SSD PRIMER:
A BASIC GUIDE TO
UNDERSTANDING
SOLID-STATE DRIVES
Application Note
Document #AN0013 – Viking SSD Primer | Rev. A
Purpose of this Document
This application note was prepared to help OEM system designers evaluate the performance of Viking solid state
drive solutions by using the same benchmarking methodology that Viking performs in it’s SSD test facility. The
SSD performance stated in the Viking SSD datasheets can be achieved by following the same Viking approach to
SSD benchmarking which has been outlined in this document.
DRAM MEMORY & FLASH STORAGE
NVDIMM, SSD, DRAM, MCP & CUSTOM
for Embedded, Industrial, Defense & Aerospace
SSD Primer
Table of Contents
1 SOLID-STATE DRIVES DEFINED 4
2 FLASH TECHNOLOGY FOR SSD’S 4
2.1 Types of NAND Flash 4
2.2 Geometry of the NAND 5
2.3 How NAND flash works 5
2.3.1 NAND Interface 5
2.3.2 NAND Attributes 6
2.4 Useful-Life and Shelf-Life of NAND 6
3 CAPACITY OF SSD’S 6
4 SPEED OF SSD’S 7
4.1 Definition of SSD Speed/Performance 8
5 SSD INTERFACES 8
5.1 USB Drives 8
5.2 Compact Flash Drives 9
5.3 CFast Drives 9
5.4 SATA Drives 9
6 SSD’S IN A SYSTEM 10
7 SSD FORM FACTORS 10
8 SSD OPERATION 11
8.1 SSD Specific Terminology 12
8.1.1 Data Compression 12
8.1.2 Data Hardening 12
8.1.3 Data Retention (Shelf Life) 13
8.1.4 Drive Life Protection (and Performance Throttling) 13
8.1.5 Endurance (Useful Life) 13
8.1.6 Garbage Collection 14
2 | VIKING TECHNOLOGY – AN0013 SSD Primer | Rev A
SSD Primer
8.1.7 Native Controllers 14
8.1.8 Over-Provisioning 14
8.1.9 Read Disturb 14
8.1.10 S.M.A.R.T 15
8.1.11 Spare Blocks 15
8.1.12 SSD Index Tables 15
8.1.13 Storage Tiering/Pairing 15
8.1.14 TRIM 15
8.1.15 Wear-Leveling 15
9 SSD’S FEATURES 15
10 FREQUENTLY ASKED SSD QUESTIONS (FAQ) 16
11 REFERENCE DOCUMENTS 17
12 ABOUT VIKING TECHNOLOGY 17
13 REVISION HISTORY 17
Table of Tables
Table 2-1: NAND Attributes ___________________________________________________________ 6
Table 5-1: USB SDD Speed Grades ____________________________________________________ 9
Table 5-2: SATA SDD Speed Grades ___________________________________________________ 9
Table 8-1: Endurance Trend _________________________________________________________ 14
Table of Figures
Figure 3-1: Example of Formatted Capacity reported by Microsoft Windows _____________________ 7
Figure 7-1: Some Examples of SSD Form Factors ________________________________________ 11
Figure 8-1: Simplified SATA SSD Block Diagram _________________________________________ 12
3 | VIKING TECHNOLOGY – AN0013 SSD Primer | Rev A
SSD Primer
1 Solid-State Drives Defined
A Solid-State Drive (SSD) sometimes referred to as an SSD flash drive, is made from
semiconductor components known as “NAND flash”. NAND flash is rewriteable non-
volatile semiconductor memory. There are several types of NAND flash and each type
has a different price/performance measure associated with it. Flash drives are usually
made with only one type of flash and are not mixed with other types of NAND.
All SSD’s include a flash controller chip to transfer data, manage flash operations and
interface to the host computer.
SSD’s come in all shapes and sizes (a.k.a. Form Factors). The most common sizes
are similar to hard disk drives used in servers, storage systems and computer laptops.
