|rackmount SSDs -
by Zsolt Kerekes,
|I've been reporting on the rackmount SSD
market since the
the main reason that most users look for rackmount SSDs today?
still mostly speed
(IOPS performance and low latency).
But as predicted in my
enterprise SSD market
silos report - a new emerging trend is for fast-enough SSD racks to find an
economic place between the performance levels of HDD arrays and the fast end of
Looking ahead further - yet another
new segment for SSD
racks - which is still to come - will be SSD arrays to implement the
lowest cost bulk / archive / cloud storage at a lower cost than high capacity
hard drive arrays.
There's a lot of complicated stuff going on in this
market at the controller, software, architecture and business levels. These were
explored in a recent home page blog -
directions in rackmount SSDs
rackmount SSD vendors?|
|Editor:- are you
searching for rackmount SSD companies? When the number of companies marketing
rackmount SSDs started heading into the 100+ region I removed the long
dangly vendor list which used to be on this page - because it was becoming
unusable. Instead I suggest using the siet search below - and insert the words "rackmount
SSD" along with another criterion which matters to you - such as iSCSI, FC
SAN, fastest etc.|
The product shown below, from
(which is no longer in business) is an example of a
accelerated SAN router which was
featured here on StorageSearch.com in
|rackmount SSD news|
|DCIG ranks top rackmount
Editor:- March 31, 2014 - If you're interested in
then DCIG has published the
2014-15 Flash Memory Storage Array Buyer's Guide (free sign-up page) -
which provides detailed comments on the strengths and weaknesses of rackmount
SSD systems from 20 different vendors - which are currently available in the
market today (includes list prices).
DCIG have created their own
multi-dimensional scoring system in which they look at component features such
as density (TB/U), software compatibility (for example ease of integration with
VMware), and management functions (dedupe, tiering, snapshots etc). DCIG has
ranked these systems overall - and compared many of them to others in the same
price band. Another useful feature of the report is a background story about the
design heritage or market history of each product.
comments:- I've read the report and think it's a good read with respect to
the raw data and detailed observations about many of the systems listed.
to the product rankings?
I think whether you agree or not -
depends on whether you would assign the same weights to each constituent in the
confidential matrix of factors which DCIG have devised.
users it will reflect your own priorities - for others - the scoring outcome
would be entirely different.
Among the SSD vendors listed in the
report - the happiest will be
Nimbus (who have been
about being #1) - and happy too should be
HP (which is #2).
vendors - whose products are best in class in a particular dimension - don't
score highly in the main list because they lose out on the "sum of all
things which DCIG think you might need" - which is an
judgement - rather than being a universal "goodness" attribute.
only company which is conspicuously absent from DCIG's list (at any rank)
Fusion-io. Does DCIG
know something we don't? That's very odd.
Pure Storage announces rackmount SSD mille-stone
March 11, 2014 - Pure
it has shipped over 1,000 of its Pure FlashArrays (fast enough rackmount
Editor's comments:- in case you didn't get that "mille-stone"
thing. "Mille" is an olde English prefix (from latin) meaning "thousand".
array history context?
Pure Storage's shipments milestone is
less signficant than IBM's
1,500 FlashSystem 840s (fast
rackmount SSDs), but more significant than
Tegile's 1,000 Zebi
storage arrays (hybrids)
- which we have also heard about in this quarter.
wanted! - VARs for Fusion's rackmounts
February 10, 2014 - Fusion-io
that its systems level (PCIe SSD inside) products will be widely available from
VARs in North America in in Spring 2014. Specifically these products include:-
Editor's comments:- It takes
less time to hatch a human baby than it has taken for Fusion-io to make the
transition from first talking about some of these integrated systems to
making them generally available.
Traditionally - SSD systems
companies hid the messy product creation process - and preferred to launch their
new rack babies when they were fully formed and ready to fly off the shelves.
That's in contrast to drive makers who often start revealing what their plans
are - long enough beforehand so that their customers can warm up to the idea.
