StorageSearch.com is about thought leadership in the SSD
from - what's in the
- how will SSDs change the
Since the 1990s
our readers have accelerated the growth of the SSD market and determined
its essence and onward direction.
|more SSD topics from A to
1" SSDs, BGA SSDs &
1.8 SSDs", MO-297, slim
2.5" SSDs - all
2.5" PCIe SSDs,
SSDs - all
silos for enterprise SSDs
11 key SSD design
2012 - changes in SSD
changes in SSD market
changes in SSD market
About the publisher
Acquisitions in the SSD
R/W & DSP ECC in SSDs
Advertising SSDs - 15
years of SSD ads
Analysts - SSD market
Animal brands in
and blogs - top 50
Architecture guide - storage
ASAPs / auto tuning,
caching, tiering SSDs
Backup - SSD enhanced
management in flash SSDs
Banner ads for SSDs
Best SSD for your
PC? - sorry I can't say
Big market picture of
versus Small - in SSD architecture
Bookmarks from SSD
- citing StorageSearch.com
Strategies in the SSD market
Buyers Guide to SSDs
Cache ratios inside
software and SSD appliances
Calling for an
end to SSD vs HDD IOPS
Can you tell me the best
way to SSD Street?
- storage interface / processors
Chips - SSD on a chip &
CIO - a
day in the life of
storage - with SSD twists
Consumer SSDs guide
Consumer SSDs - why are they
chips for SSDs
|What defines the
personality of the SSD?
Cost of SSDs
Dark matter users in
Data integrity in
|When your SSD breaks...
do you call?
hidden segments in the enterprise
markets and SSD
DSP in flash
SSD controller IP
ratings in enterprise SSDs today?
|Some are much better than others.
|Education - re SSDs
Efficiency - comparing
SSDs - overview
Endurance - in
|The risk of flash
wear-out in SSDs is
a kind of "forever war"
which is never
|Fast purge / secure erase
SSDs - market timeline
(failures in time) & SSDs
flash SSD vs RAM SSD
Forecasts for SSD market
Hard disk drives
Hard drives in an
- SSD fans disagree fundamentals
History of SSD
of data storage (all)
|Hostage to the
fortunes of SSD
SSD appliances (enterprise)
syndrome - invisible SSD capacity
Inanimate Power, Speed
& Strength Metaphors
|If they survive manufacturing they'll
survive our industrial customers too.
IOPS and flash SSDs
IOPS Comparisons -
SSDs and HDDs
- flash SSD
Legacy vs New Dynasty
|Killerbyte had a "no surrender"
attitude when it came to attacks on her personal data integrity.
MLC - in SSD jargon
MLC in enterprise
era of the SSD market
MRAM, PCM & other
threats to flash in SSDs
- mentions on this site
Oracle articles - Burleson
Petabyte SSD roadmap
protection in SSDs
Power, Speed &
Strength in SSD brands
|Planning a new SSD family?
PR Agencies & SSDs
Price of SSDs - past,
present and why?
RAID - and SSDs
RAM memory chips
RAM SSDs versus Flash
SSD capacity - server vs SAN
Reliability - in SSD
games and enterprise SSD boxes
|Can it guess what's in my next magic
Silos in the solid
SLC vs eMLC
Software (for SSDs)
to enterprise SSDs
Symmetry impacts in
zero storage? - groan...
Top SSD companies -
7 years tracker
UlltraDIMM SSDs etc
propositions for buying SSDs
VC funds in storage
Videos - about SSDs
Where are we now
with SSD software?
SSDs killed 20K hard drives
editor - StorageSearch.com
Just to clarify what I mean.
after publishing the blog you see to the right of this - I knew that some people
people who only read it quickly without understanding the concepts would say
Aha - we know what that means! Just as we suspected
- the number of enterprise SSD systems companies is about to collapse - so we
can expect to see less in the next few quarters than in the past.
is not exactly what I meant.
In the medium term... A reduction? Yes.
in the long term - a 90% reduction / collapse in distinctly different systems
boxes - compared to what we see in the market today (in Q2 2015)? Certainly.
the number of companies participating in the market will get bigger for a short
time - for at least a year - before a steady decline followed by a sharp
That's because some pieces which are needed to lubricate
these gritty changes aren't available yet and where they are - the products are
naive, the marketing is unsophisticated and the business units are under
Aside from the obvious vendors which will immediately
benefit from the changes (such as software companies of the type mentioned in
the article) we can also expect to see new systems companies entering the market
too. (Even as we head towards a converging market.)
the TAM is huge and a changing market creates new opportunities.
one of the many prepublication conversations I had about these concepts
the person I was talking to said he was so inspired that he was going to
launch his own new rackmount SSD systems company instead of doing something
very significant with another vendor. ...read
|In the summer of 1998, I
decided to launch a new web directory, which became StorageSearch.com. |
pondering many ideas, I read a book called "Creating Killer Web Sites".
This book didn't aim to cover direct marketing, and how to get new readers,
however it was EXCELLENT at analysing how you could use visual ideas in your
site. It also introduced the concept of visual metaphors.
I had seen
some of these ideas:- using animals to promote services, some years before in a
book called "Services Marketing" by Christopher H. Lovelock which I
found useful some years earlier...
|How did Mice become|
face of Storage?
now consider this...by
editor - April 21, 2015
90% of the enterprise SSD companies which
you know have no good reasons to survive.
The main reasons they will
hang around a little longer and their ranks may even swell slightly before
the big shrink are bad reasons...
Bad marketing and an inefficient
drivers, mechanisms and routes towards
consolidation in the enterprise SSD market along with some other outrageous and
However you interact with the enterprise SSD ecosystem a
useful reference point for judging the relevance of what you see happening
around you is some kind of implicit background expectation about the size and
structure of the market - which at its simplest level is measured by the number
of distinctly different vendors and major product types and economic value of
the whole market.
