Top 20 SSD
Companies - based on search volume Q3 2011 - ©StorageSearch |
rank ...company
(click logos for more info) |
editor's notes ........................................................... |
|
|
... |
|
| 1....... |
... |
 | |
Same as before.
This is
Fusion-io's 11th straight quarter in the #1 slot. Fusion-io's search volume
was 84% higher than the #2 ranked company in this list and 10x
higher than the #20 company.
I'm often asked by investors - will a
single company dominate the whole enterprise SSD market in the next 4-5
years? And if so - is it Fusion-io? (Other names get posed in this type of
question too - especially if they've just been in the news - usually from the
top 5 SSD companies list - but none so often as Fusion-io.)
My answer
is that StorageSearch's long range enterprise SSD architecture model describes
5 different types of SSD role in the enterprise datacenter in the
2016 to
2020
period. And each of these, have in turn, at least 2 different variations based
on internal architecture (big versus small,
new dynasty vs
legacy, COTS
arrays vs proprietary,
skinny vs fat
cache etc - and that's before you get into apps specific software
differentiations). The technology and market adaptations needed in these 5
interconnected silos are so different - that in my view no single
technology or company can realistically anticipate being the market leader in
all of them.
Fusion-io already has a dominant market position in one
part of one of these market silos in flash SSD acceleration at the
application server level - where the performance benefits delivered from
faster storage almost fuse into those which a CPU can get from having extra
cores - hence the name of the company.
My guess is that Fusion-io's
technologies will spread into the CPU engines of all the other SSD
enterprise silos too - in a cache role - even if in those racks it will be a
small percentage of the cost / installed SSD memory capacity. So I think it's
a reasonable assumption that Fusion-io will be one of the top 5 biggest SSD
companies (by revenue) in 2015. That's a trajectory which was hinted at in the
1st quarter of 2009
- when the company first got to the #1 slot in the quarterly top SSD companies
list - and where it's stayed since - having sustained its captivation of the SSD
market's attention by a series of product enhancements, oem partnerships,
comparative petabyte SSD shipment milestones, financial results and an
acquisition.
Each succeeding quarter that Fusion-io remains in the top
10 list - during the coming years (all the poistions in this list are
significant - not just the #1 slot) will simply confirm that if you want to
know who are the most important future companies in the SSD market? - check to
see how they rate in interest with the influential readers of StorageSearch.
main
associated SSD technologies:-
PCIe SSDs,
InfiniBand SSDs and
(recently) auto-tiering /
SSD ASAPs and SSD
software. And while we're on the subject of software -
IO Turbine would have
entered this edition of the top SSD companies in its own right at #16 - if it
hadn't been acquired by Fusion-io. |
|
|
... |
|
| 2....... |
... |
 | |
Up 2 places since the last quarter.
This is Violin's
highest ranking in this list (so far) and an amazing come-back for a company
which a few years ago had appeared to go back into stealth mode and virtually
disappeared. New management, new finance and incremental improvements to an
already hot product architecture may have been some of the factors. The main
reason Violin is here though (above more than 300 other SSD companies tracked
by our readers - and not just the 18 listed here below) is that customers
want high performance, high capacity SSD systems - and recognize price
leadership when they see it.
I don't like their
poster
ads though which - I think - are wasteful and ridiculous. More
discussion and analysis about this and Violin's SSD market milestones etc in
Violin's profile page.
main
associated SSD technologies:-
Rackmount SSDs (FC SAN and
iSCSI). |
|
|
... |
|
| 3....... |
... |
 | |
Same as before.
STEC's best
previous position in these lists was right at the top of the list - the #1
slot - in 2007 Q2.
2012
will open a new chapter in high performance 2.5" SSD marketing - which can
best be described as the "enterprise MLC credibility wars". The
winners will be decided as much by
marketing as by
technology. STEC is a major participant in this debate - which I have described
in more detail in sugaring
MLC for the enterprise.
main associated SSD technologies:-
2.5",
3.5" and smaller
SSDs for the embedded enterprise market (SAS SSDs),
military and
industrial
markets. |
|
| ... |
|
|
| 4....... |
... |
 | |
Down 1 place since the last quarter.
