Introduction & Specifications, Pricing, and Availability
These days, we rarely get to discuss a SSD that uses all Single-Level Cell (SLC) flash. Even on the enterprise side, SLC is the golden egg that only shows up from time to time, and only in products designed for extreme workloads. Multi-Level Cell (MLC) flash has ruled the roost for several years, and projections show that Triple-Level Cell (TLC) will soon rapidly gain market share from MLC.
SLC flash only writes half as much of the data to the flash as MLC. With SLC, things are simplified, and the data will only be a one or a zero. In terms of capacity, that means SLC needs twice as many NAND die per drive than a MLC part. Since the base cost comes from a wafer that is cut into flash die, and is then packaged for use on a SSD, a SLC SSD will have to cost at least twice as much as a MLC SSD.
When we first tested a modern SLC SSD, the price was a larger issue since the density was limited to 128GB at a time when 256GB SSDs were popular. At the time, the 128GB SLC drive cost more than a 256GB SSD around $275. Still, those lucky enough to purchase the SuperSSpeed S301 based drives were rewarded with lasting performance that didn't slow over time, and industry leading endurance.
In terms of endurance, SLC lasts for around 100K program / erase cycles. To put that into perspective, we went on and on about Samsung's new 3D V-NAND technology, and how backroom conversations led us to believe the new technology had around 35K PE cycles. Typical MLC flash shipping today is rated for 3K PE cycles, but that can range from 1.5K to 10K depending on the product. TLC flash ranges from 800 PE cycles to 1.5K, so the leap from MLC and TLC standards to SLC standards is significant.
SLC flash also needs less time to read and write data than MLC or TLC flash. Caching techniques from Samsung, SanDisk, and most recently, Micron, allow a drive to write small bursts to the flash as it would to SLC by manipulating the flash translation layer. This is an excellent method to increase write performance for a short period of time. For most consumers, the short burst is more than enough to increase performance for day-to-day activities. Although, performance slows when a user has written to the SLC area, and the data writes are then forced to the MLC, or worse, the TLC flash at their native speeds. Heavy workload environments that push data at high speeds to a SSD, like editing a large video, can fill the SLC buffer, and then slow by as much as two-thirds when completing the task.
The way around this for professional environments is to simply move to all SLC flash, rather than just a small amount of SLC like flash in a product.
Specifications, Pricing, and Availability
Fujitsu is a world leader in enterprise storage technology, and entered the consumer SSD market around June 2014. We've already published a report of the F100, a mainstream consumer focused MLC SSD that uses controller technology from Silicon Motion. We also took an early look at the Quadro-M, an All-in-One RAID PCIe product.
Today we're looking at the Fujitsu FSXtreme 240GB, which uses a SandForce SF-2281 controller, and 25nm Intel SLC NAND flash. The flash in the FSXtreme is a bit different from the SLC flash used in the SuperSSpeed products from a few years ago. The flash now features 16KB page sizes; the S301 used 8K page sizes. The new 25nm Intel SLC allows Fujitsu to package up to 256GB of raw SLC flash in a single SSD.
The FSXtreme is available in three capacities, 60GB, 120GB, and 240GB; these are all overprovisioned sizes so 64GB, 128GB, and 256GB flash are actually present, but hidden from the user. The reserved OP area handles background activities like garbage collection and wear leveling, even when the drive's user capacity is full.
The FSXtreme uses a SandForce SF-2281 controller for all three capacities. In our tests with the SuperSSpeed S301 128GB and 120GB models, we observed similar performance with compressible and incompressible data. The same carries over to the Fujitsu FSXtreme. The SLC flash virtually removes the number one drawback of SF-2281 products, although, the incompressible performance slowdown is often only observed while running benchmark software.
Fujitsu attached a full five-year warranty to the FSXtreme products. In the U.S., purchasing a drive isn't as easy as ordering from Amazon or Newegg; this is an area Fujitsu is working on, and we may see Fujitsu branded products on Amazon shortly. At this time, Fujitsu has led us to one of the company's partners who accepts international orders, and provides quick shipping. You can purchase the Fujitsu FSXtreme 240GB for $228.80 USD from this link, and expect delivery within a few days. The FSXtreme 120GB model sells for $130.79 USD.
Fujitsu FSXtreme 240GB SLC SSD
This is the first Fujitsu SSD we've tested that shipped with a full retail kit. The package lets you know that this is product packs SLC flash.
The back of the package lists the product's specifications, and explains the difference between SLC, MLC, and TLC flash.
Inside the package, you receive a warranty statement that doubles as a user guide, a contact for support, and the FSXtreme SSD.
