Solid state drives (SSDs) have become a popular storage technology in computers and other devices. Unlike traditional hard disk drives (HDDs), SSDs have no moving parts and instead store data in integrated circuits. A key advantage SSDs provide over HDDs is much faster read/write speeds due to not relying on physical disk rotation and arm movements. However, one question that often comes up is whether SSDs can operate without power.
How do SSDs work?
To understand if an SSD can work without power, it’s important to first look at how SSDs work. At the core of an SSD is flash memory, which stores data electronically in memory cells made up of floating gate transistors. When power is applied to the SSD, it allows data to be written to or read from the flash memory. The SSD has a controller that manages all of the read and write operations to the flash memory.
Data is stored on the SSD in blocks, which are comprised of multiple pages. Pages are typically 4KB in size, while block sizes vary based on SSD type but are commonly 256KB or 512KB. The controller maps logical block addresses from the host system to physical addresses on the flash memory. When data is written to the SSD, it is first staged in a buffer before the controller writes it to empty pages in the flash memory. Similarly, when data is read, the controller will fetch it from the flash memory pages and store it temporarily in the buffer before sending it back to the host over the SSD’s interface.
SSD power requirements
For an SSD to perform any kind of operation – like reading, writing, erasing or accessing data – it requires power. Power is needed for several key functions:
- Running the SSD controller – The controller manages all communication between the SSD and host system. It requires power to execute its firmware, temporarily store data in the buffer, map LBAs, and perform tasks like garbage collection, wear leveling, and error correction.
- Memory operations in the flash chips – Applying voltage to the flash memory cells controls whether they store a 1 or 0. Power is needed to read, write, and erase data in the memory cells through applying precise voltages.
- Maintaining stored data – The flash memory cells use electron charge to maintain the state of a 1 or 0 when power is removed. However, a small amount of power called a “keepalive voltage” is required to prevent leakage over time.
- Other components – Power is required for the external DRAM buffer, internal communications between components, monitoring systems like temperature sensors, and operating status LEDs.
In summary, power is continuously required for the SSD controller to function, perform data I/O, maintain the state of the flash memory cells, and operate other supporting components. An abrupt loss of power to the SSD would result in any data held in volatile buffers/caches being lost, while the stored data content would remain intact.
Can an SSD controller operate without power?
The SSD controller is a complex integrated circuit that is essentially a small computer system on a chip. It has a CPU, firmware, volatile memory, and interfaces to manage the SSD. Like any computing device, an SSD controller requires continuous power to operate. If power is cut, the controller will immediately stop functioning, caches will be cleared, and it will cease operations until power is restored. No operations involving the controller, including read, write, or data requests, could be carried out without power.
Some SSD controllers have capacitors or supercapacitors integrated onto their circuit boards. These can provide power for a very brief period, typically less than a second, to allow the controller to commit any cached data to the flash memory in the event of sudden power loss. However, they are not a standalone power source and the controller still requires main power to performnormal tasks.
Can flash memory store data without power?
The flash memory chips that make up the core storage in SSDs can retain data without power. This is possible due to the floating gate transistor design used in flash memory cells. When data is written to a cell, it changes the number of electrons stored on the floating gate, which modifies the cell’s threshold voltage and sets its stored value to a 1 or 0. The floating gate design acts as a non-volatile memory, meaning it retains the charge state even when power is removed from the cell.
However, there are some important caveats around how long data can be retained on an unpowered SSD:
- Charge leakage – Over time, the electron charge stored in the floating gates will slowly leak away. This leads to data retention problems where the stored state of 1s and 0s in the cells begin decaying. Most flash memory used in consumer SSDs is rated for 1 year of data retention at room temperature when unpowered.
- Applied voltage required – While no power is needed to maintain the data state, an external voltage must be applied to the flash memory to actually read the value stored in a cell. So the data contents remain trapped on an unpowered SSD.
- Support circuitry unavailable – Supporting components around the flash memory like voltage regulators, charge pumps, data buses, and output drivers require power to operate and make the stored contents accessible.
In summary, the flash memory itself has the ability to store data indefinitely with zero power. However, real-world factors limit the practical data retention timespan to months or years before the stored data is lost. And without any power applied, there is no way to access the stored data.
Boot process requirements
For an SSD to boot as a startup disk and load an operating system, it requires power. The boot process involves several power-dependent steps:
- Power is applied to SSD components and startup firmware routines are run on the controller.
- The controller loads initialization data like the flash translation layer mapping tables into its DRAM buffer.
- The SSD identifies itself to the host computer and communicates parameters like storage capacity.
- Host sends read requests to load bootloader code from reserved areas on the SSD.
- SSD controller transfers bootloader code into host memory via interface protocol.
- Bootloader executes and continues loading operating system components from SSD as needed.
This sequence highlights the back-and-forth communications and data transfers between SSD and host system required for booting. With no power to the SSD, none of it would be possible since the controller would be non-functional. So an unpowered SSD cannot boot or load an operating system under any circumstances.
Battery backup options
Because SSDs require constant power to operate, the question arises as to whether they can be powered in a computer or device that is disconnected from mains power. Some options to allow an SSD to work without external power include:
- Battery pack – Dedicated battery packs can be purchased to power an SSD when no other power source is available. These often attach directly to the SSD housing via a connector cable and hold enough capacity to power the SSD for hours of runtime.