SSD’s are smaller, faster and more rugged than hard disk drives. Today, SSD’s are
quickly replacing hard drives in applications where performance and environmental
challenges matter most. They can outperform hard drives in many platforms, including
laptops, desktops, storage systems and servers. The balance of capacity, high
performance, and reliability make SSD’s the first choice for high speed data storage in
enterprise, industrial and mobile computing applications as diverse as:
• Blade Servers, Rack servers, Storage arrays
• Tablet PC, Rugged/Mobile PCs/Laptops
• Network Cache Appliances, Telco Switches
• Event Recorder, Solid-State Recorders (SSR)
• Military/Aerospace Computers and Instrumentation
• UAV/UGV, Robotic Systems
• Single Board Computers (SBC), Vending machines
• Display systems, Video on Demand (VOD)
• Point of Sales (POS), Kiosk, Digital Signage
2 Flash Technology for SSD’s
A flash device in a single monolithic package is called a flash chip. Several flash chips
are mounted onto a printed circuit board (PCB) inside the SSD to create large storage
capacities.
2.1 Types of NAND Flash
There are at least 4 different types of NAND flash chips in use today; each
categorized according to the amount of data a memory cell can hold and the length of
its useful life.
• SLC - Single-Level Cell stores 1-bit of data
• MLC - Multi-Level Cell stores 2-bits of data
4 | VIKING TECHNOLOGY – AN0013 SSD Primer | Rev A
SSD Primer
• eMLC - Enterprise Class MLC (High Endurance MLC)
(Longer life versions of MLC that employ slower erase/programming
time to lengthen life)
• TLC - Tri-Level Cell stores 3-bits of data
• QLC - Quad Level Cell stores 4-bits of data
MLC flash is the most common flash type used in SSD’s, primarily because it has the
lowest cost, often measured as $/Gigabyte. The other NAND types generally
command higher prices due to their longer life, faster access times, higher reliability,
and greater temperature tolerances. eMLC and SLC flash are usually found inside
high-end SSD products used for industrial, military or business enterprise applications
that require high performance under heavy usage ( 24 hours/day x 7 days/week).
2.2 Geometry of the NAND
The geometry of the NAND is the physical spacing between electrical circuits inside
the NAND. As the spacing gets smaller, more memory can be packed into the chip
making it less expensive. Today, MLC NAND is in the range of 20- 35 nanometers for
memory capacities of 32Gbit and 64Gbit. However, one side effect of putting more
memory into a smaller space is the reduction in the useful life of the flash. This is
where SSD controller technology is critical and why SSD’s from different vendors are
not the same.
2.3 How NAND flash works
Flash data is stored as an electrical charge (voltage) in memory cells. Billions of
memory cells are organized into “pages” of data. The pages are then accessed by the
flash controller in “blocks” at the NAND interface. Flash is a random access medium in
which the time to access a page of data does not depend on location (unlike a hard
drive’s sequential access nature).
2.3.1 NAND Interface
There are several choices for the NAND interface to the flash controller and flash
chips are designed using only one type, which often depends on the manufacturer of
the flash chip.
• Asynchronous (a.k.a. Legacy ) access mode prior to year 2010
• ToggleDDR or ONFI access modes for year 2010+
5 | VIKING TECHNOLOGY – AN0013 SSD Primer | Rev A
SSD Primer
2.3.2 NAND Attributes
While the technology of NAND has remained essentially the same over the last
several years, the geometry continues to shrink while capacity continues to grow. The
following table shows the migration of certain NAND attributes over the last 5 years.
Table 2-1: NAND Attributes
Attribute 2006 2007 2008 2009 2010
Max Die Density 2Gb MLC: 8Gb SLC: 8Gb/ SLC:16Gb SLC:32Gb
MLC:16Gb MLC: 32Gb MLC: 64Gb
Geometry trend 9x 7x 5x 3x 2x
Geometry (nm) 90 72 50 34 25
Page Size 2112 2112 4314 4320 8640
(bytes)
Block Size (KB) 128 MLC: 256 SLC: 256 SLC: 512 SLC: 1024
MLC: 512 MLC: 1024 MLC: 2048
Toggle Interface SDR 40 SDR 50 SDR 50 DDR 166 DDR 200
(MT/s)
Packages TSOP (leads) TSOP (leads) TSOP (leads) TSOP (leads), TSOP (leads),
BGA (balls) BGA (balls)
Temperature Commercial Commercial Commercial Commercial Commercial
Range 0 – 70c 0 – 70c 0 – 70c 0 – 70C, 0 – 70C,
Industrial Industrial
-45 to +85c -45 to +85c
2.4 Useful-Life and Shelf-Life of NAND
Flash is unlike other memory semiconductors because NAND flash has a finite useful
life that wears-out over time as well as a limited “shelf-life” for holding data over long
periods of inactivity. The term “flash endurance” is the useful life of the flash and the
term “data retention” can be considered as the shelf-life of the flash.