Fusion's case - what we've been seeing in the past 2-3 quarters is not so much
the development of new rackmount product lines (because all the technology
components already existed before) but what we've witnessed instead is the
growth pangs and development of a new SSD systems business - jostling for
adequate space and recognition within an already confident module and software
business - with much of the thinking about the priorities being done out loud
and visible to the public gaze.
IBM shows off what's it's been doing with the RamSan rackmount
SSD product line it acquired from TMS - and also launches first memory
channel SSD based servers
Editor:- January 16, 2014 - For most of
the previous decade (2000 to 2009) Texas Memory Systems
was THE company which competitors aspired to match in market position when it
came to fast
In the early part of this decade (2010 to
2012) TMS lost
its monopoly on rackmounts as it inevitably had to share the expanding market
with a lot of other companies - starting with
Violin (which overtook
in 2011) and then
other companies like WhipTail,
Skyera which had all
established strong market recognition by the end of
in those latter years (from
not only was TMS competing against all those newbie rackmount vendors - but
it was also engaged in another hotly contested
part of the enterprise
SSD market in fast PCIe
SSDs - where its product line was trying to find a place somewhere in the
narrowing gaps between Fusion-io
Then a year ago - in January 2013 -
IBM completed the
acquisition of TMS (which had been announced in
August 2012) and
since then we haven't heard much about these products apart from a few
glimpses - which enabled us to observe that TMS's rackmount products had been
retained and renamed - while their PCIe products were quietly end of lifed.
week - among other things - IBM has launched a new fast rackmount SSD family -
architecture is effectively an enhanced adaptation of TMS's 8th
generation RamSan with some tweaks to incorporate newer memory, iron
out some RAS wrinkles (you can now change everything inside from the front or
back - without sliding the rack out) and a big investment to present a
software friendly face. The new software capabilities are being done by
products which are being offered as external-to-the-box unbundled
subsystems (control enclosures) for those who want them. This means that the
performance and efficiency of the raw flash array isn't compromised in any way.
new SSD box (a 2U
FC fast rackmount SSD with
upto 48TB usable capacity,
priced at $683K
approx list) is called the
Earlier this week I spent an hour talking about
this new product with Woody Hutsell
and Levi Norman -
who are both now back in the IBM branded TMS fold having both sampled the
delights of some other leading SSD companies in recent years. Woody wrote about
in a recent blog.
As I've known both of them for many years - I
couldn't help but start by saying - "This feels like one of those movies -
where they decide to make a sequel many years after - but all the actors look
much older. It's lucky for us this conversation isn't going out on YouTube."
can get a flavor of what IBM thinks it's doing with this new product - and more
details in its
document (pdf) - and I won't repeat much of that detail here.
said "It's interesting to me how much attention the flash operation is
getting within IBM's storage organization."
He went on to say
that IBM's big commitments to flash such as the $1 billion investment
announced last April
are seen within IBM as popular actions "which are important as we need to
compete." As a result - many competent people (in IBM) want to be a part
of the flashsystems effort.
Anther change in scale since TMS became
part of IBM is that the size of the development team for the flash systems
rackmount has quadrupled.
Sales are good too. IBM has shipped
over 1,500 of these flashsystems. In effect Woody said this was
limited by the fact that for 3 quarters IBM shipped everything they had planned
Woody said he thought that this alone - even without all the
other SSDs which IBM was selling into the enterprise market meant that IBM was
probably on its way to be one of the biggest vendors in the market.
said - a dominant market share in enterprise flash in 2014 might look like 5
or 10 per cent as there are hundreds of companies in the market. - We'll have
to see how things work
But my guess is that with a few assumptions about density,
channels etc this means this rackmount IBM product line has possibly been
generating about $500 million of revenue in the past year - which explains where
some of the revenue missing from competitors' reports may have gone to.
else which appeared in the briefing paper singing the praises of IBM's
expanding universe of enterprise flash product offerings - eXFlash DIMMs -
sounded to me like just another name for
to be the case) which appeared in another announcement
server announcements today - see footnotes for more.
