Puolijohdelevyt (SSD-levyt) tarjoavat suorituskykyistä ja kompaktia massamuistia sulautetuille tietojenkäsittelysovelluksille, jotka mahdollistavat nykyiset digitaaliset elämäntyylit. Niitä löytyy laitteista kuten autojen tietokone- ja navigointijärjestelmistä, ohutasiakkaista, myyntipäätteistä, monitoimitulostimista, telekommunikaatiolaitteista, tehdasautomaatiojärjestelmistä sekä kaikenkokoisista palvelinjärjestelmistä aina pienistä datakeskuksista pilvipalvelutason infrastruktuureihin.

SSD-levyn tyypillinen käyttöikä riippuu vahvasti käytetystä NAND Flash -teknologiasta. QLC- (neljän bittitason) ja TLC- (kolmen bittitason) NAND-tyypeillä on käyttöikä noin 1 000–3 000 ohjelmointi/poistokierrosta solua kohden, kun taas MLC- (kahden bittitason) NANDilla käyttöikä on noin 3 000–10 000 kierrosta. SLC- (yhden bittitason) NANDilla on pisin käyttöikä – jopa 100 000 kierrosta. Näiden teknologioiden tunnetut rajoitteet auttavat järjestelmäsuunnittelijoita valitsemaan sovellukseensa sopivimman SSD-tekniikan ottaen huomioon tarvittavan kestävyyden, hinnan ja järjestelmän kriittisyyden. TLC-NAND, jolla on lyhin kestävyys, on yleensä edullisin vaihtoehto ja siten suosittu hintatietoisissa kuluttajatuotteissa.

Vaikka kunkin NAND-teknologian tyypillinen käyttöikä voidaan ennustaa, yksittäisen levyn todellista käyttöikää ei voida tietää tarkasti. Odottamattomat viat voivat aiheuttaa tietojen menetyksiä, järjestelmäkatkoksia ja korjauskustannuksia. Seuraukset voivat vaihdella pienestä haitasta vakaviin ongelmiin. Järjestelmäsuunnittelijat tarvitsevat työkaluja, jotka auttavat ehkäisemään näitä ongelmia hallinnoimalla ja suojaamalla SSD-levyä sen käyttöiän maksimoimiseksi sekä toteuttamalla ennakoivia toimenpiteitä ennen vian ilmenemistä.

Vikaan menevä SSD voi heikentää järjestelmän suorituskykyä merkittävästi. Hidastuneet luku- ja kirjoitusnopeudet, pidentyneet vasteajat ja toistuvat virheenkäsittelyt voivat johtaa hitaaseen vasteeseen ja heikentyneeseen järjestelmätehokkuuteen.

Vian jälkeen palautus ja korjaus voivat edellyttää viallisen SSD:n vaihtamista, tietojen palauttamista varmuuskopioista ja järjestelmän uudelleenkonfigurointia – kaikki nämä voivat olla aikaa vieviä ja aiheuttaa lisäkustannuksia. Ammattiapua saatetaan tarvita ongelman diagnosointiin ja ratkaisemiseen. Joissain tapauksissa koko sulautettu järjestelmä voi vaatia vaihtamista, jos SSD-vikaa ei voida helposti korjata.

SSD-vikojen seurausten lieventämiseksi on tärkeää käyttää asianmukaisia varmuuskopiointistrategioita, toteuttaa redundanssitoimenpiteitä ja ottaa käyttöön seurantajärjestelmiä.

Artikkeli on ilmestynyt ETNdigi-lehden numerossa 1/2025. Sitä pääset lukemaan täällä.

Tässä artikkeli kokonaisuudessaan.

PROTECT YOUR DATA

Solid-state drives bring inherently high reliability to applications from entry-level consumer equipment to mission-critical systems. Proper data protection mechanisms can maximise the lifetime of the drive by taking preventative measures when needed.

Solid-state drives (SSDs) provide high-performing and compact mass storage for embedded computing applications that make today’s digital lifestyles possible. They are found in equipment such as car computing and navigation systems, thin clients, point-of-sale terminals, multifunction printers, telecommunications equipment, factory automation tools, and server systems of all types and sizes from small data centres to ­­­­cloud-level.

The typical SSD lifetime is heavily dependent on the NAND Flash technology used. QLC (quad-level cell) and TLC (triple-level cell) NAND have a lifetime between 1,000 to 3,000 program/erase cycles per cell, while the expected lifetime for MLC (multi-level cell) is 3,000 to 10,000 cycles. SLC (single-level cell) NAND has the longest lifetime, up to 100,000 cycles. These known limitations help system designers choose the optimum SSD technology for their application, depending on the required endurance, price, and mission criticality. TLC NAND, with the lowest endurance, typically comes at a lower price than the others and is usually sought for use in cost-conscious consumer products.

Although the typical lifetime range for any given technology can be predicted, the actual lifetime of any given drive cannot be known. Unexpected failure can cause data loss, system downtime, and repair costs. The consequences can range from inconvenient to catastrophic. System designers need tools that can help avoid these problems by managing and protecting the SSD to maximise its lifetime and taking preventative measures before any drive failure occurs.

