Original Link: https://www.anandtech.com/show/8689/corsair-neutron-xt-ssd-review
Corsair Neutron XT (240GB, 480GB & 960GB) SSD Review: Phison S10 Debuts
by Kristian Vättö on November 17, 2014 9:00 AM EST
The ugly truth is that the SSD market has been tough for almost all the typical PC component vendors lately. With Samsung, SanDisk, and Micron/Crucial being serious about the client market, it's not easy for other companies to find a way to provide any substantial advantage over the big brands. The fab owners will always enjoy a cost advantage, which is why we have seen a few companies backing off from the market and many more for whom SSDs have become just another series in the pool of products.
This is what could be said to have happened at Corsair. While the company has never been super aggressive on the SSD side, the past year or so has been very quiet. There have been a couple of new releases, such as the Force LS and Force LX, but nothing close to the exclusive deal Corsair scored with Link A Media (LAMD) two years ago.
One way to perceive whether a company is serious about a specific model is with the media sampling process. If there is no embargo and only one capacity is available for review, then historically the product has not been a substantial or interesting release. If, on the other hand, there is an official release date and the manufacturer samples several of the available capacities, then this can be a sign that the product might be a bit more special. This is never a hard and fast rule, and it is always welcome to be surprised.
From the title, the Neutron XT from Corsair is a member of the latter case. This is the first time Corsair has offered us the full set of drives since the original Neutron and Neutron GTX, and the Neutron XT is the first drive to ship with Phison's new S10 controller.
Phison actually sent us reference design samples of the S10 prior to Corsair, but we were asked to review the Neutron XT first (as Corsair is a household name while Phison is pretty much the opposite). However, Phison asked us to save the technical analysis of the controller and its architecture for the separate reference design article, so I'll keep to the basic details for now; stay tuned for a more in-depth analysis of the S10 in the next two weeks.
The biggest change in the S10 compared to the S8 is the upgrade to a quad-core CPU architecture. Three of the cores are dedicated to internal flash management (garbage collection, wear-leveling etc.), whereas one will handle the host operations. The NAND channel count remains at eight like in the S8, which seems to have become the standard for client-grade controllers. Capacity support goes as high as 2TB and the S10 design can also support TLC NAND, although that is not fully ready yet.
| Corsair Neutron XT Specifications | ||||
| Capacity | 240GB | 480GB | 960GB | |
| Controller | Phison PS3110-S10 | |||
| NAND | Toshiba A19nm MLC | |||
| NAND Density | 64 Gbit per Die | 128 Gbit per Die | ||
| Sequential Read | Up to 560MB/s | |||
| Sequential Write | Up to 540MB/s | |||
| 4KB Random Read | Up to 100K IOPS | |||
| 4KB Random Write | Up to 90K IOPS | |||
| Encryption | N/A | |||
| Warranty | Five years | |||
| Availability | December | |||
EDIT: Corsair had an error in the reviewer's guide, so the warranty is actually five years similar to the original Neutron series.
Corsair is offering the Neutron XT in three capacities: 240GB, 480GB and 960GB. The Neutron brand has always been more of a high-end enthusiast/gamer brand, so it makes sense to skip the 120GB model because the majority of the target population will be aiming at 240GB at a minimum.
On the NAND front Corsair is using Toshiba's A19nm MLC NAND. The 240GB and 480GB models are equipped with 64Gbit dies to provide higher parallelism and performance, whereas the 960GB model has enough NAND on its own to provide the necessary parallelism and performance with a 128Gbit die.
Like many platforms on the market, the S10 features a page-level parity scheme that Phison calls 'Page RAID ECC Parity' to protect against NAND failures. Unfortunately Phison couldn't disclose the parity ratio (i.e. how much NAND is dedicated to parity), which makes calculating the exact over-provisioning impossible, but I was told that the feature only provides page-level protection and can't tolerate a full block or die failure.
Corsair is not announcing pricing yet because the drive is not officially launching until early December, which would subject the prices to fluctuation. However, I'll make sure to update this article with the pricing information once that becomes available.