Here we get our first look at the Fujitsu SXtreme Series 240GB SSD. Everywhere you look Fujitsu lets you know this drive packs SLC flash, but doesn't throw it in your face.
The model and serial number are listed on the back of the drive. Our drive was manufactured in September of 2014.
The FSXtreme uses a 7mm z-height case design, so it will fit in your Ultrabook or notebook that requires thinner design.
We didn't find a thermal pad inside of the case, but the SandForce SF-2281 controller rarely gets hot, even under high load.
The controller employed by Fujitsu is the VB1 version of the SF-2281. This is a newer stepping model that reduces power and thermal output.
The special component is the Single-Level Cell (SLC) flash. Fujitsu uses Intel flash in all of the company's consumer SSDs, but these are the rare super die that reduce program latency, and turn up the performance.
Test System Setup and Initial Performance
Desktop Test System
Lenovo T440 - Notebook Power Testing with DEVSLP and Windows 8.1 Pro
Nearly all of the performance tests run on the desktop system, but we use a Lenovo T440 to run the power tests. The T440 is the latest addition to our client SSD test lab, and allows us to test the notebook battery life offered by a SSD with advanced features like DEVSLP enabled.
Initial Performance Evaluation - 4-Corner and then Some Tests
Sequential 80% Read 20% Write
Random 80% Read 20% Write
Our preliminary tests show strong performance in all areas except sequential write performance, which reaches 400 MB/s when the drive has already written more data than it's capacity. We call that preconditioning. The test writes so much data that the drive has to perform garbage collection and wear leveling while we finish the test. Some of the performance is lower than the claimed performance, and the preconditioning involved is the reason why.
Benchmarks - Sequential Performance
HD Tune Pro - Sequential Performance
Version and / or Patch Used: 5.50
The average sequential read performance with 64KB blocks is just over 410 MB/s. The maximum and minimum speeds remain fairly constant at 420 and 410 MB/s. SLC flash allows the SSD to remain fairly consistent when reading and writing data.
As you can see on this chart, most of the modern 256GB class SSDs vary in write performance, but a few still deliver performance with little variability; the Fujitsu FSXtreme 240GB is in that latter group.
HD Tach - Sequential Write Performance after Random Writes
Version and / or Patch Used: 18.104.22.168
After preconditioning the drive with a reasonable number of sequential and random data writes, we test the drive with HD Tach to observe the 128K sequential read and write performance. There is a little variability, but not nearly as much as we often see with MLC and TLC based SSDs.
Benchmarks - Anvil Storage Utilities
Anvil Storage Utilities
Version and / or Patch Used: RC6
So, what is Anvil Storage Utilities? Anvil Storage Utilities, is a storage benchmark for SSDs and HDDs where you can check and monitor your performance. The Standard Storage Benchmark performs a series of tests; you can run a full test, or just the read or the write test, or you can run a single test, i.e. 4k QD16.
Anvil Storage Utilities is not officially available yet, but we've been playing with the beta for several months now. The author, Anvil, has been updating the software steadily on several international forums, and is adding new features every couple of months.
We can use Anvil several different ways to show different aspects for each drive. We have chosen this software to show the performance of a drive with two different data sets. The first is with compressible data, and the second data set is incompressible data. Several users have requested this data in our SSD reviews.
0-Fill Compressible Data
SLC flash tames the SF-2281 controller's incompressible data slump. The flash is so fast that it keeps the incompressible performance nearly as high as the compressible data performance when writing data. This throws the only real negative of using SandForce controllers out the window.
Low Queue Depth Read IOPS
High Queue Depth Read IOPS
The Fujitsu FSXtreme 240GB delivers the highest QD1 4k random read performance we've measured on a single SATA SSD. We often talk about the importance of a product delivering 10K random read IOPS at QD1, but the FSXtreme achieves nearly 13K. This level of performance allows your applications to open in record time.
At high queue depths, the FSXtreme levels off at 75K IOPS; however, it's very difficult to get to the high IOPS range with a consumer SSD because the drives are so fast, and consumer workloads are not that demanding.
Low Queue Depth Write IOPS
High Queue Depth Write IOPS
The FSXtreme 240GB shows very good low queue depth random write performance that scales very well to high queue ranges, and tops out at over 90K IOPS.
Benchmarks - Mixed Read / Write Workloads
Sequential Mixed Read / Write Workloads
In this series of tests, we measure mixed workload performance. We start with 100% read, and then add data writes to the mix in 10% increments until we get to 100% writes. We believe this will be the next major area SSD manufacturers will address, after performance consistency.