final take on this? (FlashSystem 840 announcement)
IBM is now the
company to make comparisons with if you're looking for fast rackmount SSDs with
some high availability options. Particularly if you're working in a complex
environment - are a big customer and think you will be reassured by the
availability of compatible products and pre sales technical sales support.
density - in terms of rack units needed to build a
petabyte SSD - is
better than some other fast systems - but remains an order of magnitude less
Skyera - due to the
difference between IBM's use of eMLC compared to Skyera's claimed ability to
use TLC due to
controller architecture - which is 2 generations (4 years) ahead of what's
used in this particular IBM box. (Having said that - IBM does already use some
degree of adaptive flash SSD technology in other systems - by virtue of the
SSDs it designed in from
back to scary Skyera
- "On the other hand" - I said to Woody - "Skyera doesn't have
the same HA
or software in place
yet. But not everyone needs all these features."
Overall - for
competitors in the same high performance and reliability class as this new IBM
box (which includes companies like Violin,
Fusion-io etc) - IBM
can still be beaten on
price. It was ever
Footnotes - IBM's first memory channel SSD servers
SSD announcement today (alluded to above) about its new server
architecture which leverages
SSDs - and making a comparison with
PCIe SSDs - IBM said -
"Our evaluators are seeing 5-10 microseconds write latency for eXFlash
DIMMs in preliminary testing vs. 15-19 microseconds latency for PCIe-based flash
storage from Fusion-io,
Virident, and 65
microseconds latency for
Intel S3500 and S3700
We've seen increasing granularity of detail emerge about
the system characteristics of memory channel SSDs emerging in a trickle of
announcements, and experimental user reports in the past year. Now that the new
flash DIMM SSD products are becoming generally available - there will soon be
better clarity on real world costs and performance.
heads up alert for a future article about rackmount SSDs
January 15, 2014 -
hidden segments in the enterprise for rackmount SSDs is the working title
of a future article which will appear here on StorageSearch.com on
or shortly before February 14.
Segmentation is an important marketing
concept which enables vendors to operate more efficiently by identifying
discrete sets of customers within contextual frameworks - who (for specific
stated purposes) can be modeled as having similar needs. But many
significant user adoption realities still remain conspicuously absent in
enterprise SSD marketing plans. ...more about
enterprise flash storage - the 1st decade
December 11, 2013 - as we approach the end of another year - it's interesting
to note that this also marks the end of the 1st decade of flash storage arrays
in the enterprise.
This decade was an unruly period in
- accompanied from the very beginning by early doubts that flash could ever
be as reliable and
- every few years - the reliability debate was re-opened by the desire to use
newer, cheaper and intrinsically less reliable types of
flash memory to increase
competitiveness - which relied on ever more complicated internal architectures
to manage the growing raw technology deficits.
Most other IT related
markets would have reached some kind of predictable calm by now. But there's
no sign of that here in the
SSD market. Instead -
all the debates and architectural upheavels which accompanied enterprise flash
before now seem quaintly simple by modern standards. To get an idea of what
happened see my classic article -
Flash for the Enterprise".
Violin's 1st quarterly report
Editor:- November 25,
2013 - Now that Violin
is a public company - it has to publiicly report revenue etc - and in its
for the quarter ended October 31 - revenue was $28 million - which the company
says was 37% more than a year ago.
"Enterprise data center
storage is in the early stages of a major transformation to an Enterprise Memory
based infrastructure, and Violin is at the forefront of accelerating this
transformation," said Don Basile,
CEO of Violin Memory.
Editor's comments:- one of the
self-awareness repositioning adjustments which Violin has still yet to make -
is that compared to the time when the company entered the SSD market 6 years ago - when there were
only 59 companies in the whole SSD market - and only
2 of the top 10
companies at that time (including Violin) were marketing
rackmount SSDs -
we're now in a market where the number of SSD companies is 10x bigger.
I could easily give you a list of over 100 vendors today which
rackmount SSDs or
hybrids and - in the most recent quarter -
7 of the top 10 SSD
companies were marketing rackmount SSDs within their product line.
of which means - there is a much bigger market which Violin could get - if it
were not for the pesky inconvenience of all those other competitors.
