The maturation of enterprise SSDs has led to further refinement of the product stack. When enterprise SSDs began to gather steam a few years ago, nearly every product was tailored for the high endurance, write-intensive segment. The formula was simple; the high price of enterprise SSDs dictated that their destination was always high-end applications. These SSDs typically featured SAS connections, and SLC NAND, as a rule.
The expansion of enterprise SATA SSDs brought along eMLC NAND, and then finally, the widespread adoption of typical MLC NAND resulted in crashing prices for SSDs. A whole ecosystem of SATA SSD's began to crop up with an SSD truly for every usage case, from read-oriented, to mid-tier and high endurance. However, the SAS segment remained largely focused on expensive, high endurance flagship-class SSDs.
The release of the SanDisk Optimus Eco brings a lower-cost alternative into the SAS stable. The high performance expected of a SAS SSD remains in place, while the write endurance of the SSD is lower than preceding products. This provides a value-oriented SAS SSD that bucks the typical high-priced SAS trend.
The Optimus family consists of two models. The Optimus is an award winning SAS SSD designed for heavy read/write workloads, and fulfills the role of the workhorse for bleeding-edge applications. The Eco steps in and provides a mid-tier solution, with High Availability features and SAS functionality. The CloudSpeed 1000 and 1000E fulfill the role of SATA-class SSDs employed for mainstream use.
The Optimus Eco is designed for read-centric workloads, yet still features impressive random read/write speeds of 90,000/35,000, blistering sequential read/write speeds of 500/500 MB/s in normal operation, and up to 1Gb/s in wideport mode. The Eco moves to 19nm NAND produced by the SanDisk/Toshiba Flash Forward joint venture. The Guardian Technology platform delivers an endurance rating of three random Drive Writes per Day (DWPD), and seven DWPD for sequential workloads.
The Eco features capacities of 400 GB and 800 GB, in a 2.5" 9.5mm form factor. This limited capacity range on the low-end denotes a focus on high capacity for the Eco series, but it does not stop there. The Eco is also available in a 15mm chassis, with massive capacities of 1.6 and 2.0TB. This high-density trumps any performance 2.5" HDD on the market, and secures a spot for the Eco in dense applications that require high performance, significant endurance, and affordability.
While the Eco sports focus on capacity and affordability, it doesn't skimp on enterprise-class features. Power capacitors provide resilience to power loss, onboard data duplicity guards against data loss, dual port SAS provides enhanced functionality, TCG Enterprise Compliance, AES 256-bit encryption nails down security, and an amazing MTBF of 2.5 million hours provides a high level of reliability. The Eco is also backed by a five-year warranty.
There is no doubt in the performance of the original SAS Optimus; it still leads our charts as the fastest 6Gb/s SSD available on the market. The Eco has much of the same design and architecture of the original Optimus, so all of the pieces are in place for a more affordable high-end SSD.
SanDisk Eco Architecture
Before assembly, SanDisk subjects all NAND to an extensive re-characterization, and assigns more accurate P/E Ratios to the NAND. Aggregated flash management ensures that there is an even distribution of different classes of NAND present to benefit the whole. Once assembled, the Guardian Technology Platform takes over, and NAND settings are monitored and adjusted constantly. This keeps the NAND perfectly in tune over the life of the SSD. The award-winning Guardian Technology Platform is SanDisk's proprietary approach, which consists of three integral components.
As NAND geometries shrink, endurance also shrinks, and data errors multiply. Data errors are typically solved with ECC (Error Correction Code). ECC can reach a diminishing point of returns when powerful ECC algorithms create excessive overhead, leading to performance degradation, and higher latency. SanDisk utilizes adaptive DSP (Digital Signal Processing), in concert with advanced ECC algorithms to combat these data errors.
NAND is typically programmed for retention and wear metrics during NAND processing, and the settings are never altered. Adaptive programming dynamically adjusts flash parameters over the course of the SSDs life; tailoring the NAND settings at the base level. This creates exponential increases in endurance of the underlying NAND. Typical SSDs can only achieve wear leveling on each NAND package. Aggregated Flash Processing treats all of the flash elements as a whole, and provides global cross-die wear leveling.
DataGuard provides full data path protection, and F.R.A.M.E. (Flexible Redundant of Memory Elements) functionality. F.R.A.M.E. is a cross-die data redundancy feature, similar to parity, which allows for data reconstruction in the event of a catastrophic event such as a flash page, or even an entire NAND block failure.