Test Systems
For AnandTech Storage Benches, performance consistency, random and sequential performance, performance vs transfer size and load power consumption we use the following system:
| CPU | Intel Core i5-2500K running at 3.3GHz (Turbo & EIST enabled) |
| Motherboard | ASRock Z68 Pro3 |
| Chipset | Intel Z68 |
| Chipset Drivers | Intel 9.1.1.1015 + Intel RST 10.2 |
| Memory | G.Skill RipjawsX DDR3-1600 4 x 8GB (9-9-9-24) |
| Video Card | Palit GeForce GTX 770 JetStream 2GB GDDR5 (1150MHz core clock; 3505MHz GDDR5 effective) |
| Video Drivers | NVIDIA GeForce 332.21 WHQL |
| Desktop Resolution | 1920 x 1080 |
| OS | Windows 7 x64 |
Thanks to G.Skill for the RipjawsX 32GB DDR3 DRAM kit
For slumber power testing we used a different system:
| CPU | Intel Core i7-4770K running at 3.3GHz (Turbo & EIST enabled, C-states disabled) |
| Motherboard | ASUS Z87 Deluxe (BIOS 1707) |
| Chipset | Intel Z87 |
| Chipset Drivers | Intel 9.4.0.1026 + Intel RST 12.9 |
| Memory | Corsair Vengeance DDR3-1866 2x8GB (9-10-9-27 2T) |
| Graphics | Intel HD Graphics 4600 |
| Graphics Drivers | 15.33.8.64.3345 |
| Desktop Resolution | 1920 x 1080 |
| OS | Windows 7 x64 |
- Thanks to Intel for the Core i7-4770K CPU
- Thanks to ASUS for the Z87 Deluxe motherboard
- Thanks to Corsair for the Vengeance 16GB DDR3-1866 DRAM kit, RM750 power supply, Hydro H60 CPU cooler and Carbide 330R case
Performance Consistency
Performance consistency tells us a lot about the architecture of these SSDs and how they handle internal fragmentation. The reason we do not have consistent IO latency with SSDs is because inevitably all controllers have to do some amount of defragmentation or garbage collection in order to continue operating at high speeds. When and how an SSD decides to run its defrag or cleanup routines directly impacts the user experience as inconsistent performance results in application slowdowns.
To test IO consistency, we fill a secure erased SSD with sequential data to ensure that all user accessible LBAs (Logical Block Addresses) have data associated with them. Next we kick off a 4KB random write workload across all LBAs at a queue depth of 32 using incompressible data. The test is run for just over half an hour and we record instantaneous IOPS every second.
We are also testing drives with added over-provisioning by limiting the LBA range. This gives us a look into the drive’s behavior with varying levels of empty space, which is frankly a more realistic approach for client workloads.
Each of the three graphs has its own purpose. The first one is of the whole duration of the test in log scale. The second and third one zoom into the beginning of steady-state operation (t=1400s) but on different scales: the second one uses log scale for easy comparison whereas the third one uses linear scale for better visualization of differences between drives. Click the dropdown selections below each graph to switch the source data.
For more detailed description of the test and why performance consistency matters, read our original Intel SSD DC S3700 article.
 |
|||||||||
| Default | |||||||||
| 25% Over-Provisioning | |||||||||
Performance consistency has never been Phison's biggest strength and that continues to be the case with the S10 controller. The consistency is actually worse compared to the older S8 controller (i.e. Corsair Force LS) because the variance in performance is so high. I'm pretty sure the issue lies in Phison's garbage collection architecture as it doesn't seem to give enough priority for internal garbage collection, which results in a scenario where the drive has to stop for very short periods of time (milliseconds) to clean up some blocks for the IOs in the queue. That is why the performance frequently drops to ~1,500 IOPS, but on the other hand the drive may be pushing 70K IOPS the second after. Even adding more over-provisioning doesn't produce a steady line, although the share of high IOPS bursts is now higher.
