The SSD Review The Worlds Dedicated SSD Education and Review Resource | 
TEST SETUP AND METHODOLOGY
In testing all enterprise drives we focus on long term stability. In doing so, we stress products not only to their maximum rates, but also with workloads suited to enterprise environments. We use many off-the-shelf tests to determine performance, but we also have specialized tests to explore specific behaviors we encounter. With enterprise drives, you will see that we do not focus on many consumer level use-cases.
When testing SSDs, the drive is purged and then preconditioned into a steady state before capturing its performance results. We also disable all write caching on the DUT when possible, this ensures consistent results that are complaint with SNIA standards. Our hope is that we present tangible results that provide relevant information to the buying public.
SYSTEM COMPONENTS
| PC CHASSIS: | Thermaltake Urban T81 |
| MOTHERBOARD: | ASRock X99 WS-E |
| CPU: | Intel Xeon E5-2690 v3 |
| CPU COOLER: | Thermaltake Water 3.0 Ultimate |
| POWER SUPPLY: | Thermaltake Toughpower 1500W Gold |
| GRAPHICS: | MSI GT 720 |
| SYSTEM COOLING: | be quiet! Silent Wings 2 |
| MEMORY: | Crucial Ballistix Sport DDR4 2400MHz |
| STORAGE: | Crucial MX200 500GB |
| OS: | Windows Server 2012 R2 |
| RSTe DRIVER: | 4.3 |
| MOBO FIRMWARE: | 1.70 |
| INTEL DC P3608 FIRMWARE: | 8DV10171 |
This Test Bench build was the result of some great relationships and purchase; our appreciation goes to those who jumped in specifically to help the cause. Key contributors to this build are our friends at ASRock for the motherboard and CPU, be quiet! for the cooling fans, and Thermaltake for the case. We have detailed all components in the table below and they are all linked should you wish to make a duplicate of our system as so many seem to do, or check out the price of any single component. As always, we appreciate your support in any purchase through our links!
For this specific drive we tested both single volume and RAID 0 performance. In order to RAID the volumes presented by the SSD we utilized RSTe 4.3. The RAID 0 strip size was left at default (128K) and the volume allocation size was set to 100%.
SNIA TESTING
The Storage Networking Industry Association has an entire industry accepted performance test specification for solid state storage devices. Some of the tests are complicated to perform, but they allow us to look at some important performance metrics in a standard, objective way.
SNIA’s Performance Test Specification (PTS) includes IOPS testing, but it is much more comprehensive than just running 4KB writes with Iometer. SNIA testing is more like a marathon than a sprint. In total, there are 25 rounds of tests, each lasting 56 minutes. Each round consists of 8 different block sizes (512 bytes through 1MB) and 7 different access patterns (100% reads to 100% writes). After 25 rounds are finished (just a bit longer than 23 hours), we record the average performance of 4 rounds after we enter steady state.
- Purge: Secure Erase, Format Unit, or vender specific
- Preconditioning: 2x capacity fill with 128K sequential writes
- Each round is composed of .5K, 4K, 8K, 16K, 32K, 64K, 128K, and 1MB accesses
- Each access size is run at 100%, 95%, 65%, 50%, 35%, 5%, and 0% Read/Write Mixes, each for one minute.
- The test is composed of 25 rounds (one round takes 56 minutes, 25 rounds = 1,400 minutes)
Unlike our other performance tests, the SNIA tests only last for a relatively short period of time each (1 minute), but they cover many more access patterns and transfer sizes. All tests were done at a QD of 256. Here we have the RAID 0 results. Looking at 100% read performance we can see that overall it is very solid. We don’t see the rated 850K 4K IOPS, but still 745K is very impressive. When looking at the 4K write IOPS we can see that the Intel SSD DC P3608 reached nearly 170K IOPS. Compared to the Intel SSD DC P3700 800GB we reviewed last year, it completely blows it away! We will take a closer look at 4K, 8K, and seq. performance in the coming pages so we will see if it can really achieve the stated 850K.
LATENCY
To specifically measure latency, we use a series of 512b, 4K, and 8K measurements. At each block size, latency is measured for 100% read, 65% read/35% write, and 100% write/0% read mixes.
Looking at the results the RAID 0 configuration is on the left and single volume is on the right. At QD256 average latency we can see that overall it is very well controlled, especially for the 4KB file size with all results under 2ms in RAID 0. 8KB files size the average is double the 4KB latency, which is expected and is at about 3ms. All of the single volume’s results are double that of RAID 0.
Maximum latency is very well tamed. The highest latency we see for RAID 0 is 28ms for the 8KB file size and 25.6 for the single volume results. 0.5 and 4KB file sizes’ maximum latency reached 16ms and 15ms respectively during 100% write for RAID 0 and are a bit higher for the single volume. Again, overall in RAID 0, it beats out its P3700 bother in terms of latency.
Random adjectives, desperate efforts to “humanize” the tech resulted in this huge review to contain next to no information at all.
There is no easy way to say this: software RAID 0 on PCIe is simply retarded.
Thanks for your thoughts
Now just make it affordable
Well, for enterprise it is very affordable for what you get. If you are concerned about consumers/enthusiasts I can see where you are coming from, but this is not meant for them. Next year, however, we may be seeing performance like this trickle down.
More than likely next year
As an enterprise product I can see it as a high-end workstation device but not a server device. The lack of RAIDability seems to limit its use to caching and high-speed scratch work area.
I’ve been informed that PCIe hardware RAID will be available on the Skylake CPU and the Xeon version when it comes out later. Now we’re talking………
so this is a preview, not a review… where are the comparisons to P3700 and PM951?
I don’t have access to those drives. We reviewed the P3700 in another system. Because of that as well as a change in our testing methodology, we cant not graph them side by side. Looking at the P3700’s specific review you can gauge for yourself the approximate performance difference between the two.