- when the company says that the enterprise market is still at the "early
stages" of adopting SSD systems - in one way I'd agree with that - insofar
as the upside potential for this whole market in revenue is more than 10x more
than it is today - but on the other hand - while Violin has got good
recognition and has established a track record in some segments of the
enterprise SSD market - for example in
and fast SSDs - I
don't rate it as a strong runner in the race for some of the
technology developments which are emerging in the rackmount market.
- the segments in which Violin is already arguably a leader - are big enough.
I often say to CEOs in enterprise SSD companies that an important
business strategy for surviving to play in the bigger SSD market of the future
- is to get better at what they do now - and to focus on just one or two
strategic product types which they can execute well and be the #1 or #2 -
rather than making the mistake of enlarging their product lines - and being an
"also compete" company which is never quite good enough to get the
business - because it operates in too many diverse application segments in
which it can't possibly be the best solution.
what are others saying
You may decide that my ratios are too timid - I said to Skyera's
CEO - Rado Danilak.
If so - scare us!
22, 2013 - I spoke recently to Skyera's CEO about
SSD utilization efficiencies and the problem of being a newbie vendor in the
enterprise SSD market. You can see more in the article -
New report from the Spur Group tracks "share of voice"
in rackmount SSD market
Editor:- October 16, 2013 - The Spur Group
a new market report service -
Share of Partner
Voice: Storage and SSD ($4,995) which compare the sales channels for 4
encumbant rackmount storage platform vendors:
Dell against the channel
efforts of selected rackmount
SSD suppliers including:-
Whiptail and others.
Spur Group says its "Share of Partner Voice" web based methodology
is rooted in "extracting website content from over 76,000 technology
service providers, hosting companies, software developers and other customer
facing technology suppliers through a proprietary crawler."
Ross Brown, senior principal
and the lead author of the report, said - "Our approach to measuring Share
of Partner Voice® shows that Violin Memory is outpacing the market, with
more than twice the brand presence of its nearest competitor. The report also
highlights why the market for
companies is accelerating, as the channel adoption is happening very quickly and
often at the expense of incumbent platforms."
comments:- As you know I'm a great believer in the value of using raw web
data to get insights into fast changing emerging markets - which is why I
launched the Top SSD
Companies series - 6½ years ago.
The raw data types and
methodologies which The
Spur Group use are completely different to ours. They look at different
types of web activity and are based on different population samples - so the
results and inferences are going to have different use cases too.
growing ecosystem for storage
and SSD market data companies is a healthy sign that more economic
activity in future will become SSD-centric.
But a bigger SSD market
means bigger risks too.
This is a complex and innovative market - so
it's very easy to miss key products, companies and trends in the SSD market
compared to the comparative ease with which you can assess what's happening in
the "no great surprises here" technology museum segments
elsewhere in the enterprise.
Cisco to acquire WhipTail
Editor:- September 10, 2013
- Cisco today
announced its intent to
acquire privately owned WhipTail for
approximately $415 million in cash and retention-based incentives.
are focused on providing a converged infrastructure including compute, network
and high performance solid state that will help address our customers'
requirements for next-generation computing environments," said Paul Perez, VP
and GM, Cisco Computing Systems Product Group.
as we learned in June - WhipTail has designed an elegant and scalable software
and hardware architecture to
solve the ever
changing SSD rackmount riddle game - which can compete in cost sensitive
iSCSI environments or
hold its head up among the
installations while also offering resilient
WhipTail entered the
Top 20 SSD companies list
in Q4 2011 and has
been in the Top 10 part of these lists in every quarter since
new products from Pure Storage, the significance of 38KB, and a
door opens for future government business
Editor:- May 29, 2013 -
new models in it rackmount SSD family of non-disruptively upgradeable HA storage
which double the performance of the 2U controllers to 400K 8K IOPS and
provides 12TB of raw storage in 2U. In a related
the company said - the average size of I/O requests its customers see
is 38KB - which is why Pure Storage has stopped using 4KB IOPS metrics
in its sales literature.