Sequential Mixed Workload Bandwidth
SATA can't read and write data at the same time; it's half-duplex, so the controller has to be able to read, then write data quickly. Our tests show that SandForce SF-2281 controllers are the best at this transition, but paired with SLC flash, the SF-2281 takes it to another level. There isn't a lot of drop off at all; it just takes the workload and runs with it.
Sequential 80% Read / 20% Write Bandwidth
The test uses 50% entropy. Entropy is just a fancy name for incompressible data. Since the SLC flash negates the compressible / incompressible roll off, and the controller already does well with mixed workload (read and write) testing, the FSXtreme killed this test.
If you manipulate a lot of sequential data, like reading and writing audio / video files when editing, the FSXtreme 240GB is the fastest SATA SSD money can buy for this purpose.
4K Random Mixed IO
The FSXtreme is optimized for QD8 and QD16 random mixed workloads. Again, this falls inside the A/V editing realm where you have the main file manipulation, the reading and writing from plug-ins, and cache.
PCMark 8 Consistency Test
Futuremark PCMark 8 Extended - Consistency Test
Version and / or Patch Used: 2.0.228
Heavy Usage Model:
Futuremark's PCMark 8 allows us to wear the test drive down to a reasonable consumer steady state, and then watch the drive recover on its own through garbage collection. To do that, the drive gets pushed down to steady state with random writes, and then idle time between a number of tests allows the drive to recover.
1. Write to the drive sequentially through up to the reported capacity with random data.
2. Write the drive through a second time (to take care of overprovisioning).
1. Run writes of random size between 8*512 and 2048*512 bytes on random offsets for ten minutes.
2. Run performance test (one pass only).
3. Repeat one and two, eight times, and on each pass, increase the duration of random writes by five minutes.
Steady state Phase:
1. Run writes of random size between 8*512 and 2048*512 bytes on random offsets for 50 minutes.
2. Run performance test (one pass only).
3. Repeat one and two, five times.
1. Idle for five minutes.
2. Run performance test (one pass only).
3. Repeat one and two, five times.
PCMark 8's Consistency test provides a ton of data output that we use to judge a drive's performance. Here we see the three states of performance for select SSDs: light use, consumer steady state, and worst case.
Storage Bandwidth - All Tests
Storage Bandwidth - Heavy Load
The heavy workload test shows that the Fujitsu FSXtreme 240GB is the second fastest 256GB class SSD on the market today under heavy load. The only drive faster is the SanDisk Extreme PRO 240GB.
Storage Bandwidth - Typical Consumer Load
Under typical consumer workloads, the FSXtreme 240GB remains very fast, but a couple of lower price value drives manage to perform a little faster thanks to advanced cache schemes.
PCMark 8 Consistency Test - Continued
Total Access Time - All Tests
The access time test measures the total latency across all 18 tests. This is one of, if not the most important of the tests we run at this time for consumer SSDs. When your latency is low, your computer feels fast, it's just that simple.
Total Access Time - Heavy Load
Total Access Time - Typical Consumer Load
Of all the tests in this review, I feel the latency results from PCMark 8's advanced storage test is the most relevant to consumers. The FSXtreme 240GB stays competitive throughout the tests, but again falls short of hitting the standard set by the Extreme PRO.
Benchmarks - Power Testing and NAS Cache
Bapco MobileMark 2012 1.5
Version and / or Patch Used: 2012 1.5
Developer Homepage: http://www.bapco.com
Test Homepage: http://www.bapco.com
MobileMark 2012 1.5 is an application-based benchmark that reflects usage patterns of business users in the areas of office productivity, media creation, and media consumption. Unlike benchmarks that only measure battery life, MobileMark 2012 measures battery life and performance simultaneously, showing how well a system design addresses the inherent tradeoffs between performance and power management.
There are consumer 256GB class SSDs that give a notebook better battery life than the Fujitsu FSXtreme 240GB, but there are also several that do much worse.
Power Limited Performance
In a notebook's power-limited state, the CPU, memory, PCIe bus, and SATA bus operate at lower speeds to reduce power. This increases the battery life, but also means nearly all SSDs perform at the same performance level. Two SLC drives are on this list, the Fujitsu FSXtreme 240GB we're testing today, and the SuperSSpeed S301; both deliver a bit higher performance in a power-limited state than the MLC and TLC drives. There is a four point difference between a 2.5" HDD found in several notebooks from the manufacturer and a MLC SSD on this test, so each point carries a lot of weight. There is a five point difference between that same OEM HDD and the two SLC drives. This shows that the Fujitsu FSXtreme is very efficient when the busses are limited to save power.
