage to maintain a sizable gap in the read test, but the writes are only a tad faster.

The 960 EVO 1TB looks more and more exceptional each time we put it to the test. The 250GB drive fares much better with RoboBench’s write tests than it does with IOMeter’s, but the two versions of the drive are clearly in separate performance classes. Next, we’ll boot Windows off these drives to see how they fare as primary storage.

 

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.

The 960 EVOs are among the faster drives to boot that we’ve tested, but so are the 850 EVOs. All that new hotness doesn’t translate to a faster arrival at the Windows desktop.

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.

As usual, the drives land willy-nilly in the productivity load test rankings. Let’s see if games turn up anything interesting.

Nope, nothing interesting. 960 EVOs work just fine as game library drives, but so do 850 EVOs. Or Trion 150s. Or Intel X25s. Get my drift?

That wraps up our test suite. Click next to read about our test methods, or skip straight ahead to the conclusion.

 

Test notes and methods
Here are the essential details for all the drives we tested:

  Interface Flash controller NAND
Adata Premier SP550 480GB SATA 6Gbps Silicon Motion SM2256 16-nm SK Hynix TLC
Adata Ultimate SU800 512GB SATA 6Gbps Silicon Motion SM2258 32-layer Micron 3D TLC
Adata XPG SX930 240GB SATA 6Gbps JMicron JMF670H 16-nm Micron MLC
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
Crucial MX300 750GB SATA 6Gbps Marvell 88SS1074 32-layer Micron 3D TLC
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 960 EVO 250GB PCIe Gen3 x4 Samsung Polaris 32-layer Samsung TLC
Samsung 960 EVO 1TB PCIe Gen3 x4 Samsung Polaris 48-layer Samsung TLC
Samsung 960 Pro 2TB PCIe Gen3 x4 Samsung Polaris 48-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 TC58 15-nm Toshiba MLC
Toshiba OCZ VX500 512GB SATA 6Gbps Toshiba TC358790XBG 15-nm Toshiba MLC
Transcend SSD370 256GB SATA 6Gbps Transcend TS6500 Micron or SanDisk MLC
Transcend SSD370 1TB SATA 6Gbps 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, 950 Pro, RD400, and 960 Pro require 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
The 960 EVOs put up great numbers across most of our test suite, so we have high expectations for their final standings. The 1TB drive especially threatens to shake up the top end of the overall performance chart. To compare each drive, we take the geometric mean of a basket of results from our test suite. Only drives that have been through the entire current test suite on our current rig are represented.

 

The 960 EVO 1TB clambers all the way to the summit, landing just a step below the 960 Pro 2TB. The drive turned in an outstanding performance, even eclipsing its more expensive older brother in several of the tests. It’s no surprise that it landed so high. The 960 EVO 250GB doesn’t reach such nosebleed-inducing heights, but it still shows a sizeable performance increase over its 850 EVO predecessor. Samsung succeeded in giving the 960 EVO 250GB an NVMe-facilitated performance bump over the 850 EVO, but really hit it out of the park with its 1TB drive. Performance-wise, the RD400, 950 Pro, and 750 Series fall in between the two versions of the 960 EVO, so their prices may need some adjusting to remain competitive.

So the 960 EVOs’ performance is certainly up to snuff, but what about the bang-for-buck? Our scatter plots will give us the answer. Use the buttons to switch between views of all drives, only SATA drives, or only PCIe drives. The most compelling position in these plots is toward the upper left corner, where the price per gigabyte is low and performance is high. Prices for the 960 EVOs and 960 Pro 2TB are based on Samsung’s price sheets, since they aren’t actually for sale yet.


The 960 EVO 1TB is in a great spot, perhaps even the sweet spot. It represents a huge savings over the 960 Pro while reaching almost the same level of performance. The 960 EVO 250GB isn’t quite as enticing, but it delivers near-950 Pro 512GB performance for less money. That’s still a pretty amazing leap for the solid-state storage market.

This tectonic shift in the price-to-performance arena leaves Samsung’s own 950 Pro and Intel’s 750 Series drives looking outclassed. Toshiba isn’t standing still, though. Its OCZ RD400 512GB MLC drive has recently gotten some price cuts that let it hang with Samsung’s latest. Despite debuting at near-950 Pro prices, the RD400 512GB is down to $269 on Newegg or even less on Amazon for the version with an included PCIe adapter. Competition is a great thing.

Samsung 960 EVO 1TB
November 2016

In summary, we like the 960 EVO 250GB, but we like like the 960 EVO 1TB. We previously gave the 960 Pro 2TB our Editor’s Choice award, but the 960 EVO 1TB deserves it even more. 48 cents per gigabyte is substantially more than many are used to paying for storage these days, but we think the performance increase over SATA drives finally justifies the expense for the well-heeled. It’s more affordable than the 63 cents per gigabyte that the 960 Pro demands while delivering close to the same performance.

Intelligent TurboWrite, copper-film heat dissipation, third-generation V-NAND, and NVMe combine to give the 960 EVO 1TB a huge performance edge over almost any other drive on the market. It’s part of the formation of a new intermediate tier that the market has sorely needed in its transition to NVMe. The only real reason anyone should go for the 960 Pro over the EVO is for its five-year warranty and increased endurance. For sheer performance per dollar, the 960 EVO 1TB is the new king.

Latest News

Dish Network
Streaming News & Events

Dish Network Reaches Settlement Deal in Lawsuit Against iFit

Japan
News

Japan Entertains More Taiwan Chip Companies as China Decoupling Grows

A recent report by Reuters suggests that Japan’s continuous push to revive its semiconductor industry has increased the expansion of more Taiwanese chip companies on its shores.  These companies are...

AI Chatbot Costs Air Canada Hundreds Of Dollars In Compensation
News

Chatbot’s Mistake Costs Air Canada Hundreds Of Dollars In Compensation

Air Canada’s AI chatbot misrepresented the company’s discount policy to certain customers, as a result of which they are now forced to pay a hefty compensation to those affected. The...

Leaked Data Shows China Cyber Firm Hacked Govt Agencies
News

Leaked Data Shows China Cyber Firm Allegedly Hacked Govt Agencies, Businesses, Organizations

“Gmail Is Shutting Down” — Google Shuts Down Rumors
News

“Gmail Is Shutting Down” — Google Shuts Down Rumors 

Google CEO Sundar Pichai Says That AI Can Help Against Cybercrime
News

Google CEO Sundar Pichai Says That AI Can Help Against Cybercrime

Bitcoin’s
Crypto News

First Bitcoin Zero-Knowledge Rollup Citrea Secures $2.7M In Seed Funding