Editor's comments:- The rack density
of Pure Storage's new
(approx 10TB usable/ U for a 10U stack using a mix of 2U controller shelves
which manage 2U storage racks which each have upto 24x 512GB
SAS SSDs inside) is
low by comparison to industry leaders.
From a business point of view
this means their architecture may not be the first choice for customers looking
for huge installations of SSD (scaling to hundreds of
petabytes) or who
have datacenters in expensive city locations where increasing the square
footage of storage or server cabinet space is simply not an option - and for
whom higher density (SSD TB/U) outweighs any hypothetical considerations such
as lower cost per terabyte to buy.
The company says it has shipped
hundreds of units.
How you interpret that will depend on your
perspective. It's small by comparison to competitors which have been in the
market longer - but also indicates that the product works and the company is
capable of developing business in the difficult to reach smaller customers who
were orphaned by the earlier stages of enterprise SSD adoption which focused
more on SSD-CPU equivalency as the economic justification rather than simply
displacing HDD arrays. Pure Storage's business model is the harder nut to crack
- but is a bigger market opportunity - which has no clear leaders at the
In another interesting
today Pure Storage today said it is receiving an investment from
In-Q-Tel (IQT), the investment firm that
identifies innovative technology solutions to support the missions of the US
Intelligence Community. The partnership will allow Pure Storage to further
develop its FlashArray technology to meet the unique needs of IQT's government
This is just speculation on my part - but one of the big
headaches for users with sensitive info on hard drives is the cost and secure
logistics associated with
when storage drives reach end of service life.
The only trustworthy
way to deal with hard
drives is to physically shred the drives into little pieces. SSDs present
different challenges for this end of the data lifecycle - but also offer more
technical solutions - such as built-in
- but that adds to the cost of the drives. Within the context of an SSD rack -
it should be possible to integrate reliable software based data shredding -
which would save costs
for customers who factor in these considerations.
exciting new directions in rackmount SSDs
May 29, 2013 - If you're an enterprise user who is already sold on the idea of
using more SSDs - what could be better than a great new SSD drive?
you're an SSD vendor looking for the magic formula to open up vast new
untapped markets for SSDs - what kind of solution do you need to offer to
attract enterprises who aren't at the sharp end of the performance pain
curve, are content with the speed they get from HDDs and who aren't even
looking at SSDs for their network storage?
This is a problem which has
been occupying the SSD industry's smartest product architects for years. And
their answer to both questions is the same - although the product details vary
according to the target market. - It's a new type of SSD box.
generation of enterprise SSD rackmounts is breaking all the rules which
previously constrained price, performance and reliability. The sum impact of
cleverly designed SSD arrays is systems which are many times more competitive
than you would imagine from any tear-down analysis of the parts.
new SSD folksy wisdom - "you can't second guess an enterprise flash array
from knowing what drives are in it" - may soon have to join the - "you
no longer judge
an SSD from simply knowing its memory".
Anyway - to get back to
my headline today - the new math of this new reality SSD box trend - what's
behind it, where it's going, and some of the vendors driving it - are explored
in my recent home page blog on StorageSearch.com
directions in rackmount SSDs. ...read the
Kaminario drops PCIe and turns to SAS to get costs down in new
Editor:- April 18, 2013 - "You don't have to be
an investment bank like JP Morgan to afford our style of fast, scalable high
availability SSD systems any more" - was the key message I got talking to
VP Business Development at Kaminario earlier
this week when discussing with me aspects of the company's newest series of
FC SAN compatible SSD
arrays - the
K2 v4 (6TB usable per U at a cost of $10K to $15K per TB) which was
Phil was referring to the expectation that their products -
which in the first generation were entirely
RAM based SSDs - and
then moved onto RAM / flash hybrids and then mostly pure flash (the flash
components being implemented in the previous generation of K2's by
Fusion-io's PCIe SSDs
- a relationship direction which I suggested in a much earlier briefing
conversation with Kaminario's CEO few years ago BTW ) - had acquired a
reputation of being out of reach pricewise - and not just in a class of their
own for resilience and
of the ways that Kaminario has pulled off the affordability trick is to drop
PCIe SSDs as the internal flash components and use instead
said before that in the enterprise arrays space - "SAS is the new SATA"
- because there are so many companies which have moved into this segment
that there's stiff competition. Unlike the PCIe SSD market -which is mostly sold
on high performance - the SAS market includes a number of vendors who have been
R/W ECC to enable them to use cheap flash to build reliable
Because Kaminario still has a lot of
RAM cache in
its server based architecture - it doesn't need the raw
and performance of
FIO's ioMemory to deliver multi-gigabyte throughput at the rack level. And
another factor is that Fusion-io itself is on course to become a significant
supplier of rackmount SSDs (although not aimed at the same kind of customers.)