There are several methods used to extend the longevity of SSD’s; the most important
being the flash controller, not the flash, since endurance continues to drop as device
capacities increase. The intelligence of the flash controller manages how often each
flash memory cell is being used and maximizes the useful life of the SSD by
“managing” the technical limitation of flash endurance and data retention. (See the
section on SSD Operation)
3 Capacity of SSD’s
Capacity refers to the maximum amount of data an SSD can store and is described in
3 different ways.
1) raw capacity (unmanaged physical storage capacity of the flash)
2) usable capacity (raw capacity less any capacity reserved for the flash controller,
over-provisioning etc.
6 | VIKING TECHNOLOGY – AN0013 SSD Primer | Rev A
SSD Primer
3) formatted capacity (usable capacity less the capacity reserved for formatting
and use with the operating system software)
Raw capacity is never completely available to the user, since flash must reserve a
certain amount of capacity for the internal flash controller and its management tasks.
Usable capacity is the total amount of disk space available to the operating system
software, but generally not to the user. The actual usable capacity is based on a level
of “over-provisioning”, which is a portion of raw capacity reserved for the SSD flash
controller to perform housekeeping and optimization of flash usage. (i.e. “wear
leveling” tables to extend SSD life, spare flash cells to replace bad flash cells that
accrue over time, file pointers and mapping, etc.) Today, SSD usable capacities are
described in Gigabytes (GB) ranging from 25GB upwards to 2TB (Terabyte). It is
typical for client grade SSD’s to be marketed with their raw capacity, however, there is
always some amount of over-provisioning included. For example, a 256GB client
SSD may actually have about 245GB usable space. Due to their usually lower
performance and feature set, additional over-provisioning is not always useful. Client
SSD’s marketed with their usable capacity are going to have a feature set closer to
that of an Enterprise SSD. Enterprise class SSD’s are specified by their usable
capacity. While an Enterprise SSD may contain 256GB of raw flash to ensure high
performance and 5 years of endurance in datacenter applications, the usable capacity
is more typically 200GB.
Formatted capacity, in most cases, is the true measure of capacity that is available to
the user. It varies depending on the type of format and type of operating system. An
example of how the actual formatted capacity can be found is by clicking on the disk
drive icon (or ‘My Computer” in Microsoft Windows) to show its properties.
Figure 3-1: Example of Formatted Capacity reported by Microsoft Windows
4 Speed of SSD’s
The SSD data transfer rate depends on a number of factors including flash controller
architecture, NAND interface, NAND type, and host system interface.
7 | VIKING TECHNOLOGY – AN0013 SSD Primer | Rev A
SSD Primer
4.1 Definition of SSD Speed/Performance
Although “throughput” and “bandwidth” are common ways to describe performance,
SSD speed is really based on two important read/write metrics:
1) Throughput between the host and SSD for sequential writes and reads to
contiguous blocks of flash memory. Sequential bandwidth speed is measured
in megabytes per second (MB/s).
2) Operational performance between the host and SSD for writes and reads to
random blocks of flash memory. Random read/write performance is measured
in the number of Input/Outputs Per Second (IOPS).
The most significant speed metric to an SSD user will depend on the intended
application. Applications that involve frequent transfers of small and random amounts
of data to the drive will benefit from high IOPS numbers which provide a quick
response time to the host (i.e. virtualization, caching, databases, OLTP etc.).
Applications that involve moving large sequential amounts of data (i.e. video
streaming, data acquisition, data backup/restore, event recording, image processing
etc.) will benefit from high throughput (i.e. MB/s).
The SSD flash controller can be setup using the operating system to optimize for
IOPS or bandwidth (MB/s) in order to increase the speed of small data transfers or
large data transfers, and to match the size and type of data access.
