Embedded system designers need more from a storage system than higher capacities.
Today's applications require enhanced performance, reliability, and security,
all of which can effectively be met with advanced solid-state storage. This
technology offers many tangible benefits, including multiyear product cycles, no product wearout, the ability to accurately forecast usable storage
system life, and security options beyond encryption. But
OEMs continue to design in substandard storage based on five
prevalent myths.
Myth 1: It's too expensive
Not necessarily. The
megabyte capacity of traditional storage products far exceeds
user requirements in many applications. Yet to maintain average
selling prices, hard-drive manufacturers offer increasingly
higher-capacity hard drives, forcing users to buy 40 Gbytes or
more of storage when their application requirements may be
for as little as a few gigabytes.
A recent study concluded that most enterprise system OEM
applications require less than 4 Gbytes. An edge router is a typical
application that stores both an operating system and data
log files using less than 4 Gbytes. Since there is cost parity
between hard drives and solid-state at this capacity, solid-state
storage is no longer too expensive.
Advances in the last 15 years have made solid-state storage
more affordable today than ever. Working with a typical $250
storage budget, companies can purchase the same capacity in
either solid-state or hard drive. For the same cost as a hard
drive, they can gain the additional benefits that come with solid-
state. So, the cost per useable gigabyte now favors solidstate
storage over hard drives in many applications.
Myth 2: Superior hard-drive performance
Field failure rates for hard drives are up to 15 times greater
than datasheet specifications, according to a Carnegie Mellon
University report dated February 2007. And according to a February
2007 Google study, once a hard drive has its first scan
error, whether it be for reallocation, offline reallocation, or probational
counts, it is 39 times more likely to fail within 60 days.
When hard drives are used in more demanding applications
that extend duty cycles or in applications with vibration, temperature
variation, or other environmental challenges, field
failure rates are much higher.
Hard drives will always be better in applications requiring
massive amounts of storage capacity, since solid-state is a
long way away from offering a cost-effective terabyte of storage.
But for many applications, solid-state storage offers better
durability and a significantly lower total cost of ownership.
Myth 3: Costly product requalifications
Not so. Many flash card and hard-drive manufacturers force
requalifications on their customers due to product development
techniques they employ to minimize their bill of materials
costs. Conversely, developers of advanced solid-state storage
technology leverage product development strategies that
enable them to continually introduce technology advances
without triggering product requalifications for users.
Myth 4: Unavoidable wearout
Not with solidstate
technology. Storage systems should not wear out or fail
during the required deployment cycle, as the costs associated
with unscheduled downtime, field maintenance, product
recalls, lost revenue, and customer goodwill are significant.
Solid-state storage technology features patented technologies
such as robust wear-leveling and error correction code
(ECC) algorithms, as well as early warning systems that forecast
useable life to virtually eliminate the chances of storage
system wearout. The endurance of a traditional flash card simply
cannot match that of an advanced storage technology that
offers wear-leveling over the entire drive and 6-bit ECC.
Myth 5: Limited security options
In the past,
embedded system design challenges due to the small footprint
and low-power requirements for storage systems prevented storage
security options beyond basic encryption technology. Leading
manufacturers of advanced solid-state storage technology
offer an array of advanced user-selectable security options that
prevent IP theft, protect application data, and manage data
security via the host system, not the storage product.
Total cost of ownership
Many designers assume
that low-cost storage solutions will lower their total cost of storage
ownership, but a wide array of factors must be considered. It
is far more complex than simply calculating the cost differential
between various storage products. A lower individual unit cost
per storage product is only relevant if all products being compared
deliver the same benefits to the user.