With Apple’s MacBook Air standardizing on flash based storage system for their ultra-portable laptop, can we see this trend picking up with enterprise storage as well? Why is the industry excited about the Solid State Storage? Is the Solid State technology trouble-free and mature enough? What are the types of flash based storage and what capacities do they come in today? These are some questions well be answering in this article.
What are Solid State Drives/Disks?
Solid State Devices (SSD’s) are electronic components on circuit boards, basically silicon chips. They have already been used to store data in a number of applications (Random Access Memory – RAM in computers, for example). Now, the technology has progressed enough to enable higher storage densities, sometimes equaling magnetic media based rotating hard disks (HDD). Two types of SSD’s that are commonly used in such storage applications include DRAM and FLASH based storage systems.
Why is the industry excited about Solid State Storage Devices?
Solid State Storage devices have number of advantages (when compared to rotating hard disks). Some of them are,
- Lower power consumption (around 1/3rd of what a hard disk might require).
- Because of lower power consumption – lesser cooling.
- SSD’s are silent as there are no moving parts.
- Higher I/O rates/ performance.
- Storage capacities almost equal to hard disks.
- Very high Random Access Read performance/ speeds.
- Shock and Vibration resistance.
- Operation at a wider temperature ranges and altitudes.
- SSD’s come in a variety of shapes and sizes – they can be much smaller than their equivalent hard disks. They, also have lesser weight.
- Low latency (especially useful for real time applications).
- Error correction and management algorithms can be applied to SSD’s. Like hard disks, they can find out how much capacity is worn out.
- SSD’s can implement features like hardware based RAID (for redundancy) and hot pluggable drives (for minimizing downtimes), etc.
- MTBF (Mean Time Between Failures) rates are claimed to be higher for SSD’s but they are still prone to individual component failures.
What are the limitations (Dis-advantages) of Solid State Disks/ Drives?
When compared to magnetic media based rotating hard disk drives,
- For SSD’s, the cost per bit is still higher. While that is coming down (50% cost reduction in two years with 50% increase in capacity), hard disks also have a similar price reduction/ capacity increment trend.
- SSD’s are being used for large scale data storage, only since recently but hard disks have been around for about 5 decades. Hard disks have been field tested for quite a long time and the technology is mature enough (known limitations with available workaround techniques) for storing huge amounts of data.
- While SSD’s do not have mechanical chances of failure, there are chances for other types of failure like component failures, connection and solder failures, firmware failures etc.
- For sequential access workloads, magnetic hard disks are better. Hard disks are also better with Random Write performance.
- There are software encryption limitations for SSD’s (This is because, SSD’s cannot overwrite data but they write in available free space).
- The more prominent type of SSD’s – MLC based Flash SSD’s suffer from limitations like limited number of erase-write cycles before wear out,
corrupted data when accessing adjacent memory cells, etc.
What are the different types of commonly used SSD’s?
There are basically two types of SSD’s that are commonly used : DRAM based SSD’s and Flash based SSD’s. DRAM based SSD’s are costlier, and hence the flash based SSD’s are more popular.
There are two types of Flash based SSD’s: NOR based SSD’s and NAND based SSD’s. The NAND based Flash SSD’s are being used more as they have higher chances of mitigating the dis-advantages of Flash based SSD’s.
There are again two types of NAND based Flash SSD’s:
SLC (Single Level Cell): The Single Level Cells are a type of NAND based SSD that store a single level of charge in each cell and hence contain single bit of information in each cell. They support around 1,00,000 erase program operations before wear out, which is considerably higher than the other type called MLC – Multi Level Cell. They have higher write speeds and higher bit error rates when compared to MLC.
MLC (Multi Level Cell): The Multi Level Cells are a type of NAND based SSD that store four different charge states in a cell and hence contain two bits of information in each cell. They support around 10,000 erase program operations before wear out, which is considerably lower than the SLC. They have lower write speeds and lower bit error rates when compared to SLC.
MLC based Flash SSD’s are being manufactured in bulk and have a distinct cost advantage over SLC based Flash SSD’s, currently. Vendors are finding out ways to mitigate the dis-advantages of MLC, in order for them to be accepted in enterprise segment. On the long run though, SLC based Flash SSD’s might be more popular in the enterprise segment.
Currently, it may be better for IT departments to balance out storage requirements by moving frequently accessed and mission critical (latency sensitive) applications in to faster, more expensive media like SSD’s and less frequently accessed bulk storage in to slower, less expensive media like the hard disks. So, it is recommended to adopt Tired Storage with a combination of hard disks as well as flash drives, based on the application.
What are the storage capacities supported by SSD’s?
Commonly available SSD’s can go up to 512 GB of storage capacity, while hard disks are available up to 2-3 TB.
How are SSD’s implemented?
- SSD’s (flash disks) are sometimes integrated in to the existing storage arrays.
- They are implemented as rack mountable Stand-Alone SSD systems too.
- Some of them are implemented as In-Server solid state storage where flash chips mounted on printed circuit boards are integrated in to the data stream through common bus connector types like PCI-Express, etc.
In case you have any more points to add or any questions, you can use the comment form below or contact us using the contact form. You could keep yourself updated with computer networking technologies by subscribing with your email address in the box titled “Get email updates when new articles are published”.