5 SSD Interfaces
There are several host system interfaces available for SSD’s.
• USB - Universal Serial Bus
• CF - Compact Flash
• Cfast - Compact Flash with a Serial ATA interface
• SATA - Serial ATA
• SAS - Serial Attached SCSI
• FC - Fibre channel
• PCIe - PCI express
• Ethernet
SAS, FC, PCIe and Ethernet SSD’s are usually for advanced users or business
enterprises needing extra performance and reliability. Because of the sophistication of
these drives, they are beyond the scope of the introductory document on SSD’s.
Contact Viking for more information.
5.1 USB Drives
The speed range of USB SSD’s is dictated by the version of the flash controller used,
as shown in the table below:
8 | VIKING TECHNOLOGY – AN0013 SSD Primer | Rev A
SSD Primer
Table 5-1: USB SDD Speed Grades
USB Gbps
Version 1.0 Up to 0.0015
Version 1.1 Up to 0.012
Version 2.0 Up to 0.48
Version 3.0 Up to 5
Notes:
Gbps, Gigabits per second
Later versions that are usually faster are backwards compatible to work in
slower/older computer ports.
5.2 Compact Flash Drives
CompactFlash (CF) cards use an ATA / PC card / CompactFlash Interface and can
support various data transfer modes. The transfer rates can be in the range of
40MB/sec or 3K IOPS. Compact flash cards measure slightly under 1.5 x1.75 inches
and are available in two thickness.
• Type 1 (0.13 inches)
• Type 2 (0.26inches )
5.3 CFast Drives
The host system interface for CFast drives is SATA with additional enhancements to
the specification such as sleep modes, current limiting and hot swap capability.
CFAST drives use the compact flash card (CF) form factor with similar mechanical
dimensions, but employ a different connector.
5.4 SATA Drives
Serial ATA (SATA) SSD drives support the ATA command protocol. Similar to USB
drives, SATA drives can also operate at different speed grades based on the version
of the flash controller, as shown in the table below:
Table 5-2: SATA SDD Speed Grades
SATA Gbps
Version I Up to 1.5
Version II Up to 3.0
Version III Up to 6.0
Later versions of SATA, have higher data transfer rates, but are backwards
compatible to work at slower speeds in slower/older computer ports..
9 | VIKING TECHNOLOGY – AN0013 SSD Primer | Rev A
SSD Primer
6 SSD’s in a System
The number of SSD’s that can be deployed in a system is only limited by the number
of sockets (a.k.a ports) available. Most computers today come with multiple USB and
multiple SATA ports making it possible the run several SSD’s at the same time, either
internal or external to the system. However each SSD must be compatible to its port
(SSD interface must match the host system interface for the port).
7 SSD Form Factors
Historically SSD’s, were designed with the same dimensions and packaging as the
hard drives they were replacing. Today, however, SSD’s can break away from
traditional form factors and take full advantage of the flexibility that Solid-State
technology affords.
The popular “legacy” HDD form factors are:
• 1.8” (which is actually 2.1” wide)
• 2.5” (which is actually 2.7” wide)
Newer form factors can shed the traditional plastic or metal cases and be incorporated
into the system as bare board drives that are either reside in sockets or are soldered
directly to the printed circuit board. Examples include:
• SSD’s mounted in the memory sockets of the computer (SATADIMM)
• Board mounted SSD’s hard soldered to the computer motherboard
• MO-297 (a.k.a. SlimSATA, a 2” x .5” bare board SSD with a socket)
• PCIe SSD (SSD mounted on the PCI Express computer bus in the computer
• MO-300 (mSATA) mini-PCIe SSD measures 50.80mm x 29.85mm x 4.85mm
• M.2
10 | VIKING TECHNOLOGY – AN0013 SSD Primer | Rev A
SSD Primer
2.5” 1.8” “MO-297 (SlimSATA)
PCIe card mini PCIe MO-300 (mSATA) embedded USB
Figure 7-1: Some Examples of SSD Form Factors
Note: Pictures are not actual size and are enlarged or reduced
Large capacity SSD are also available as separate external storage systems and can
be connected to a computer or network of computers using a Fibre Channel or
Ethernet interface. These “storage appliances” provide massive Solid-State storage
capacities and use specialized architecture to provide redundancy and higher data
reliability (i.e. RAID).
