Solid-state drives have become the standard storage device in the current generation of laptops and ultrabooks, and are popular upgrades for desktop PCs and older model laptops. Their method of storing data inside semiconductors with no moving parts alleviates some of the data loss risk associated with older mechanical drives—for example, an SSD can better absorb physical shocks and drops compared to a mechanical hard disk drive. Though the technology is generally very reliable, SSD endurance can be affected by user behavior and requires care and maintenance that differs from a mechanical hard drive.
Measuring SSD life
The lifetime of a solid-state drive is measured in program/erase (P/E) cycles. A P/E cycle describes the sequence of events when data is written to a NAND flash memory cell inside the drive, is erased, and then rewritten. Every SSD wears down and fails after a certain number of P/E cycles. Manufacturers rate SSD endurance with an estimated number of P/E cycles; most SSDs have P/E ratings between 3,000 and 4,000 cycles.
Does this mean you should tally up each time you add and remove files from a drive? It doesn’t quite work like that. P/E ratings apply to each block of storage, and drives contain millions of blocks. Operating systems act with a drive’s memory controllers to tidy up how data is written to the blocks, so simply writing and erasing the drive’s capacity 4,000 times will not kill the drive. In most instances, the SSD inside your laptop or ultrabook should outlast your device—even if you are writing and erasing 50 GB per day.
Related: What NASA’s Mars Rover Tells Us About SSD Life
What about in more extreme conditions? Electrical engineering-types have undertaken gung-ho attempts to purposely break consumer SSDs with P/E cycles. Experiments show that consumer drives are good for about a year’s worth of constant P/E ridiculousness. Do understand that with all electronic devices, there are lemons. Drives do fail from manufacturer defect, but this usually happens inside the three year warranty that comes with new SSDs. So keep your paperwork.
That said, in normal circumstances, SSD endurance can be aided by certain behaviors and system setups.
Enable TRIM support. TRIM is a hardware configuration specific to solid state drives that can be set in Windows. TRIM helps the operating system know the best place to store new data, and when the user deletes data from the drive it immediately wipes the blocks and sub-blocks (called pages) where the that data is stored. It translates into faster performance and longer life.
How do you enable TRIM support in Windows? Navigate to the Command Prompt and type:
fsutil behavior query DisableDeleteNotify
There are two possible outcomes:
- DisableDeleteNotify = 0 : TRIM is already enabled in Windows
- DisableDeleteNotify = 1 : TRIM is not enabled → to enable it, type:
- fsutil behavior set DisableDeleteNotify 0
Don’t treat it like an HDD. Forget about disk defragmentation and other software-based disk optimization utilities used for optimizing spinning hard drives. These will not help SSD performance and may result in unnecessary P/E cycles.
Do run manufacturer’s management software. Most SSD manufacturers will include software utilities that will allow you to monitor and install patches and firmware updates. Last year, Samsung 840 EVO owners were able to repair faulty SSD performance by downloading an update from Samsung Magician, for example. It is always advisable to stay current with your hardware and there is no additional cost, making this a no-brainer.
Place large media files on hard drives. If your system has both SSD and HDD storage, it is good practice to movies and photo albums on the spinning disk storage. Media files tend to be saved, played, and moved more than application files, which is harder on SSDs. If you have a choice, HDDs are a better bet for big files like these.
Upgrade the RAM capacity. Someone over at Tom’s Hardware ran some experiments that pointed to increasing RAM capacity resulting in better SSD endurance. A system with 16 GB saw performed 42-63 percent fewer writes to the SSD, depending on the application, compared to a system with 4 GB.
Power down instead of hibernating. When you hibernate a computer, it dumps everything in RAM into storage. That takes a toll (albeit a tiny one) on a drive. Powering down a computer eliminates this. With fast SSD boot times, there is really no excuse not to power down a computer when not in use.
Pick up a higher capacity drive. Larger drives tend to last longer mostly because there are more physical blocks of memory to write on. Do not buy a larger drive just for this advantage. Buy a larger drive because users tend to accumulate data faster than they think, and a full SSD tends to slow down when it nears capacity.
Leave some empty space. A team of engineers at Anandtech explored SSD performance degradation in near-capacity SSDs. Without getting overly technical in the explanation, it is recommended that users keep around 25 percent free space on your SSD for peak performance and durability.
Related: Symptoms of Impending Drive Failure, and What to Do
Final thoughts
Hardware endurance, particularly storage drive endurance, should remain critical to users because it is tied to data—the most important technological asset there is. Having a plan or system to for backing up data is paramount in basically any computing scenario. I repeat: back up your data! Understand that disk storage is a commodity and all drives, both spinning and solid state, will fail eventually even though users can and should take measures to optimize their systems for SSD endurance and performance.
2 comments
[…] Related content: 7 Tips to Maximize SSD Endurance […]
There is a lot of “information” on the web saying an SSD will outlive the life of a typical consumer. I found that to be criminally false. You really, really, do have to be aware of write management, and in the end these things will, in fact, die sooner than you might expect.. My example is a 256GB Samsung 830 in a Linux HPTC. It went online 4 years, 2 months, and 5 days ago (Yea, SMART). It is now at 25% wear life remaining. I did all the tricks, day one. My /var, /lib, /home, and media directories are all shunted to hard disks. Huge RAM (32GB), TRIM enabled, no pointless directory updates, and 7GB unallocated partition on the SSD. The file system on that SSD averages only about 35% allocated. It suffers very, very, few writes beyond routine system software updates, and most updates actually land on one of the hard disks (libraries). So, estimated life… 6 years at $300 bucks (at the time). A better solution, at least for me going forward? Raid a few hard drives. All 8 of my hard disks (Media storage) are still tooling along, entirely error free, and have had the living “c**p” beat out of them..
In case one might think something else is writing the disk behind my back, well no. Linux has some pretty good monitoring tools. Since a system reboot, due to a major system update a week ago, the SSD has been fed a total 2.9 MB of data (the result of a fairly sizeable follow-up to the update).