 |
|
| leading the way to the
new storage frontier |
..... | |
|
| |
| .. |
many StorageSearch readers have asked...
Zsolt Kerekes,
editor - November 21, 2011
The market dynamics which underly
the growth expected in the enterprise
SSD market in 2012 to 2015 are well understood by all the business leaders
I've spoken to in the past year - and by
SSD market analysts
too - but they are less transparent in the investment community - from where
most of these questions come from.
So this is a condensed and updated
version of the explanations I've been giving to readers who ask me how all the
growth rates they're hearing about from different enterprise SSD companies stack
up.
So - will the enterprise SSD market really be big enough for all
these companies [list] to grow?
I'd like you to think about this....
In
2010 less than 1% of datacenter servers had any SSD acceleration built
in or directly connected.
In 2016 - 100% (nearly) of all new datacenter servers
will include SSD acceleration and most will also include substantial amounts of
SSD attached bulk storage. It could be a few years later. But there won't be
much of a market for new servers which are 10x slower than the slowest SSD
accelerated models.
That's not to say that the enterprise SSD market
will be 100x larger in revenue in 5 6 years although I do believe it
could grow to be a $50-100 billion market by
2020.
The reasons you cant just scale up and multiply past or present SSD market
revenue numbers by numbers varying from x10 to x100 to predict the future
market size are:-
- users will need
less servers than they need now to do the same work if each SSD
accelerated server is 3x to 10x faster. (This reduces the multiplication
factor.)
- new SSD enabled data markets will create new demand for servers and SSDs
which don't exist today. (This increases the multiplication factor - but in a
way that's impossible to predict reliably.)
- users will deploy significantly more SSD per server in future years than
they do now. (Because they will have more confidence in what SSDs can do for
their business and because SSDs will become more competitive as storage devices
- compared to HDD RAID -
and won't just be used because of their ROI as apps accelerators.)
- SSDs will get cheaper.
(So a 2010 $100K budget will buy more SSD in future years.)
Whatever
the actual SSD market scale-up will be - x20, x50 or x100 - only time will tell.
But along the way it will be a staggeringly high revenue growth rate in the
next 3 to 5 years.
Simple message for the next few years is that no
single oem or group of oems will be able to service all the demand for
enterprise SSDs
That means large sales opportunities for enterprise SSD makers who
make it easy for their customers to buy from them and offer credible products
and credible pricing.
End users who buy enterprise SSDs will have different preferences
about who they buy from (think - preference distribution curve). So an SSD oem
who focuses on Fortune 500 accounts for example will not necessarily be the
supplier of choice for small and medium sized user sites.
Where do
resllers fit into this picture?
In my view - SSDs are not yet near to
being a reseller market. The level of technical knowledge required to ensure
that an SSD solution is needed and will actually do what is expected requires
analysis skills which most resellers don't have.
Auto-teiring products
and the transition towards a pure SSD market will change that dynamic gradually.
Going
back to the original question... I think there is enough room for all the SSD
companies named in the reader's [list] to grow / maintain high double digit
growth.
For the next 2-3 years it will be an easy ride. Then it will
get tougher because it will become clearer which SSDs are better adapted for
particular market apps.
I think there's room for maybe 10 to 15
enterprise SSD makers to each surpass multi-billion dollar revenues in the
market shape of the future. Because they will be operating in 3 to 5
distinct market silos
where the characteristics of the SSDs are dramataically different.
More
about that later. In the meantime you may be interested in these other
enterprise SSD articles.
SAS SSDs
PCIe SSDs
RAM SSDs
this way to the petabyte
SSD
MLC
flash in enterprise SSDs
11 Key Symmetries in
SSD design
what do enterprise
SSD users want?
High Availability
enterprise SSD arrays
how fast can your SSD
run backwards?
Yes you can! - swiftly
sort Enterprise SSDs
2012 - Year of the
Enterprise SSD Goldrush
the Problem with
Write IOPS - in flash SSDs
Data Integrity
Challenges in flash SSD Design
don't all PCIe SSDs
look pretty much the same?
RAM SSDs versus Flash
SSDs - which is Best?
the New Business Case
for auto accelerating SSDs
Rackmount SSDs - open
vs proprietary architectures
if Fusion-io sells more
does that mean Violin will sell less?
what do I need to
know about any new rackmount SSD? | |
| . |
| HA / FT
enterprise SSD arrays |
|
| | |
| ....... |
 | |
|
 |
| ....... |
If Megabyte spent less time on email
he'd get more SSD
articles written. | |
| ....... |
|
|
| ....... |
|
|
| . |
 |
| . |
|
|
| . |
why so
fast?
where to heading?
inside enterprise SSD market growth |
In September 2012 - I realized
that an important article was missing from my core list explaining why the
enterprise SSD market is growing so fast and where its revenue heading.
3 key earlier articles had been:-
But something was missing.
And I realized that an
important 4th article was needed to alert readers to another new trend I've
been seeing in 2012 - which will make the enterprise SSD market even bigger
still than predicted by all those earlier models.
It was the home page
blog iin september 2012 - but now it's got a new permalink.
You can't
really understand the business dynamics of the enterprise SSD market without it.
See why for yourself in -
the big market impact
of SSD dark matter | | |
| . |
| |