8 SSD Operation
The flash controller writes and reads NAND blocks over several pathways, called
channels on the flash side and converts them into sectors that are accessible over a
single pathway on the Serial ATA (SATA) host computer side as shown below.
11 | VIKING TECHNOLOGY – AN0013 SSD Primer | Rev A
SSD Primer
Figure 8-1: Simplified SATA SSD Block Diagram
The complexity of the flash controller to optimize the write/read performance, reliability
and life of the SSD is a major differentiator among different SSD suppliers. Low-end
SSD’s often have low cost flash controllers that provide only basic functions for storing
data under light or infrequent workloads and do little to increase flash life. These are
typically referred to as client-grade SSD’s.
Enterprise class SSD’s perform under continuous operation, thereby requiring more
sophisticated flash controllers and longer life flash that can withstand the higher-end
workloads. The firmware and controllers for these SSD’s are designed to ensure
reliability under heavy usage , extended SSD life, maximum SSD speed (IOPS or
throughput) and provide for advanced error detection and correction..
8.1 SSD Specific Terminology
To get a basic understanding of how an SSD works, it is important to understand its
associated terminology. Many client SSD’s do not have all these features.
8.1.1 Data Compression
Some controllers have the ability to compress data by analyzing repeatable data
patterns and writing those patterns only once to the flash. Controllers with data
compression capability can reduce the file size for data that is not 100% random (i.e.
encrypted, zipped, PDF etc.) thus enhancing write amplification to minimize flash write
cycles and increase SSD life.
8.1.2 Data Hardening
In the event of an unexpected power loss to the SSD, the flash controller and NAND
memory will use internal self sustained power to finish any write cycles currently in
12 | VIKING TECHNOLOGY – AN0013 SSD Primer | Rev A
SSD Primer
process to the flash and also complete all internal housekeeping tasks like table
updates. (i.e. this can occur on removal of the SSD from the computer or upon system
power failure that kills power to the drive while it is actively writing data). Data
hardening prevents data corruption that would result from an ungraceful shutdown of
the flash, particularly MLC NAND, which is more susceptible to errors than SLC
NAND.
Typically, a client grade SSD does not have this sort of data protection, however,
enterprise class SSD’s typically have the ability to self sustain power to finish these in-
flight writes and gracefully shut down the SSD.
Flash memory devices exhibit unexpected behavior upon power failure. It is important
to note that if a host write is in process when power is lost, that data will be lost and/or
the contents of other MLC flash may become corrupted. Data that has been
programmed or erased during a power failure can be unreliable, even if the data
appears to be intact. Power failure can also corrupt existing data in flash prior to
power failure. If possible, the host should issue a Standby Immediate ATA8 command
(Opcode E0h) before powering down of the SSD
8.1.3 Data Retention (Shelf Life)
SSD’s have a limited time (Shelf-Life) for holding data before it is automatically lost
due to inactivity or after very long term storage. This is called Data Retention and it is
the measure of time from when data is last written to the time when it can no longer be
reliably retrieved.. Best case data retention is 10 years for storage of a drive that has
had light or moderate use before reaching its end of life limit and the SSD storage
temperatures is approx. 75F (25C). The worse case would be 3 months storage
(inactivity) IF the drive has reached its end of life AND the SSD is subjected to storage
temperatures of 100F+ (40C+) . Exposure to higher temperatures exacerbates the
issue and shortens data retention. Generally speaking, SLC NAND flash will have a
longer shelf life than MLC NAND flash.
8.1.4 Drive Life Protection (and Performance Throttling)
Some advanced flash controllers can be programmed to ensure that the life of the
SSD will not expire before a certain period of time. This is accomplished by adjusting
the speed of program/erase cycles (writes) to a level which will set the drive life to a
predetermined amount of time; often the warranty period for the drive. Read
operations are unaffected by this approach to protecting drive life.
8.1.5 Endurance (Useful Life)
Endurance is a measure of the maximum # of rewrites or erasures that the flash can
withstand before it can only perform reads and can no longer perform writes.