For average client workloads, this shouldn't be an issue because drives never operate in steady-state and IOs tend to come in bursts, but for users that tax the storage system more there are far better options on the market. I'm a bit surprised that despite having more processing power than its predecessors, the S10 can't provide better IO consistency. With three of the four cores dedicated to flash management, there should be plenty of horsepower to manage the NAND even in steady-state scenario, although ultimately no amount of hardware can fix inefficient software/firmware.
 |
|||||||||
| Default | |||||||||
| 25% Over-Provisioning | |||||||||
 |
|||||||||
| Default | |||||||||
| 25% Over-Provisioning | |||||||||
TRIM Validation
To test TRIM, I filled the drive with sequential 128KB data and proceeded with a 30-minute random 4KB write (QD32) workload to put the drive into steady-state. After that I TRIM'ed the drive by issuing a quick format in Windows and ran HD Tach to produce the graph below.
And TRIM works as expected.
AnandTech Storage Bench 2013
Our Storage Bench 2013 focuses on worst-case multitasking and IO consistency. Similar to our earlier Storage Benches, the test is still application trace based – we record all IO requests made to a test system and play them back on the drive we are testing and run statistical analysis on the drive's responses. There are 49.8 million IO operations in total with 1583.0GB of reads and 875.6GB of writes. I'm not including the full description of the test for better readability, so make sure to read our Storage Bench 2013 introduction for the full details.
| AnandTech Storage Bench 2013 - The Destroyer | ||
| Workload | Description | Applications Used |
| Photo Sync/Editing | Import images, edit, export | Adobe Photoshop CS6, Adobe Lightroom 4, Dropbox |
| Gaming | Download/install games, play games | Steam, Deus Ex, Skyrim, StarCraft 2, BioShock Infinite |
| Virtualization | Run/manage VM, use general apps inside VM | VirtualBox |
| General Productivity | Browse the web, manage local email, copy files, encrypt/decrypt files, backup system, download content, virus/malware scan | Chrome, IE10, Outlook, Windows 8, AxCrypt, uTorrent, Ad-Aware |
| Video Playback | Copy and watch movies | Windows 8 |
| Application Development | Compile projects, check out code, download code samples | Visual Studio 2012 |
We are reporting two primary metrics with the Destroyer: average data rate in MB/s and average service time in microseconds. The former gives you an idea of the throughput of the drive during the time that it was running the test workload. This can be a very good indication of overall performance. What average data rate doesn't do a good job of is taking into account response time of very bursty (read: high queue depth) IO. By reporting average service time we heavily weigh latency for queued IOs. You'll note that this is a metric we have been reporting in our enterprise benchmarks for a while now. With the client tests maturing, the time was right for a little convergence.
The Neutron XT offers a healthy performance boost over the Force LS in our 2013 Storage Bench, but overall ends up in the middle-class. The Phison S10 platform appears to be very competitive with Silicon Motion's SM2246EN (ADATA SP610 in the graphs), but it can't challenge the high-end drives from Samsung, SanDisk, and OCZ.
AnandTech Storage Bench 2011
Back in 2011 (which seems like so long ago now!), we introduced our AnandTech Storage Bench, a suite of benchmarks that took traces of real OS/application usage and played them back in a repeatable manner. The MOASB, officially called AnandTech Storage Bench 2011 – Heavy Workload, mainly focuses on peak IO performance and basic garbage collection routines. There is a lot of downloading and application installing that happens during the course of this test. Our thinking was that it's during application installs, file copies, downloading and multitasking with all of this that you can really notice performance differences between drives. The full description of the Heavy test can be found here, while the Light workload details are here.
The same story continues in our 2011 Storage Benches. The Neutron XT is an upgrade – and actually a fairly substantial one – from the Force LS, but it doesn't break any records. Also, what's a bit surprising is that the performance drops as the capacity increases because the 240GB is the fastest of the Neutron XT family. This isn't uncommon because the larger capacities have more LBAs to track, which requires more processing power for internal tasks, but it's still something that's worth noting.
Random Read/Write Speed
The four corners of SSD performance are as follows: random read, random write, sequential read and sequential write speed. Random accesses are generally small in size, while sequential accesses tend to be larger and thus we have the four Iometer tests we use in all of our reviews.
Our first test writes 4KB in a completely random pattern over an 8GB space of the drive to simulate the sort of random access that you'd see on an OS drive (even this is more stressful than a normal desktop user would see). We perform three concurrent IOs and run the test for 3 minutes. The results reported are in average MB/s over the entire time.