When
SandForce came to market in April 2009 - it got a surge of interest from SSD
makers who wanted to tackle the high performance end of the small form factor
SSD market which had previously been owned by STEC.
Pliant Technology -
which was founded with the very same idea - still hadn't launched any products
at that time - and when it did it - it got swept aside by SandForce partners.
SandForce's shipments growth in 2010 surprised
analysts because
it was so much faster than anyone had wildly predicted. At that stage they
were the only game in town facing dozy marketing from STEC and a Pliant which
in marketing terms acted as though it was still waiting for permission to take
off.
The situation now in Q4 2011 is different. SandForce driven SSDs
still own the biggest slice of this market - but 3 newcomers have also entered
the space with full SSDs or controllers which are catching up in performance.
Overall the market for fast small SSDs will more than double in 2012 - but it
will be harder for SandForce driven SSDs to maintain their very high market
share of that.
main associated SSD technologies:-
flash SSD Controllers |
|
| ... |
|
|
| 5 (joint) |
... |
 | |
Up 4 places since the last quarter.
SanDisk's
best previous position in these lists was right at the top of the list - the #1
slot - in 2007 Q3.
No one can hear
you scream in the vacuous consumer SSD market - where market share in recent
years has been worth about as little as eyeballs in the dotcom bubble. That's
why SanDisk earlier this year acquired Pliant.
StorageSearch.com's
prediction for 2012 is that it be a
very good year indeed for enterprise SSD makers who have good products and
understand which orbit they fit into in the SSD solar system. Will SanDisk be
able to learn how to market enterprise SSDs any better than the company which
it acquired? A lot of people are looking to see what it's going to do in this
space. It's rare for companies rooted in semiconductor culture to succeed in
enterprise system markets. By the time you've digested the inputs from
enterprise customers - your fab is already shipping the next generation product.
SanDisk has to know what it's going to do in years 2 and 3 of its enterprise SSD
roadmap before it's even 1/2 way through year zero.
main associated
SSD technologies:- SAS SSDs,
notebook SSDs,
and (soon) PCIe SSDs
|
|
|
....... |
|
|
| 5 (joint) |
... |
 |
... | |
Down 1 place since the last quarter.
You'd
think that a company which has been actively at the forefront of the fast
enterprise SSD market longer than anyone else might have started running out of
ideas by now or slowing down. But in the last 4 years TMS has shown that it's
not going to be bit flipped aside by changing market fashions in memory
technologies or form factors - although it was indeed almost the last of the
enterprise SSD makers to introduce MLC (naughty flash) into its
product lines this year. Despite that you can still get "good"
flash (SLC) - and the "angelic" memory which doesn't wear out (RAM)
within TMS's
confusingly named
- but respected "RamSan" product line. The RamSan SSDs are
differentiated by density , performance and market application - while retaining
the common unity of always being among the fastest products in their class.
The
secret to this flexibility is that Texas Memory Systems has designed more ASICs
/ FPGAs than any other company in the enterprise SSD market - and under the
tutelage of their founder Holly
Frost - their engineers have the confidence born from a decades long
company culture of doing this fast hardware stuff for mission critical
applications. Will the growing importance of
software in the SSD market
change any of that? The way things work in the storage market is that if you're
a hardware company who needs complicated software - you license it or buy a
company. You rarely have to write it yourself.
main associated SSD
technologies:-the
fastest SSDs, PCIe
SSDs, Rackmount
SSDs |
|
| 7 |
... |
 | |
Down 1 place since the last quarter.
OCZ
has been one of the hardest companies for me to characterize - since they first
entered the top 10 SSD companies list in the
2nd quarter of 2010.
Just when I thought I had the picture - they moved on and did something else.