When the endurance limit has been reached, the SSD is usually “retired” from service
and can only be used as a read-only device for archiving or long term storage (i.e. for
backup data). This is the point in time where data retention (explained above) is most
critical.
13 | VIKING TECHNOLOGY – AN0013 SSD Primer | Rev A
SSD Primer
Flash endurance depends on the type and capacity of the NAND flash inside the SSD,
as shown in the table below.
Table 8-1: Endurance Trend
Characteristics 2006 2007 2008 2009 2010
Monolithic Die 2Gb MLC: 8Gb SLC: 8Gb SLC: 16Gb SLC: 32Gb
Density MLC: 16Gb MLC: 32Gb MLC: 64Gb
Geometry (nm) 9x 7x 5x 3x 2x
Endurance SLC:100,000 SLC: 100,000 SLC: 100,000 SLC: 100,000 SLC: 100,000
MLC: 10,000 MLC: 10,000 MLC: 5000 MLC: 3000
eMLC NAND (Enterprise class MLC) is ~ 10,000-30,000 write or erase cycles
8.1.6 Garbage Collection
Garbage collection refers to a flash controller’s housekeeping processes, transparent
to the host computer, that recycle any flash space that contains obsolete data, which
has already been modified and rewritten to a new flash location inside the SSD.
8.1.7 Native Controllers
A native flash controller is a single monolithic ASIC or FPGA, not needing a
separate ”interface” chip to communicate with the host computer. Native controllers
can operate at the raw speed of the host without any interface “bridge” chip.
8.1.8 Over-Provisioning
Over-Provisioning is the difference between raw capacity and actual usable capacity
based on a portion of the NAND that the SSD controller reserves for internal functions,
such as storage of the controller firmware, flash housekeeping and usage
optimization, flash endurance “wear leveling”, garbage collection, spare block
replacement of bad sectors, SSD index tables, file pointers, and logical to physical
mapping. SSD over-provisioning also works like HDD short stroking (using only the
outer/faster tracks on the disc platter). High levels of over-provisioning will lengthen
the life of the drive at the expense of reduced usable capacity. Over-provisioning
ranges can vary as a percentage of raw capacity.
8.1.9 Read Disturb
Excessive reads from the same flash memory cells can cause data loss from the
nearby unread memory cells. Some flash controllers prevent Read Disturb data loss
from multiple reads at the same flash memory address by tracking reads and
performing read wear-leveling before cell structure is negatively impacted (i.e.
repeated reads of the controller firmware stored in flash, repeated reads of system
operating system software stored on flash etc..)
14 | VIKING TECHNOLOGY – AN0013 SSD Primer | Rev A
SSD Primer
8.1.10 S.M.A.R.T
S.M.A.R.T. (Self-Monitoring, Analysis, and Reporting Technology) helps warn system
software, the system administrator, or a user of an impending drive failure, while time
remains to take preventive action.
8.1.11 Spare Blocks
Reserved drive capacity, known as spare blocks, are used to replace any sections of
flash that have become defective.
8.1.12 SSD Index Tables
A section of flash that contains the logical addresses and their mapping to the physical
flash addresses These index tables contain file pointers used by the operating system,
to indirectly access the data using the SSD flash controller as an intermediary.
8.1.13 Storage Tiering/Pairing
Storage Tiering or Pairing refers to the practice of using SSD’s and HDD’s in the same
system to gain the benefits of both types of storage devices (i.e. SSD’s for fast boot
up and high speed data I/O to frequently accessed data; HDDs for large, low cost
storage/archiving).
8.1.14 TRIM
A host initiated software command to the SSD to allow the SSD to reclaim any flash
space that has previously allocated to older files that have been deleted by the host.
8.1.15 Wear-Leveling
A flash controller housekeeping process, transparent to the host computer, that
prevents premature endurance wear-out of any flash cell by evenly distributing
Write/Erase cycles over the entire flash capacity. Wear leveling extends the
endurance of the overall SSD by preventing repeated write/erase to the same flash
locations; particularly relevant for file systems and file management algorithms that
repeatedly write/erase identical logical locations which are physically mapped.
