2.4.3 | Performance |
|
|
|
|
|
|
|
|
|
|
|
| Notes | 200GB | 100GB | 50GB |
|
|
|
|
| |
Maximum Burst Transfer Rate |
| 300MB/s |
|
| |
|
|
|
|
| |
Peak sequential 128KB read/write data transfer rate (MB/s max) | [1] | 240/220 |
|
| |
|
|
|
|
| |
Sustained sequential 128KB read/write data transfer rate (MB/s) | [2] | 200/100 | 200/60 | 200/30 | |
|
|
|
|
| |
Peak 4KB random read/write command rate (IOPs) | [3] | 30,000/25,000 |
|
| |
|
|
|
|
| |
Sustained 4KB random read/write command rate (IOPs) | [4] | 25,000/5200 |
| 25,000/2600 | |
|
|
|
|
| |
Sustainable 4KB Random combined IOPS for 5 year Endurance | [5] | 10,500 |
|
| |
(65%/35% R/W, 70% Duty Cycle) |
|
|
|
|
[1]Testing performed at Queue Depth = 32, Sequentially Preconditioned drive, using IOMeter 2008.6.28.
[2]Testing performed at Queue Depth = 32, Sequentially Preconditioned drive, using IOMeter 2006.7.27.
[3]Testing performed at Queue Depth = 32, Randomly Preconditioned drive, using IOMeter 2008.6.28.
[4]Testing performed at Queue Depth = 32, Randomly Preconditioned drive, using IOMeter 2006.7.27.
[5]Testing performed at Queue Depth = 32,
Note. IOMeter is available at http://www.iometer.org/ or http://sourceforge.net/projects/iometer/.
IOMeter is licensed under the Intel Open Source License and the GNU General Public License. Intel does not endorse any IOMeter results.
Peak performance is defined as the typical best case performance that the product will be able to achieve when the product is preconditioned as mentioned and host commands are aligned on 4KB boundaries.
Sustained performance is defined as the typical worst case performance that the product will be able to achieve when the product is preconditioned as mentioned and host commands are aligned on 4KB boundar- ies. Write values also take into account the worst case performance throttling that may occur to ensure the product meets specified reliability specifications.
Due to the nature of Flash memory technologies there are many factors that can result in values different than those stated in this specification. Some discrepancies can be caused by bandwidth limitations in the host adapter, operating system, or driver limitations. It is not the intent of this manual to cover all possible causes of performance discrepancies.
When evaluating performance of SSD devices, it is recommended to measure performance of the device in a method that resembles the targeted application using real world data and workloads. Test time should also be adequately large to ensure that sustainable metrics and measures are obtained.
2.5Physical characteristics
Maximum height
Maximum width
Maximum length
Max weight
7.0mm (0.276 in)
70.1mm (2.76 in)
100.53mm (3.958 in)
135 g (0.298 lb)
6 | Pulsar Product Manual, Rev. A |