EverGuard is a third generation host power-loss protection approach that protects against unexpected power events. Discrete ploy-Tantalum capacitors supply power to flush all data to NAND during power loss. Tantalum is rated for high temperature environments, and does not degrade over time. Additional architectural features reduce power hold-up time requirements through advanced controller power management, and lower pages reserved for power-down writes.
Optimus Eco Internals
SMART CloudSpeed 1000E Internals
The Eco comes bearing the branding from SMART Storage Systems, the company that originally developed the core IP behind the Guardian Technology Suite, and was recently purchased by SanDisk. SanDisk branding will follow the scheme outlined on the first page of this evaluation, though the pretty stickers are not the real story behind the Optimus Eco.
The construction of the Optimus Eco, which consists of precision milled aluminum alloy, speaks volumes to the care taken in the design of this SSD.
Popping apart the case brings the precision milling into view. There are recessed grooves that allow the heat producing components to come very close to contacting the case. Even though there is no direct surface mating, the extreme close tolerance design provides such a slim gap between the components and case, that heat is wicked efficiently into the case, which then acts as a large heat sink. This ingenious design forgoes the use of thermal pads, yet still delivers the same thermal efficiency.
The top of the PCB holds eight of the BGA-mounted 32GB NAND packages, and the remaining eight populate the bottom of the PCB. The Toshiba 19nm NAND is jointly manufactured with SanDisk through their Flash Forward venture.
A total of 12 power loss protection emplacements are placed close to the NAND packages. The Tantalum capacitors provide the SSD with enough power hold-up to flush data to the lower NAND pages. There are three DRAM packages (for a total of 768MB) above the Marvell controller.
The Optimus Eco utilizes the Marvell 88SS9185-BLN2 controller. This key component provides flexibility for firmware customization, which is an important consideration for deploying the custom SanDisk firmware that allows the Guardian Suite to provide massive improvements in endurance, and performance.
Test System and Methodology
We utilize a new approach to HDD and SSD storage testing for our Enterprise Test Bench, designed specifically to target long-term performance with a high level of granularity.
Many testing methods record peak and average measurements during the test period. These average values give a basic understanding of performance, but fall short in providing the clearest view possible of I/O QoS (Quality of Service).
'Average' results do little to indicate performance variability experienced during actual deployment. The degree of variability is especially pertinent, as many applications can hang or lag as they wait for I/O requests to complete. This testing methodology illustrates performance variability, and includes average measurements during the measurement window.
While under load, all storage solutions deliver variable levels of performance. While this fluctuation is normal, the degree of variability is what separates enterprise storage solutions from typical client-side hardware. Providing ongoing measurements from our workloads with one-second reporting intervals illustrates product differentiation in relation to I/O QOS. Scatter charts give readers a basic understanding of I/O latency distribution, without directly observing numerous graphs.
Consistent latency is the goal of every storage solution, and measurements such as Maximum Latency only illuminate the single longest I/O received during testing. This can be misleading, as a single 'outlying I/O' can skew the view of an otherwise superb solution. Standard Deviation measurements consider latency distribution, but do not always effectively illustrate I/O distribution with enough granularity to provide a clear picture of system performance. We use histograms to illuminate the latency of every I/O issued during our test runs.
Our testing regimen follows SNIA principles to ensure consistent, repeatable testing. We measure power consumption during precondition runs. This provides measurements in time-based fashion, with results every second, to illuminate the behavior of power consumption in steady state conditions. Power consumption can cost more over the life of the device than the initial acquisition price of the hardware itself; significantly affecting the TCO of the storage solution. We also present IOPS-to-Watts measurements to highlight efficiency.
Our test pool features SSDs of varying capacity, and it is important to bear this in mind when viewing results. The first page of results will provide the 'key' to understanding and interpreting our new test methodology.
The original Optimus easily bests any other 6Gb/s SSD on the market. To keep things in perspective, we are comparing the Optimus Eco to the fastest 6Gb/s SSD on the market, its predecessor the Optimus, and the CloudSpeed 1000E provides a glimpse of the speed of a SATA SSD. All three SSDs are 400GB, and we test both SAS SSDs in dual port configuration.
4K Random Read/Write
We precondition the Optimus Eco for 18,000 seconds, or five hours, receiving reports on several parameters of workload performance every second. We then plot this data to illustrate the drives' descent into steady state.
This chart consists of 36,000 data points. This is a dual-axis chart with the IOPS on the left, and the latency on the right. The red dots signify IOPS during the test, and the light green dots are latency measurements during the test period. We place latency data in a logarithmic scale to bring it into comparison range. The lines through the data scatter are the average during the test. This type of testing presents standard deviation and maximum/minimum I/O in a visual manner.