9 SSD’s Features
There are many features of SSD’s that are common to hard drives such as
• Data Encryption (AES-128, AES-256)
• Secure Erase
• CRC and ECC operation
• Bad block management
• Industry standard host interface
15 | VIKING TECHNOLOGY – AN0013 SSD Primer | Rev A
SSD Primer
Since this document is a basic flash SSD Primer, any additional information on hard
drive features that are available for SSD’s, can be obtained by contacting Viking.
10 Frequently Asked SSD Questions (FAQ)
Q1: What are the benefits of SSD's over HDD's in the enterprise environment?
A1: SSD’s can offer 10x-100x IOPS performance over HDD’s and reduce the total cost of
ownership (TOC) using fewer drives. This also reduces power consumption and
cooling and the smaller form factor reduces space requirements.
Q2: What is the best usage for SSD's in the Enterprise storage environment?
A2: Storage tiering or pairing of SSD’s with HDD’s. Viking SSD’s can cache high speed
data I/O to frequently accessed data while HDDs provide large, low cost storage for
archiving infrequently accessed data.
Q3: What is Viking’s fit in the Enterprise SSD market?
A3: Viking can leverage it’s expertise in creating small form factor modular solutions
using controllers, DSP’s, NAND flash and DRAM to create products having value
and
performance to fit our customer applications.
Q4: How will SSD's complement Viking’s existing memory business?
A4: Viking will continue to market a variety of products to provide customer solutions.
Please visit Viking’s website at http://www.vikingtechnology.com for a complete
overview of the Viking’s product offering.
Q5: What type of SSD's are planned for the future?
A5: A Viking sale representative can provide you with our SSD roadmap for our SATA,
SAS and PCIe product lines. These new SSD’s are available in variety of form
factors and performance levels.
Q6: What type of Flash memory is used in the Viking SSD’s?
A6: Viking offers single-level cell (SLC) and multi-cell (MLC) NAND flash SSD’s.
Q7: What is the reliability of the Viking SSD’s?
A7: Enterprise SSD reliability is based on a number of factors including bit error rate
(BER), Mean Time Between Failure (MTBF), write endurance and data retention.
The values for these parameters are specified in the individual product datasheets.
Q8: How are Viking SSD's priced?
A8: Viking’s SSD focus is on enterprise, industrial and military markets. We also offer
lower cost client SSD’s without power failure protection, less S.M.A.R.T. features and
with lower random read/write IOP performance.
Q9: Who are typical Viking’s SSD customers?
A9: Viking supports all Tier-1 and Tier-2, global enterprise OEMs and ODM as well as
16 | VIKING TECHNOLOGY – AN0013 SSD Primer | Rev A
SSD Primer
leading industrial and military companies worldwide. Our mission is to develop and
deliver high-technology products that optimize the value and performance of our
customers’ applications in the Networking & Communications, Enterprise Computing
& Storage, Defense & Aerospace, and Embedded & Industrial markets.
11 Reference Documents
• Wikipedia information on SSD’s
http://en.wikipedia.org/wiki/Solid-state_drive
12 About Viking Technology
Viking Technology develops and delivers innovative high-technology products that
optimize the value and performance of our customers’ applications. Founded in 1989,
Viking Technology has been providing Original Equipment Manufacturers (OEMs) with
industry leading designs, engineering, product support and customer service for 20
years. For more information visit http://www.vikingtechnology.com.
13 Revision History
Related Products Viking Element SSD Datasheets
Author
Date 4-28-12
Document Number AN0013
Date Edited by Alteration to previous document revision
9/14/17 Revise logo. Change company address
17 | VIKING TECHNOLOGY – AN0013 SSD Primer | Rev A
Global Locations
US Headquarters India Office Singapore Office
2950 Red Hill Ave.
Costa Mesa, CA 92626
A 3, Phase II, MEPZ-Special Economic Zone
No 2 Chai Chee Drive
NH 45, Tambaram, Chennai-600045
Main: +1 714 913 2200 Singapore, 109840
India
Fax: +1 714 913 2202
For all of our global locations, visit our website under global locations. For sales
information, email us at [email protected]
DRAM MEMORY & FLASH STORAGE
NVDIMM, SSD, DRAM, MCP & CUSTOM
for Embedded, Industrial, Defense & Aerospace