- USB power bank – A USB power bank with sufficient wattage output can readily power a typical 2.5″ SSD connected via a USB 3.0 or USB-C cable. High capacity power banks can provide runtimes of several hours on a full charge.
- Supercapacitors – As mentioned earlier, some SSDs have supercapacitors built into their circuit boards to provide brief power continuity in the event of sudden external power loss. However, these are not sized to act as a usable alternative power source.
The large storage capacities and fast transfer speeds of SSDs require considerable power compared to storage media like flash drives. As a result, power sources have to be fairly large and robust to provide practical operation time without external power connected. Small batteries or capacitors only act as brief buffers rather than full SSD power sources.
Operating power requirements
To get a sense of the power levels required to operate an SSD, here is an overview of typical power draw:
- 2.5″ SATA III SSD – Active: 2-3W Idle: 0.2-0.5W
- M.2 NVMe SSD – Active: ~5W Idle: ~0.5W
- PCIe Add-in-Card SSD – Active: 10-15W Idle: 1-2W
Active power describes when the SSD is receiving heavy read/write traffic and the controller is processing at full performance. Idle is light traffic in standby mode.
Based on average power draw, here is how long example backup power sources could sustain common SSD form factors:
| SSD Type | Battery Pack | USB Power Bank |
|---|---|---|
| 2.5″ SATA SSD | ~5 hours (50Wh pack) | ~15 hours (20,000mAh) |
| M.2 NVMe SSD | ~3 hours (50Wh pack) | ~10 hours (20,000mAh) |
| PCIe AIC SSD | ~1 hour (50Wh pack) | ~3 hours (20,000mAh) |
This illustrates how power demands increase for higher performance SSD form factors, resulting in much shorter potential runtimes from battery sources. Nonetheless, well-sized batteries or power banks can power most SSDs for meaningful periods of time without external power.
Supercapacitor data persistence
As mentioned earlier, some SSDs integrate supercapacitors onto their circuit boards to provide brief power to protect data in the event of unexpected power loss. For example, if power is unexpectedly disconnected, the supercapacitor allows the SSD controller enough time to transfer data from caches to the non-volatile flash memory.
Supercapacitors have high power density but low overall energy storage. This makes them ideal to handle sudden power transitions, but unsuitable as a standalone SSD power source. Here are typical supercapacitor capabilities on consumer SSDs:
- Capacity – Around 20-40 Farads
- Persistence time – Up to 1 second at max power draw
- Voltage – 5V rated output
While the persistence time is very brief, it is sufficient to provide the milliseconds to microseconds needed to save cached data before power is completely lost. But the low capacity means supercapacitors cannot power SSD operation on their own beyond initial data persistence.
Custom supercapacitor solutions
There are specialized SSD solutions that integrate much larger supercapacitors or ultracapacitors to act as temporary power buffers. For example:
- Industrial SSDs – Designed for unreliable power environments, these use large external supercapacitors that can provide seconds to minutes of power backup during outages.
- DIY enthusiast builds – Hobbyists have created SSDs with external capacitors banks that store enough power to sustain write caching and other short term needs.
With large and expensive enough supercapacitor banks, power backup duration could be extended substantially. But cost, size, weight, and safety concerns generally make this impractical compared to standard battery solutions.
Non-volatile memory technologies
Aside from flash memory, there are some other SSD technologies that have innate non-volatility and can retain data indefinitely without power:
- 3D XPoint – Developed by Intel and Micron, 3D XPoint stores data by changing the resistive state of memory cells. It doesn’t require power to maintain cell data state.
- Magnetoresistive RAM (MRAM) – Stores data in magnetic tunnel junctions. The magnetic state is retained with no power needed.
- Ferroelectric RAM (FeRAM) – Uses ferroelectric layer in cell capacitors to code binary states that are persistent without power.
These technologies theoretically allow data persistence and retention with zero standby power. However, discreet ICs made with these non-volatile memories currently have relatively low capacities and high costs compared to NAND flash. As a result, their usage is confined to special purposes like on-device buffers rather than as primary storage media in SSDs currently.
Final considerations
To summarize the key points on whether SSDs can operate without power:
- The SSD controller requires constant power to execute firmware, temporarily cache data, and perform all management operations.
- While flash memory itself retains data when powered off, voltage must be applied to read or access the stored contents.
- Batteries or power banks can be used to power SSDs without external power for limited runtimes.
- Supercapacitors in SSDs provide very brief data persistence but lack capacity for sustained operation.
- In the future, non-volatile memories like 3D XPoint may enable storage and data access with zero standby power.
So in practical usage today, SSDs will always require external power sources to function and access their stored data. But battery backup options as well as advances in persistent memory technologies may enable newer power-free SSD operation modes going forward.
Conclusion
SSDs require continual power to execute controller operations, access flash memory contents, run support components, and perform data I/O. While flash retains data without power over temporary timespans, batteries or external power must be connected to utilize SSD functionality. Integrated supercapacitors provide very brief power to retain caches on sudden outage but lack capacity to operate SSDs standalone. Future non-volatile memory may enable storage that both retains data and is accessible without external power, but this is not yet practical in SSD implementations. For now, SSDs remain dependent on constant power sources to deliver their fast performance and robust storage capabilities.
