
Virtualization and Cloud executives share their predictions for 2015. Read them in this VMblog.com series exclusive.
Contributed article by Saul Zales, CEO, Contour Semiconductor
What’s next for solid-state storage
‘Tis the season for
industry watchers to ask those of us on the innovation side what the
coming year will bring. We've certainly offered predictions and educated
guesses about the state of solid-state storage on this blog, but it now
behooves us to make a few bolder statements, and a few disclaimers.
(Your mileage may vary. Some restrictions apply. Void where prohibited.)
Solid
state storage technology has been proven and accepted. This seems
obvious, but it becomes more important into 2015 because solid state
will put continued pressure on the market space for high-performance
traditional hard drives.
Those
vendors that purchase mass quantities of hard drives, like EMC, NetApp,
HP, Dell and IBM understand memory is the most important component.
Supply is short, and not likely to improve in the foreseeable future.
NAND manufacturers are presently constrained, and memory component
prices are liable to go up in 2015. Many system suppliers will continue
to face challenges associated with a short list of memory suppliers..
Let's
think about how that affects other ends of the landscape. Major end
users like Amazon Web Services, Yahoo, or Facebook are very concerned
about getting performance that is appropriate for their business needs.
They're buying a lot of components - and a lot of solid-state storage.
The current memory supply situation is unlikely to resolve for several
years.
Emerging flash-based storage vendors, which tend to be more flexible and spry when it comes to pricing and configurations, are making inroads into the big players' sales. We'll have to see if this can continue if they're obliged to compete for the static NAND supply.
At
Contour, our Diode Transistor Memory (DTM) is scoring between NAND and
DRAM in terms of performance, so we're in a good space for these storage
applications, and a good position when NAND prices increase.
Looking
ever more forward, the growing pains caused by NAND's limited
production could be fully realized. This couldn't come at a worse time.
The Internet of Things will expand. It's not unrealistic to expect this
category to grow to hundreds of millions of devices in the next few
years. That's hundreds of millions more devices that all require memory.
Many of these devices will need NAND for local storage (though some
won't - again, your mileage may vary, and we mean that literally:
automotive microcontrollers also need memory).
For example, the wicked popular Nest
home automation devices use NAND, so they could be challenged to keep
up with demand. Micron has said that the standard memory in smartphones
available in China is growing from 8 GB to 32 GB - a 4x increase and a
4x challenge. Anybody producing anything with standalone memory (and we
expect to see a lot of that at CES) will be challenged.
Where's
all that memory for mobile and IOT applications going to come from?
This is another segment where DTM comes to the rescue. As a phase-change
based technology, it's not only fast; its endurance and reliability is
much better. (You can read why here.)
Perhaps
the better question is not "where" but "who:" who will be supplying all
that memory? We are hopeful that it will be Contour Semiconductor's
development partners who take advantage of the opportunity to supplant
NAND with memory technology that is superior in performance, endurance
and retention. (Some assembly required. Participating locations only.
Any similarity to actual storage technologies, living or dead, is purely
coincidental.)
##
About the Author
Saul Zales has
30 years of experience in all management aspects of the semiconductor
industry, including engineering, high volume manufacturing
relationships, marketing, business management, equity investments and
licensing. Saul previously served as executive vice president of
corporate business development at Fusion-IO, Inc. and was vice president
and general manager of corporate development at Numonyx B.V. prior to
that company's acquisition by Micron. Saul was also at Intel Corporation
for twenty-four years, last serving as director of its flash memory
business, where he guided the company's flash technology joint ventures,
joint development programs and technology licensing. Saul has a degree from the University of Pennsylvania.