Original Link: http://www.anandtech.com/show/8420/amd-radeon-r7-ssd-240gb-review
AMD Radeon R7 SSD (240GB) Reviewby Kristian Vättö on August 28, 2014 6:00 AM EST
In 2011 AMD took the first step in expanding the Radeon brand and partnered with Patriot and VisionTek to provide AMD branded memory. With the launch of the Radeon R7 SSD AMD is continuing this strategy by jumping into the SSD market. Just as they did with memory, AMD is partnering with a third party that handles the development, manufacturing and support of the product, which in the case of the R7 SSD is OCZ.
While the front is covered by AMD branding, the back label is OCZ branded and there is no effort to hide the fact that the drive is made by OCZ. Both companies are very open about the partnership as AMD mentions OCZ as the partner in the first sentence of the R7 SSD press release and OCZ even lists the R7 as one of their products on their website. I am glad that there is no secrecy regarding the origin of the R7 SSD because there is already enough shady things going on in the SSD industry (e.g. Kingston switching the NAND in the SSDNow V300 and PNY using multiple controllers in the Optima).
OCZ is a logical partner for AMD because the company is now owned and financially backed by Toshiba, which also gives OCZ prime access to NAND. Given the always haunting NAND shortage, a partner without direct access to NAND will more likely run into supply issues that may hurt AMD's brand as consequence. Additionally, OCZ's Barefoot 3 platform has been geared towards the higher-end market (gamers, enthusiasts and professionals) from day one and has proven to be a good performer.
As I mentioned in the launch pipeline, AMD is not really bringing anything new to the market. The number one goal with the R7 SSD and expansion of the Radeon brand is to make it easier for novice PC builders to pick parts and at the same time ensure that the parts they purchase are high quality and provide good performance. For someone that has never built a PC, the available selection can be fairly overwhelming, so AMD is trying to make the part selection smoother by providing the CPU, GPU, RAM and SSD. Obviously, the R7 SSD also gives AMD a great opportunity to offer more extensive bundles and with aggressive bundle pricing AMD could boost its CPU sales too.
|AMD Radeon R7 Series SSD Specifications|
|Controller||OCZ Barefoot 3 M00|
|NAND||Toshiba 64Gbit A19nm MLC|
|4KB Random Read||85K IOPS||95K IOPS||100K IOPS|
|4KB Random Write||90K IOPS||90K IOPS||90K IOPS|
|Steady-State 4KB Random Write||12K IOPS||20K IOPS||23K IOPS|
|Endurance||30GB/day for 4 years|
The R7 SSD is based on OCZ's Barefoot 3 M00 controller, which is the faster version running at 397MHz, while the M10 version that is used in the ARC 100 and Vertex 460/450 runs at 352MHz instead. The NAND is the same Toshiba's 64Gbit A19nm MLC as in the ARC 100 and in fact even the part number is an exact match.
What separates the R7 SSD from OCZ's other SSDs is the endurance. While the ARC 100 is rated at 20GB for three years (21.9TB total) and the Vector 150 is at 50GB for five years (91.2TB total), the R7 hits the middle ground by offering 30GB of writes per day for four years (43.8TB total). OCZ did some wear-leveling and garbage collection optimizations to achieve the higher endurance in the R7 but otherwise the R7 should be very similar to OCZ's ARC 100.
Sadly there is still no support for low power states and TCG Opal 2.0 / eDrive. The lack of TCG Opal 2.0 support I can understand since that is not the most important feature for gamers and enthusiasts, but by not having support for low power states (HIPM+DIPM and DevSleep) the Barefoot 3 platform is simply not competitive in the mobile space.
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 Drivers||Intel 184.108.40.2065 + 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|
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 Drivers||Intel 220.127.116.116 + Intel RST 12.9|
|Memory||Corsair Vengeance DDR3-1866 2x8GB (9-10-9-27 2T)|
|Graphics||Intel HD Graphics 4600|
|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 tells us a lot about the architecture of these SSDs and how they handle internal defragmentation. 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 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.
The IO consistency is very similar to the ARC 100 but the R7 is maybe slightly faster. Compared to the Vector 150 and Vertex 460 there is a small decrease in consistency as performance occassionally drops below 10K IOPS, but on average IOPS of 15-20K is excellent for a client drive. The same goes for IO consistency with 25% over-provisioning – the R7 is not as good as the Vector 150 and Vertex 460 but it is still one of the best performing client SSDs.
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|
|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, AdAware|
|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.
Performance in our 2013 Storage Bench is typical Barefoot 3, although it appears that both the R7 and ARC 100 are more optimized for small IOs given that they provide a lower service time, yet the data rate is marginally slower.
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.
On the other hand, the R7 does not do that well in our 2011 Storage Benches. The performance is still acceptable but it is definitely not the fastest drive around.
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 performance remains similar to other Barefoot 3 SSDs. Peak performance has never been Barefoot 3's strength, which is why especially random read performance seems slow for a modern drive, but the lack of peak performance is compensated by sustained consistency.
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.
Sequential read performance receives a nice upgrade from the Vector 150 and Vertex 460, although that comes at the cost of sequential write speed. It is quite common that an increasing in one benchmark leads to a decrese in another as in the end firmware design is about finding the right balance for IO priorization.
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.
Click for full size
Slumber power consumption remains high due to the lack of low power state (HIPM+DIPM) support, which is a dealbreaked for a mobile user but for desktop users the power increase is negligible. Load power consumption is much more reasonable with the max staying below 3W.
To be frank, the R7 does not bring anything new to the table. Performance and feature wise it provides something in between the ARC 100 and Vector 150, but there's nothing earthshaking that would separate it from the masses. Sure, a four-year warranty is unique but with Samsung and SanDisk offering 10-year warranties in their high-end SSDs, the buyers who value a long warranty will likely choose something other than the R7.
|NewEgg Price Comparison (8/24/2014)|
|AMD Radeon R7||$100||$160||$290|
|OCZ Vector 150||$85||$140||$280|
|OCZ Vertex 460||$90||$140||$245|
|OCZ ARC 100||$75||$120||$240|
|Samsung SSD 850 Pro||$130||$200||$400|
|Samsung SSD 840 EVO||$90||$140||$250|
|SanDisk Extreme Pro||-||$200||$380|
|SanDisk Extreme II||$70||$140||$295|
|Intel SSD 730||-||$190||$340|
|Intel SSD 530||$90||$140||$250|
The pricing does not make the R7 any better. Right now it is more expensive than the Vector 150, which provides higher performance and a longer warranty along with a higher endurance rating. The R7 must come down in price to make any sense, although even then the R7 will fall into the infamous middle-class. As I have said before, I only see two segments in the client market: the mainstream/value and the high-end market. In other words, the average Joe should just buy the SSD that provides the best value and frankly the MX100 has been dominating that market for the past couple of months. The high-end market, on the other hand, values absolute performance and features, and the R7 cannot compete with the 850 Pro and Extreme Pro there.
It is quite obvious that AMD is going after novice users with the R7 SSD. For someone with very little or no understanding of computers, the AMD branding can provide a peace of mind since AMD is fairly large and visible brand in the industry, but anyone who understands what they are buying should look elsewhere. To put it simply, there are better SSDs available for less money.
Ultimately AMD's biggest opportunity with the R7 SSD is bundles. I would not mind seeing the R7 as a part of an AMD component bundle because there is nothing wrong with the drive, but the pricing must be much more competitive than what the standalone drive currently is. If the total bundle price can put the cost of the SSD roughly in line with other value drives like the MX100, it provides good performance overall; it's just not worth the current price premium when purchased on its own.