A failing SSD can significantly degrade system performance. Slow read/write speeds, increased access times, and frequent error handling can lead to a sluggish response and decreased overall system efficiency.

After failure, recovery and repair can involve replacing the faulty SSD, restoring data from backups, and reconfiguring the system, all of which can be time-consuming and result in additional costs. Professional assistance may be needed to diagnose and resolve the issue. In some cases, the entire embedded system might need to be replaced if the SSD failure cannot be easily rectified.

To mitigate the consequences of SSD failure, it is essential to have proper backup strategies, redundancy measures, and monitoring systems in place.

SILICON MOTION INTELLIGENT DATA PROTECTION SERIES

Among the tools available for managing SSDs safely and securely, built-in data protection mechanisms can both prevent and mitigate expected lifetime-based failures. Additionally, they can provide protection against hazards such as prolonged duration, excessive read access, temperature fluctuations, overtemperature, power instability, hacking,, all of which can shorten the life of the drive.

FerriSSDs come with a set of intelligent data protection mechanisms that meet the storage needs of different application systems and can be effectively activated in various environments where data loss may occur.

These mechanisms include IntelligentLog, which comprehensively records the health status of the SSD to let system developers understand its current condition and take appropriate measures (see picture below). When the SSD approaches its usage limit, it is possible to proactively inform the host of any potential danger. In terms of read and write cycles, data migration is initiated to ensure data integrity and prevent any loss. It also records major event logs for faster and more efficient debugging. In addition, IntelligentScan performs automatic scanning and repairs, ensuring that data stored in the secure area can be read and written to correctly. It increases the duration of data storage and extends the lifespan of the SSD (see figure 1).

Figure 1. IntelligentScan & DataRefresh functions can detect high-risk factors before the data is lost.

OVERTEMPERATURE PROTECTION

Exposure to high temperatures can present a number of hazards to SSDs, including accelerating the degradation of the NAND flash memory cells, which over time, can shorten the lifespan of the drive. Excessive heat can also affect the integrity of the stored data and increase the risk of read or write errors. Performance degradation can also occur, including slower access times, decreased transfer rates, and latency issues, resulting in decreased overall system performance and responsiveness.

To prevent such problems, FerriSSD IntelligentThermalTM monitors the drive’s temperature and activates cooling mechanisms if a certain threshold is exceeded. It effectively chooses between host-controlled thermal management (HCTM) and drive-controlled thermal management (DCTM) to control temperature and prevent SSD damage in high-temperature conditions.

POWER LOSS

Power loss can affect data protection in several ways. If power is suddenly lost during a write operation, the SSD may not have enough time to complete the data writing process. This can result in partially written data or metadata inconsistencies, causing the stored information to be corrupted or inaccessible. Also, power loss during remapping processes such as wear levelling, which is performed regularly to distribute write operations across memory cells evenly, can lead to data mapping errors or incorrect tracking of data locations. This can cause data loss or fragmentation, affecting the overall reliability and accessibility of the stored data.

In addition, data temporarily stored in cache can be lost if a power outage occurs before the cached data is written into the main storage, resulting in data inconsistencies or file system corruption. Also, power loss can interrupt the updating or synchronization of metadata used for organising and managing the stored data, leading to corruption or inconsistencies that can cause issues during boot-up or result in the loss of file system integrity.

To protect against power disruptions, FerriSSD IntelligentShieldTM ensures that all data being transmitted is securely stored in a designated safe area, preventing data loss caused by power outages. The drive initiates this mechanism as soon as instability in the power supply is detected and also activates FerriSSD’s IntelligentFlushTM. This enables instant data migration at a consistent speed, allowing data to be swiftly moved to a secure storage block (see figure 2).

Figure 2. IntelligentFlush enables instant data migration at a consistent speed, alowing data top be swiftly moved to a secure storage block.

COMPLETING PROTECTION FOR DEVICES

The data protection features also include preventing cyberattacks that could compromise the SSD's management functionality. Because all data transmission and protection mechanisms within the SSD are managed by the firmware, safeguarding them against attacks is critical. Ferri’s IntelligentGuard technology achieves this by enhancing authentication and protection mechanisms of the SSD's firmware (see figure 3).

Figure 3. IntelliGuard technology maintainsa boot security by enhancing authentication and protection mechanisms of the SSD´ s firmware.

Additional features include IntelligentImage, a preloading mechanism for embedded systems that allows data to be loaded onto a secure block prior to surface mounting assembly. This protects the data against manufacturing hazards such as damage by high reflow soldering temperatures and significantly enhances production line efficiency.

Finally, IntelligentZones supports a multi-Name Space design, allowing the SSD to be tailored according to the requirements of different system designers. The SSD is divided into different storage blocks, and data is placed in the appropriate block based on the design. This ensures data integrity as well as effective utilization of SSD capacity, thereby enhancing economic efficiency.