And just when I thought they would never get their heads around an issue (like
enterprise SSDs or IP) they did something significant which again proved me
wrong. But I think I have a better idea of what OCZ is all about now. Here
goes...
If there is a science to the art of selling and marketing SSDs
- based on the true scientific principle - which is you test an idea - you see
if it works - and if it does you model it and then do some more testing and
gradually get more confident about trying something else - and then learn from
that too - then OCZ is the master of pragmatic SSD marketing.
In the
relatively short span of years in which they've been involved in the SSD market
they've dipped their market thermometers into more SSD niches than any other
company I can think of. And as they've gained confidence from learning what's
hot and what's not they've boldly moved on at a pace of innovation - for example
in the auto-tiering SSD market - which is breathtaking.
Will all their
market experiments work? It's hard to believe they all will - but nevertheless
the company's reaction times for SSD market experimentation and the knowledge
they are building up across a wide swathe of the market means that if there is
any gold in them there SSD treasure hills they will be among the first to find
it and know how to mine it too. And if there isn't any gold in a particular SSD
hill they're nimble enough to get out lightly before digging in too deep.
main
associated SSD technologies:-
PCIe SSDs,
2.5" SSDs,
notebook SSDs
and flash SSD Controllers,
auto-tiering / SSD ASAPs
. |
|
|
....... |
|
| 8....... |
... |
 | |
Up 3 places since the last quarter.
There
was a time - prior to 2005 - when if a flash SSD company announced that its
drives had been qualified for use in a space mission by NASA - that would have
opened the doors to most of the flash SSD market - because the enterprise
applications for flash SSDs were then barely in their infancy.
RAM SSDs ruled the
enterprise roost and "flash SSD" meant "rugged drive".
SSDs from a company called
Memtech did once fly
past Mars (or crash into it) I can't remember which. Anyway when Foremay did
one of these NASA type announcements recently - most of you wouldn't have seen
it or cared - because the rugged embedded market is just a single color in the
flash SSD spectrum. But a single color can still light the way into the top 10
SSD companies list in 2011 - as this listing proves.
main associated
SSD technologies:- embedded SSDs for the
military and
industrial markets
and also PCIe SSDs. |
|
|
....... |
|
| 9....... |
... |
 | |
Down 2 places since the last quarter.
The
EU mergers and monopolies commission has been investigating
WD's
proposed acquisition of Hitachi GST (which the EU regulator said it was
considering in the context of Seagate's proposed acquisition of Samsung's HDD
business). The regulator said a decision will be made before December. So the
industry has to continue waiting to see how WD will leverage the wider
enterprise SSD footprint it will get from integrating products and interface
technologies from Hitachi GST.
main associated SSD technologies:- high
reliability embedded
2.5" SSDs and
notebook SSDs
and (soon - following acquisition of
Hitachi GST)
SAS SSDs and
FC SSDs too. |
|
|
....... |
|
| 10 (joint) |
... |
 | |
Same as before.
In this
quarter RunCore announced it is building the biggest SSD factory in China
to enable the company to cope with the increasing international demand for its
products. The company also entered the
1" / SSD on chip
market.
main associated SSD technologies:-
notebook SSDs,
industrial SSDs
and military SSDs |
|
|
....... |
|
| 10 (joint) |
... |
 | |
Up 4 places since the last quarter.
In
a market as crowded at 2.5"
SSDs - most of which use the same small set of industry standard
SSD controllers inside
- why should anyone care about another new SSD?
SSD specifiers are
learning that there's more to the
SSD reliability mix
than endurance
and data integrity
when the power's on and everything is running smoothly. It's when the
SSD power
slams suddenly off that you find some SSDs fail and others don't - for
detailed design reasons which are specific to each SSD maker.
In
this quarter SMART announced that its Guardian protection technology won a best
of show award at the Flash Memory
Summit .
main associated SSD technologies:-
2.5" SSDs,
SAS SSDs,
military SSDs |
|
|
....... |
|
| 12 ....... |
... |
 | |
Up 5 places since the last quarter.