Kaminario didn't want to say which SAS product they're using. They
might say later. But it doesn't really matter.
The K2 v4 also
demonstrates that the key IP component in Kaminario's box is SSD software.
When I suggested that future boxes could equally well discard SAS SSDs if
2.5" PCIe SSDs
offered a better set of characteristics - Phil agreed that the company wasn't
tied to any particular internal SSD drive form factor or interface.
has paid Taneja Group
to do some new testing on the performance aspects of simulated hard faults.
These will be very useful for customers - and take the uncertainty out of the
picture - giving hard numbers for various scenarios.
For example - when
running at just under 200K
5GB/s throughput - an entire node (controller) was removed to simulate a fault.
I/O resumed after 23 seconds and performance dropped by less than 15% for 2
minutes before recovering fully.
Nimbus brings flash SMART plus stats to SSD rackmounts
March 25, 2013 - Nimbus
Data Systems today
new software APIs which support its proprietary
HALO OS based family
of rackmount SSDs
- and report on hundreds of real-time and historical metrics such as:-
flash endurance, capacity utilization, latency, power consumption, deduplication
rates, and overall system health. Another new feature is that sys admins can
monitor their Nimbus
SSD arrays via new apps on Android / Apple phones and tablets.
CEO and founder of Nimbus Data said the new software framework would enable
cloud architects and enterprise customers to gain greater insight into their
flash storage by viewing internal aspects of their flash storage which
mattered to them - rather than simply relying on benchmark indicators which
have been cherry picked by vendors or reviewers
EMC samples XtremIO flash arrays
Editor:- March 5,
2013 - EMC today
new models of PCIe
SSDs which the company claims offer nearly 60% better
TCO than (unnamed
competitors) due to new levels of
half - height, half - length PCIe SSDs are currently available in
2.2TB, while SLC models upto 1.4TB will ship in the 2nd quarter.
also said it's sampling flash
arrays which are designed and
using the big
SSD controller architecture based on leveraging IP from its acquisition
comments:- the industry has been anticipating flash SSDs which use
XtremIO's RAID busting
Details are sketchy right now - but the
from throwing away the old drive array design rulebook and starting again with
a flash foundation while at the same time having control of the complete
SSD software stack can
be impressive - as I learned last year talking to Rado Danilak
CEO of another leading company taking this approach -
we expect EMC's array pricing to come down to Skyera levels?
never happen - because EMC's business carries the legacy burden of too many
hard drives and too many old suits.
But what we could see
instead - is EMC's flash arrays coming down to a price point where the
customer pain is low enough to delay many of them from switching away to
other flash. Which means EMC could still have a future in the
|Pure Storage tells CNBC
about enterprise flash|
|Editor:- June 3, 2013 - "You're not the only
company that plays in this market - correct?" - said Becky Quick, co-anchor of
CNBC's squawkbox in a leading
question to Scott
Dietzen, CEO of Pure Storage
"We weren't the first to package flash for
storage" said Dietzen. "What we did figure out uniquely was
how to get flash into a place where it was price competitive with (enterprise)
Editor's comments:- "uniquely"
must have another meaning I haven't learned yet.