NAS Cache Performance and Considerations
In this test we use an off the shelf QNAP TS-653 Pro six-bay small business NAS (review coming later this month), and pair it with four Hitachi 7,200 RPM HDDs designed for use in NAS products. We tested with and without the Fujitsu FSXtreme 240GB as a cache drive in a few random performance tests. The connection back to the host system is a single gigabit Ethernet connection. The TS-653 Pro does not have 10GbE capability, but if it did, we suspect the random performance would have increased even further.
I also test NAS and server storage performance at TweakTown. A trend is growing and is expected to get much larger in the coming years. Netgear, QNAP, and Synology already allow end users to add a SSD to the storage stack to increase random data performance over the network. We suspect Thecus will follow suit as well with a possible announcement in January at CES.
The problem with most users taking the low-cost route with consumer SSDs is the limited endurance. Samsung's 850 Pro series tops the write endurance category with 150TB data writes to the drive. SanDisk and other manufacturers use the WMI scale that weighs the data writes differently; random data (the data most likely to be placed on a NAS cache drive) wears the WMI scale faster than sequential data.
Until now, it's been very difficult for me to recommend a consumer SSD for use in a NAS for cache. Endurance plays a very large factor, and while consumer MLC drives will work, most of us build a NAS and stick it on a shelf for a decade with little thought going to the product other than daily use. I'm not convinced a consumer MLC SSD is the right product to pair with since nearly all of the data writes are small random bits.
On the other hand, the Fujitsu FSXtreme product line has the endurance to handle the workload over an extended period of time.
This isn't our first go around with a consumer / prosumer SSD using SLC flash, and I truly hope it's not our last. These products are usually made in very limited numbers since SLC flash is expensive, and given the price, only hardcore users purchase them. The long-term payout is very good for those willing to invest in such a high quality product. In our experiment with SuperSSpeed and MyDigitalSSD for the TweakTown Chris Ramseyer Special Edition 120GB SLC SSD, we had zero returns and zero known failures. I expect the same will come true for the Fujitsu FSXtreme products as well since the same formula went into making these products.
Currently, the 240GB model costs roughly $230, so the same cost formula was used as well. The 240GB FSXtreme costs about double the price of a mainstream consumer 256GB class SSD right now, but the dollar amount has shrunk this time around. I'm thinking about purchasing more drives for a high performance, high capacity RAID array, even though I already own several high performance SSDs in groups for RAID. Some may read into that as an endorsement, but just to set the record straight, it is definitely an endorsement.
Users don't get a lot of bells and whistles with the FSXtreme products. For the money, you get a drive and a some paperwork. Given that most computer cases now have at least one 2.5" drive bay, a desktop adapter bracket plays less of a role now than it did a few years ago. Fujitsu does cover the FSXtreme with a five-year warranty, which is a nice addition if you need it.
A few reviews have come from Asia for the FSXtreme 120GB and 240GB drives; most of those utilized the basic four-corner tests. The sequential performance on this product isn't as high as some other products, mainly the Extreme PRO and 850 Pro, the other two drives competing in the ultra-high performance product category. Many buyers look at the sequential performance heavily and all, but ignore the random performance, especially the random read performance that is so important to the user experience. The FSXtreme 240GB is the leader in random read performance right now, and delivers over 12K IOPS in the category.
In our PCMark 8 advanced storage test, the FSXtreme 240GB was only outperformed by the SanDisk Extreme PRO that has enough SLC area from the flash translation layer to keep the performance very high under extended heavy workload sessions.
Given the performance, endurance, and relatively low price of the Fujitsu FSXtreme, I think this product should be popular with users seeking the ultimate performance and endurance under heavy workloads. I really like the idea of using this drive as a cache for NAS products, but it also doubles as an amazing SSD for all workloads, including some previously reserved for entry-level enterprise products.
|Quality including Design and Build||96%|
|Bundle and Packaging||91%|
|Value for Money||96%|
The Bottom Line: Without a doubt, this is the most versatile SSD in this price range. SLC lasts 10x to 20x longer than MLC, and the latency is lower as well. The Fujitsu FSXtreme 240GB is the drive enthusiasts want as long as the extra cost doesn't get in the way.
PRICING: You can find products similar to this one for sale below.
United States: Find other tech and computer products like this over at Amazon.com
United Kingdom: Find other tech and computer products like this over at Amazon.co.uk
Australia: Find other tech and computer products like this over at Amazon.com.au
Canada: Find other tech and computer products like this over at Amazon.ca
Deutschland: Finde andere Technik- und Computerprodukte wie dieses auf Amazon.de