The trend towards smaller silicon geometries and more bits per cell in flash devices has dramatically reduced the cost-per-gigabyte for NAND flash-based SSDs, accelerating deployment in mainstream applications. However, these changes have also reduced the reliability characteristics of flash devices, resulting in lower endurance, reduced data integrity and shorter data retention. Advanced flash management technologies help deliver enterprise-class reliability with even consumer-grade flash.
DuraClass™ technology is a set of NAND flash management features that work together to deliver world-class solid state drive (SSD) endurance, reliability, performance and power efficiency. DuraClass technology differentiates Seagate® SandForce® flash controllers from the competition.
SandForce DuraWrite architecture reduces the amount of data written to the NAND flash memory inside solid state drives (SSDs) for longer effective endurance and higher performance. Storing less data also creates additional free space for dynamic over-provisioning, generating even more improvements. These innovations lower the total cost of ownership in both enterprise and client applications.
After writing to a block of flash memory, it must be erased prior to writing new data, a procedure known as a program/erase cycle. Since the longevity of solid state drives is related to the cumulative number of P/E cycles, writing less data extends SSD life by decreasing wear on the NAND flash memory.
DuraWrite architecture reduces the amount of data before writing it to flash memory. With less actual data to read and write, I/O operations require less bandwidth resulting in higher performance.
Solid state drives perform periodic clean-up of flash memory blocks containing deleted information, a process called garbage collection. DuraWrite architecture uses less space for data thus creating space for additional dynamic over-provisioning, enabling garbage collection to operate much more efficiently and to take much less time. Since garbage collection shares I/O bandwidth with read and write requests from the user, faster garbage collection translates into higher performance.
More free space for additional dynamic over-provisioning enables garbage collection to operate with far fewer data movements. Less data movement reduces write amplification, which increases SSD life.
The degree of benefit from DuraWrite architecture depends on the degree of entropy, or randomness, in the underlying data. Low entropy data offers more potential for data reduction, and will see greater benefits from DuraWrite architecture. For typical workloads DuraWrite architecture can reduce data by 50% or more*, yielding a write amplification of less than 1.0. In contrast, without a data reduction technology such as DuraWrite architecture, write amplification by definition must be more than 1.0.
Reducing the amount of data stored on flash offers tremendous endurance, performance and cost benefits. These benefits span myriad solid state storage applications, and Seagate DuraWrite architecture is a robust and proven solution in both client computing and enterprise applications, such as hyperscale datacenters.
* Source: Seagate internal testing
NAND flash manufacturers have continually reduced the cost of each generation of flash memory, accelerating the adoption of flash storage. Each new generation of flash memory uses fewer electrons per cell to store data, and as a consequence, data integrity suffers and the expected life of the NAND is reduced. The transition from MLC (multi-level cell) to TLC (three-level cell) further exacerbates the problem because each cell must store more information. Seagate SHIELD technology uniquely combines adaptive code rates, smart handling of transient noise, and a multi-level error correction schema to deliver industry-leading error correction that helps transform the latest NAND flash memory into robust storage solutions.
SHIELD technology is a unique implementation of low density parity-check (LDPC) code that combines hard-decision, soft-decision, and digital signal processing (DSP) to provide a comprehensive error correction solution for flash memory. Previous ECC techniques based mainly on Bose-Chaudhuri-Hocquenghem (BCH) codes strain to meet the broad spectrum of product requirements with newer flash. SHIELD technology innovations enable SSD manufacturers to deliver enterprise-class product life and data integrity, even using less expensive memory with higher error rates.
The frequency of read errors varies across the blocks of any NAND flash memory chip. To deliver optimum reliability with minimal overhead, SHIELD technology implements a variable code rate per block that increases ECC for weak blocks and reduces ECC for strong blocks. The allocation of ECC space is dynamic, so the number of ECC bytes for a given block increases over time as a block ages and generates more errors. During the beginning of life (BOL) of the flash, the unused ECC space is automatically converted to additional over provisioning (OP) space. Then as the solid state drive (SSD) approaches its end of life (EOL), part of the OP is gradually consumed to enable a much stronger ECC.