Note that the IOPS and Latency figures are nearly mirror images of each other. This illustrates that high-granularity testing can give our readers a good feel for latency distribution by viewing IOPS at one-second intervals. This should be kept in mind when viewing our test results below.
We provide histograms for further latency granularity below. This downward slope of performance happens very few times in the lifetime of the device; typically during the first few hours of use, and we present the precondition results only to confirm steady state convergence.
Each QD tested includes 300 data points (five minutes of one second reports) to illustrate performance variability. The line for each QD represents the average speed reported during the five-minute interval.
4K random speed measurements are an important metric when comparing drive performance, as the hardest type of file access for any storage solution to master is small-file random. 4K random performance is a heavily marketed figure; one of the most sought-after performance specifications.
The Optimus Eco averages 97,092 IOPS, with a 4K random read workload at QD256. This falls just slightly below the Optimus, with 100,290 IOPS. The read speed of the Eco is simply phenomenal, with the full performance of the higher endurance model on tap with the Eco.
Garbage collection routines are more pronounced in heavy write workloads; this leads to more variability in performance. The Eco chugs along with an average of 34,424 IOPS, with a bit more variability than the original Optimus, which averages 44,326 IOPS at QD256. When considering price, this is excellent performance for the Eco.
Our write percentage testing illustrates the varying performance of each solution with mixed workloads. The 100% column to the right is a pure write workload of the 4K file size, and 0% represents a pure 4K read workload.
The Eco provides solid performance as we move into the heavier write workloads. Perhaps most interesting is the very similar performance to the original Optimus across the board. It is very clear that these two drives share very similar performance characteristics, in spite of the difference in write endurance.
The Eco easily outperforms the SATA-powered CloudSpeed 1000E.
We record the power consumption measurements during our precondition run. We calculate the stated average results during the last five minutes of the test, after the device has settled into steady state.
The Eco hits the sweet spot in power consumption, between the original Optimus and the CloudSpeed 1000E. Dual-port SAS is always going to come with a bit of a power premium due to the nature of the interface, but the inclusion of 19nm NAND helps to bring the power consumption of the Eco down to only 6.21 Watts, well below the 7.48 average of the original Optimus.
IOPS to Watts measurements are generated from data recorded during our precondition run, and the stated average is from the last five minutes of the test. The CloudSpeed 1000E wins the efficiency challenge due to its SATA interface, but the Optimus Eco and the regular Optimus nearly tie with 5,546, and 5,744 IOPS per Watt, respectively. The higher performance of the original Optimus allows it to squeak by with the win in this test.
8K Random Read/Write
8K random read and write speed is a metric that is not tested for consumer use, but for enterprise environments, this is an important aspect of performance. With several different workloads relying heavily upon 8K performance, we include this as a standard with each evaluation. Many of our Server Emulations below will also test 8K performance with various mixed read/write workloads.
The average 8K random read speed of the Eco falls slightly short of the original Optimus at an impressive 66,297 IOPS at QD256, while the Optimus chugs away at 77,268 IOPS. Both SAS SSDs reach their best performance at QD64, and then fall off slightly under heavier loads.
The Eco again has an increase in variability, but manages to pull off an average of 18,568 IOPS across the board. The original Optimus averages 23,726 IOPS.
Once again, the Eco mirrors the performance of the original Optimus across the different write percentages.
Both SAS SSDs deliver better latency than the SATA SSD, though the CloudSpeed 1000E performs well in this test.
Power consumption for the Eco averages 6.31 Watts during the measurement window, much lower than the original Optimus.
All three SSDs are closely matched in this test, with the original Optimus nudging the Eco with 3,189 IOPS per Watt, and the Eco delivering 2,936 IOPS per Watt.
128K Sequential Read/Write
The 128K sequential speeds reflect the maximum sequential throughput of the SSD using a realistic file size encountered in an enterprise scenario. The Eco manages to best the original Optimus at QD256, with an average speed of 596 MB/s, though we see speeds in excess of 650MB/s at QD128.
The Eco delivers an impressive write speed of 580MB/s at QD256, though the original Optimus manages to deliver an astounding 696MB/s.
The Eco lags behind the original Optimus in this testing, though its speeds easily trump anything available with a SATA connection.
The 1000E is remarkably consistent, delivering 99% of I/O at 60-80ms, and 72% at 80-100ms.