There's
a growing realization in the enterprise market that RAM SSDs will be a permanent
part of the SSD toolkit in high transaction volume datacenters which have
any kind of traditional hierarchical data architecture. And the only reason for
buying them is to solve problems which flash SSDs create or can't solve
technically because of RAM's superior latency, bandwidth, truly symmetric
IOPS,
better reliability (and sometimes lower floor price). So readers are curious
to know what type of systems sit at the top end of the
fastest SSDs list.
Kove's
Xpress Disk embodies how fast you can
go in an FC SAN /
IB connected RAM
storage appliance - with 8 microsends latency and nearly 30 gigabytes / s
bandwidth. It's good to know that - when you need them - such products are
there.
main associated SSD technologies:-
RAM SSDs |
|
|
....... |
|
| 13 ....... |
... |
 | |
Up 3 places since the last quarter.
This
is the best ranking so far for Virident - which is the last of the serious
enterprise flash PCIe SSD makers to only offer "die-hard" SLC - in
their product line.
Wise product specifiers look beyond such
transient claims as which
is the fastest PCIe SSD? - and are more interested in scalable architecture
and sustainable roadmaps. Virident was the first company in this segment to
talk about maintaining speed even when (storage capacity is) full. But they
aren't the only fast flash SSD company to offer spike-free IOPS
performance.
Considering how noisy the SSD market is with marketing
communications - Virident is quiet. They don't say much and extracting data
from them is harder than it should be. Despite that their Director of Systems
Engineering - Shirish Jamthe
did present a paper in August at the
Flash Memory Summit called -
a
Close Look at PCIe SSDs (pdf) which gives you some idea of their thinking.
main
associated SSD technologies:-
PCIe SSDs |
|
|
....... |
|
| 14 (joint) |
... |
 | |
Up 5 places since the last quarter.
BiTMICRO's
best previous position in these lists was right at the top of the list - the #1
slot - in 2008 Q1
BiTMICRO
has been designing fast high capacity flash SSDs since the 1990s. They were the
first SSD company to show that a single
3.5" FC flash SSD
could speed up a 25,000 user email server - better than an
HDD RAID. That was
in 2004!
Today BiTMICRO has returned to its roots as an SSD innovator in mission
critical industrial and military markets.
main associated SSD
technologies:-
industrial SSDs,
military SSDs |
|
|
....... |
|
| 14 (joint) |
... |
 | |
First appearance in the top SSD companies
list.
This is the first time in the
history of
the SSD market that an ISV has appeared in the top SSD companies list.
FlashSoft would have been joined in that distinction by
IO Turbine - who
would have come in at the very next position down this list if they hadn't
been gobbled up in the qualifying quarter by Fusion-io.
main
associated SSD technologies:-
auto-tiering / SSD ASAPs,
SSD software |
|
|
....... |
|
| 16 (joint) |
... |
 | |
Up 4 places since the last quarter.
When
this PCIe RAM SSD came to market in
May 2009 - it
disproved the theory that RAM SSDs are always more expensive than flash SSDs -
with a price which was then and still is 5x lower than a typical
enterprise PCIe flash SSD. OK - to be fair - the capacity is about 100x
lower too. But in the right environment and used as a cache it can beat the
pants off some well known 2.5" flash SSDs in performance and value for
money - according to the benchmarks shown on DDRdrive's home page.
main
associated SSD technologies:-
RAM SSD,
PCIe SSDs |
|
|
.. |
|
|
| 16 (joint) |
.. |
 |
... | |
First appearance in the top SSD companies
list.
With Nimbus - we have arrived at the point in this list
where it's the integration of a unique home grown software stack (more than 10 elapsed
years in the thinking) which makes the difference.
This software
stack is integrated into an unremarkable hardware architecture which is
conservative - and similar in style to 100+ other open COTS SSD hardware arrays
- and if it was only naked hardware it would be good enough - but with the
software it's attractive.