Flash memory suffers from various types of noise including program/erase cycling, retention, and read disturb. Such noises may cause an HLDPC decode failure, triggering SLDPC and the associated performance penalty. The SHIELD error recovery policy includes a suite of techniques designed to identify transient noise sources, and to prevent additional failures within the same page, block or area of the chip.
SLDPC decoding using digital signal processing (DSP) techniques achieves better error correction than HLDPC decoding, but it triggers two types of latency associated with collection of additional channel information and decoding. To minimize the overall performance impact of error correction, SHIELD technology implements a multi-level retry schema that applies progressively stronger decoding methods. Multi-processor parallelism is utilized to further reduce latency.
The Seagate's flash controller engineering team has a long history of advancing the state-of-the-art in error correction. This experience and engineering expertise is leveraged in the SHIELD error correction technology for SandForce flash controllers. SHIELD technology will be available with Seagate SandForce SF3700 family flash controllers.
RAISE technology provides additional measures of protection against data loss beyond what ECC can do. The combination of Seagate SHIELD error correction and Seagate RAISE data protection provides an uncorrectable bit error rate (UBER) of 10^29, nearly one quadrillion times better than other controllers. Together, these technologies allow manufacturers to build enterprise-grade solid state drives with consumer-grade flash memory.
RAISE technology writes data across multiple flash die to enable recovery from a page or block failure. RAISE technology delivers RAID-like data protection in a single drive environment, and without the write overhead associated with RAID parity. Moreover, RAISE data protection operates within a single drive without impacting performance. SF3700 family products provides an additional measure of protection that can recover from the failure of an entire die. For mission-critical applications, auto-reallocation recursively restores full die protection after a failure. For low capacity storage, fractional RAISE offers page- and block-level protection by allocating a portion of a die.
Data security is a broad concern spanning many industries, including storage. Nearly all solid state drives (SSDs) today store data directly to flash memory without performing any encryption. Passwords typically protect these systems from would-be thieves, but in the hands of a skilled technician, the flash memory of an SSD can be accessed directly with a special “clip”. Some systems use software encryption to protect data stored on an SSD, but software encryption consumes valuable host resources and creates overhead that slows access to the SSD. Seagate SandForce flash controllers solve this problem by using dual automatic hardware encryption to protect the information it stores on flash and to prevent unauthorized access. Automatic hardware encryption transparently encrypts every piece of data written to the drive.The encryption is done inside the SSD without any dependence on the host system and without slowing down the data transfer. Seagate SandForce flash controllers are compliant with the TCG Opal specification making it easier for SSD manufacturers to achieve compatibility with security management applications. The result is simplified deployment and management of SandForce Driven® self-encrypting SSDs in corporate laptops.
Given ever-increasing data volumes, the cost of data storage is a critical concern. While solid state drives (SSDs) offer numerous performance benefits over hard drives, the cost per gigabyte (GB) of flash memory is still considered high compared to traditional data storage alternatives. Seagate®SandForce® DuraWrite™ Virtual Capacity (DVC) is a unique feature of Seagate SandForce flash controllers that extends the available storage space beyond the stated capacity of the underlying flash memory, helping to lower the cost per gigabyte of flash storage.
DVC is a unique data reduction technology, which takes advantage of space normally used as additional over provisioning in an SSD. DVC returns a portion of this extra capacity to the user as additional storage space. DVC results in an increased storage capacity for the same physical flash memory, thereby reducing the cost per GB of delivered capacity to the user. With typical database applications, internal Seagate testing has shown that DVC can be used to more than triple the capacity for storing user data.
The total increased capacity provided by DVC will vary based on the compressibility of the data itself. Traditional operating systems do not comprehend storage devices that change capacity over time. For this reason, DVC provides a dedicated SMART attribute to enable applications to control and monitor the remaining free physical space and manage the amount of valid data to ensure proper performance and behavior. For customers who can modify or create their own storage operating system, this monitoring and control capability can be completely automated.
DVC offers greater advantages in applications with lower entropy data workloads, such as the boot drives and caching drives found in many hyperscale datacenter and cloud computing servers. Applications with higher entropy data workloads still benefit from this technology, but the capacity increase will not be as high.