Random read speed gets a boost compared to the Force LS and the Phison finally appears to be on-par with the others.
Random write speed is also good, although the 240GB model could perform better at lower queue depths.
Sequential Read/Write Speed
To measure sequential performance we run a 1 minute long 128KB sequential test over the entire span of the drive at a queue depth of 1. The results reported are in average MB/s over the entire test length.
Here we see the performance drop when the capacity increases again. The 960GB model is quite noticeably slower in write speed compared to the rest of the capacities, although at 355MB/s it definitely isn't slow.
AS-SSD Incompressible Sequential Read/Write Performance
The AS-SSD sequential benchmark uses incompressible data for all of its transfers. The result is a pretty big reduction in sequential write speed on SandForce based controllers, but most other controllers are unaffected.
Performance vs. Transfer Size
ATTO is a useful tool for quickly benchmarking performance across various transfer sizes. You can get the complete data set in Bench. Both read and write performance seem to scale nicely with the transfer size and the Neutron XT is equal to the competition.
Click for full size
Power Consumption
Currently the S10 controller doesn't support slumber or DevSleep power modes, which is why the idle power consumption turns out to be high compared to the competition. Phison told me that the slumber power (HIPM+DIPM) implementation is nearly complete and a new firmware should follow in about a month, but in the meantime the S10 and Neutron XT aren't really suitable for mobile use. The load power consumption is more moderate, although there are more efficient drives and controller solutions on the market.
Final Words
From the benchmark results, the S10 is a fair successor to Phison's S8 platform. However, in terms of overall performance against the compeition, it is difficult to be impressed. Despite the added processing power in the controller, the expectation of a big increase in performance over the S8 was a little misguided, and as it stands the S10 and Neutron XT are more mid-market products.
One result to point out is the performance consistency. One purpose of the extra CPU cores is to increase performance in steady-state by handling the flash management, but that does not seem to be happening as efficiently as one might hope. If this was a value-market drive, the consistency would be okay since the drive would be subject to mostly light workloads. The Neutron branding is a higher end product, which puts the performance directly against the branding. As the branding is marketed for enthusiasts and professionals, the consistency in harsher IO workloads does matter. Phison has told me that they have a more consistent firmware in development that is geared more towards the prosumer market, but it is possible that it will also find its way into high-end client drives (although I don't know if Corsair has any plans of adopting that firmware).
UPDATE: Corsair told me that while the samples we received are already production-level candidates, discussions about firmware updates are already ongoing with Phison. In other words, there is a chance that the final retail units will carry a newer firmware, but if that happens I will of course retest the drive and provide an update.
Aside from performance, the high idle power consumption due to the lack of slumber power support limits the S10's and thus the Neutron XT's market to desktop users. Mobile users are more geared towards low powered devices and there are drives with proper power management support available. As I mentioned on the previous page, the slumber power states will be implemented to the firmware in about a month, although it will be difficult to estimate when it will arrive to the Neutron XT as the new firmware must go through Corsair's validation as well before a public release.
As pricing has not been announced yet, it's hard to draw any final conclusions. The Neutron branding hints that it is aimed at the high-end market, but on the other hand Corsair must be competitive in price to provide any advantage over the other high-end drives (especially SanDisk's Extreme Pro and Samsung's 850 Pro). I don't want to speculate too much on the pricing or whether the Neutron XT is positioned competitively because we are only a couple of weeks away from the launch.
All in all, while it is an improvement over the S8, the S10 (and by extension, the Neutron XT) lacks a 'wow' factor that would really help it stand out from the rest. The TLC NAND support in the S10 is certainly a nice feature, but I don't want to praise the feature until I see a drive shipping with the S10 and TLC NAND. The lack of increase in consistency despite the increase in CPU power is one point of concern that Phison needs to address to have a competitive product at the higher end of the market. That being said, it is difficult for any SATA 6Gbps platform to provide a massive advantage over another because of the link throughput and AHCI software stack limitations. We are also still at least a couple of quarters away from seeing more PCIe solutions in the market. Once that happens, the doors for differentiation open up again.
Facebook
Twitter
RSS