Attractive enough for eBay to disclose
recently that it is using 100TB of Nimbus SSD storage inside its
infrastructure. And eBay is no newbie to SSD accelerators having publicly
revealed that it was using (different) SSDs to accelerate its online
transactions in our news pages as long ago as
December 2000.
main
associated SSD technologies:-
iSCSI SSDs,
FC SAN SSDs,
IB SSDs |
|
|
.. |
|
| 18 (joint) |
... |
 | |
Same as before.
A year ago
when when I extended the top SSD companies list to 20 companies (instead of
just 10) that sucked EMC up into the list of significant SSD companies.
At
that time (as we've seen since confirmed in reports of
shipped SSD petabytes)
they needed all the help they could get. They were asleep and didn't get the
SSD plot. It looked like they mistakenly believed or arrogantly expected
that a 1990s style RAID architecture wrapped around a bunch of small form
factor SSDs with some error correcting layers of software and management GUIs
was what the world needed to advance to the next stage of the solid state
storage market.
"...There's no reason Fusion-io should've
come out in front of us, and we are catching up..." said EMC's President and COO
Pat
Gelsinger - in an
article
by - SiliconANGLE.com (in August) - re flash in storage arrays.
As
I commented in our pages -
there were in fact many good reasons that EMC had been locked out of the
thinking of anyone looking seriously at the enterprise SSD market. And if EMC
still wasn't awake and not doing anything very useful to progress SSD market
adoption in the enterprise there were plenty of other companies who were willing
to show the way...
In some ways EMC's SSD product line is further
behind the market curve today than is was a few years ago. But as more software
gets added into the enterprise SSD mix - that's where other vendors could start
slowing down. My guess is that EMC will improve its SSD offerings (they can't
get any worse) by acquisition
and licensing. And by the time we get to 2016-2020 EMC will have a place
in medium performance bulk storage SSDs and SSD backup systems. But it will be 1
of 10 or more similar size enterprise SSD companies - and will probably have a
smaller share of the storage market in the future than it does today. Having
said that - the storage market will be much bigger then - so investors
probably won't lose sleep over it.
main associated SSD technologies:-
Rackmount SSDs,
PCIe SSDs, and
auto-tiering / SSD ASAPs
|
|
|
.. |
|
| 18 (joint) |
... |
 | |
Down 3 places since the last quarter.
Considering
what it had to gain (or lose) from the industry transition to solid state
storage - Seagate stood on the sidelines for many years too long - saying
SSD was a silly game, it didn't like the rules and didn't want to play.
So
by the time that Seagate (said it had) changed its mind and wanted to play
after all - the other players didn't really care either way - and it was no
real surprise to most spectators that Seagate wasn't thought good enough
to be picked as team captain.
main associated SSD technologies:-
PCIe SSDs,
2.5" SSDs. |
|
|
.. |
|
| 20 ....... |
... |
 | |
Down 8 places since the last quarter.
Intel
has shipped some truly terrible SSDs - which demonstrated the company's lack
of knowledge about the fundamental building blocks needed to design a reliable
storage drive.
In the summer of 2011 - after yet another SSD /
firmware recall - this led me to comment - "If Intel's SSD design
business was a horse - it would have been shot a long time ago and put out of
its misery..."
Intel's customers have painfully learned that there's a lot more to
designing SSDs than soldering a bunch of memory chips to a controller and host
interface. Intel's designers should have known that too. In the long term I
expect Intel will buy
an SSD company or license more SSD technology.
main associated
SSD technologies:-
notebook SSDs,
2.5" SSDs |
|
2 companies dropped out of the top
20 list in this quarter:-
Pliant (acquired)
and Toshiba.
Companies
hovering just below the the top 20 in this quarter include:- (in alphabetic
order) ACARD,
LSI,
Samsung,
Solid Access
Technologies, SolidFire
and WhipTail Technologies. |