The Eco averages 7.15 Watts during the measurement window; well below the 9.33 Watts from the original Optimus.
The Eco averages 81 MB/s per Watt, while the original Optimus pulls off 54 MB/s per Watt. The power efficiency of the Eco really shines through on this round.
Database/OLTP and Webserver
This test emulates Database and On-Line Transaction Processing (OLTP) workloads. OLTP is the processing of transactions such as credit cards and high frequency trading in the financial sector. Enterprise SSDs are uniquely well suited for the financial sector with their low latency, and high random workload performance. Databases are the bread and butter of many enterprise deployments. These are demanding 8K random workloads with a 66% read and 33% write distribution that can bring even the highest performing solutions down to earth.
The Eco averages 33,794 IOPS at QD256.
The Eco averages an impressive 5.39 Watts during the precondition run.
The Eco nudges past the original Optimus with 6,209 IOPS per Watt.
The Webserver profile is a read-only test with a wide range of file sizes. Web servers are responsible for generating content for users to view over the internet, much like the very page you are reading. The speed of the underlying storage system has a massive impact on the speed and responsiveness of the server that is hosting the website, and thus the end-user experience.
The Eco averages 19,587 IOPS at QD256, marking the only blemish on its performance in the test. The 1000E and original Optimus rise to the challenge. Much of this performance differential is due to the awesome performance of the 1000E in this pure read workload.
The Eco averages 7.56 Watts during the measurement window.
The Eco averages 2,591 IOPS per Watt, while the 1000E shows its considerable strength in this pure read workload.
The Emailserver profile is a very demanding 8K test, with a 50% read and 50% write distribution. This application is indicative of the performance of the solution in heavy write workloads.
The Eco averages 32,782 IOPS at QD256.
The Eco averages 5.38 Watts in steady state.
The Eco averages 6,060 IOPS per Watt; easily besting the other competitors in this test.
Over the last year, the 2.5" performance HDD segment (10K, 15K) has been under assault from datacenter-class SSDs. Several features from 2.5" HDDs have helped them retain some presence in the market; price per GB, capacity, and High-Availability from the SAS interface immediately jump to mind.
The Optimus Eco lowers the ceiling on the price of a top-tier SAS SSD, with the 2TB SSD coming in at an MSRP of $3,999, to roughly $2 a GB. This price projection was made in March, before SanDisk purchased SMART. It makes sense that access to the fab, and volume pricing, could push this price even lower. 10K SAS HDDs are hovering around $1.80 per GB, allowing the Eco to begin encroaching upon the 2.5" performance segment in price.
Adding 2TB of capacity puts the Eco well out of reach of 15K HDD's from a capacity standpoint. In today's shrinking datacenter, density is king; there isn't a spindle solution on the market that can provide the Eco's density in a 2.5" form factor.
Marrying this high capacity and relatively low price point to the SAS interface, with expanded high availability and enhanced management features, really begins to remove key advantages enjoyed by the current crop of performance HDD's.
Endurance, which was once a major concern, is sufficient for the Eco. The fastest 2.5" HDD on the market cannot come close to writing its own capacity with random data in one day; it simply isn't fast enough. The Eco provides three Drive Writes Per Day (DWPD) of random data, exponentially more capacity than an HDD can even write in a day, and still holds the same five-year warranty as an enterprise HDD. The 2.5 million hour MTBF is equally as impressive.
Without even touching on the Eco's performance (which blasts away the fastest of SATA SSDs with ease), or efficiency (which is better than most SAS SSD's and obliterates HDD's), we can come to the easy conclusion that the Eco is going to enjoy massive success in the datacenter. Do we feel that the performance HDD segment is going to disappear? Of course not. However, with solutions like the SanDisk Optimus Eco, there certainly are sharks in the swimming pool. The transition to SSHD's will help stave off the encroachment from SSDs in the performance segment, but we expect the pure HDD performance market to continue to feel pressure from the new breed of high-capacity SSDs.
The Eco makes good gains on lowering the overall power consumption compared to the original Optimus, surely due to the switch to 19nm NAND. Mixing in sequential read speeds above the maximum attainable from any SATA port at 596MB/s, and blistering random read IOPS in the 97,000 IOPS range, guarantees performance beyond competing solutions in the 6Gb/s realm.
The Optimus Eco retains the functionality, and some performance characteristics of the bleeding-edge SAS SSD's, creating a value-market segment for SAS SSD's where one really hasn't existed before. The Eco is a great addition to the SanDisk portfolio, and easily wins our TweakTown Best Value award.