As a commentator on the future of the enterprise SSD
market since before the
modern era of SSDs
I've played an active part in setting expectations about the long term growth
prospects in the market - measured by vendor count, significantly different
product types, and revenue ceilings.
We're now nearing a pivotal
point in the enterprise SSD market where the long held assumptions I helped to
encourage (especially how many leading systems vendors there will be in the
market at the same time) are about to be change dramatically.
changes in customer thinking has already begun. But it will take about 5 or
6 years for all the effects to ripple through. That seems like a long way away.
But take it from me - users would make these changes more quickly if they
That's the impression I've got from prepublication
The ideas in this article could be a catalyst to
speed up the urgency of change in user thinking. If you want to skip ahead -
scroll down to the headline - "Now imagine this..."
Why will the enterprise SSD market change from a game played
concurrently on the same wide open range by hundreds of vendors (if you can
be bothered to count them) and which yet still manages to satisfy the
egotistical aspirations of so many self styled winners (can they all be
right?) this beloved wild spirited contact sport of ours (which
afford not to take part in) maybe it's better to ask not just "why"
but also "how" would it ever change into a sedate board
game resembling something more like bridge crossed with chess - in which
every space on the table will be precisely ordered and precious and the player
seats themselves and cushions will have to be reserved years in advance?
me feel a little sad and nostalgic even thinking about it.
As to the
why? Whatever you may think about mergers and acquisitions and suchlike
behavior that's got very little to do with it.
The real pressure for
change comes from 3 sources:-
- changes in user behavior
It's always been true that SSD
users change their behavior as they learn from their own experience in deploying
SSDs. The lessons they learn after 2 or 3 deployment generations help them to
conceptualize for themselves what they can expect to get done better from SSDs.
In earlier phases of the enterprise market (2007 to 2010) it was difficult
for users to disentangle the individual contributory causes of successes and
failures as there were too many changing variables in technology.
it came to performance modeling for examples - users couldn't be certain
how much of the disruptive changes such as memory types (RAM to SLC to MLC
etc), computer architecture (including placement of SSDs in server vs SAN), and
successive software generations (emulation of HDD arrays, caching of legacy
storage, bypassing legacy application stacks) etc could be considered as a
onetime improvement as opposed to being part of a predictable and scalable
activity. Most vendors didn't know either. Or if they thought they did - they
didn't necessarily agree about what had been achieved, the best way to do it or
the direction of future change. And comparisons between vendors were problematic
for users in the early days flash arrays when there were too few products to
growing user awareness from experience of similar things
of the SSD user base has changed.
It has moved away from being a
market in which most users had little or no experience of using mission
critical SSDs towards one in which most users already have had several
generations of deployments with SSDs. Or they can read about others like them
who do have this experience.
Users are applying the lessons they have
learned - the good about the technology along with the bad about the
predictability and reliability of vendor continuity roadmaps.
now realize that in their own self interest they have much to gain from
abstracting the benefits they get away from the diverse feature sets of any
single supplier towards a minimalist set of common must-have features in 2 or 3
distinct SSD box types which will satisfy all their needs while giving them
independence from failed or greedy suppliers.
Web scale companies have
been following these types of abstraction strategies for years as many were
already well down the road to double digit Nth generation SSD deployments. At
their scales their economic options included creating some of the missing links
in software for their own uses and their own special applications needs.
most enterprises - IT is a tool and not their primary business - so the
investment in home grown SSD virtualization and management tools would be nuts -
given the massive degree of risk and changeability in the market. But software
is a part of the solution.
what limits the shape of the future consolidated market?
- the emergence of new software solutions which can satisfy common
management functions in a scalable fashion and in a way which is credible over
multiple product generations.
It's reasonable to ask why isn't there
already a better availability of industry wide useful enterprise software
which can replace and abstract away the mish mash of different chaotic
patches, tools and feature sets which arise in the systems products of SSD box
makers and which are sometimes as different to other products from the same
supplier in different product generations as they are different to those from
The reasons are...
The enterprise SSD
software market started
chaotically with independent parallel developments in multiple places, and
without a common core idea of unification or longevity related roadmaps.
hasn't been going long. The first useful enterprise SSD software products
emerged in 2009.
Most early SSD software vendors didn't accomplish very much before
they got acquired. The SSD makers who acquired them were in most cases
looking for simple tools to make their current product lines easier to sell
rather than as the basis for industry-wide platforms which cut across multiple
product types and generations. There were some exceptions - but there's more to
SSD software than caching and virtualizing different types of SSDs.
recently the entire enterprise SSD market hasn't been big enough to attract the
scale of investments needed to deliver industry-wide abstraction tools. So the
business case for this type of solution has been risky.
such solutions come from?
Some of the solutions may come from the
future work of companies like
Primary Data while
another center of gravity is in the application server type of SSD box and the
mainstream SDS market which includes companies like
Maxta. But the
application server is only one orbit in the enterprise architecture solar
system not the whole big bang.
SDS vendors who want to take
advantage of this consolidation trend will have to extend their data reach to
support other box types (not just more of their own).
entry directions for newcomers into this market include existing box companies -
who may determine that abandoning their original business plans as systems
companies and morphing into software business entities (or spinning off their
software divisions) is a better option than retaining the characteristics of a
combined systems business in which the software side is unable to leverage
from user experiences and revenue contributions from outside the core box
And other directions from which we may expect to see newcomers
enter this aspect of the software market are:- the productization and
repurposing of cloud and webscale software which already exists, and of course
- new start ups. But the
funding of pure play
startups will likely have to be billion dollar plus - because it takes a lot of
software to support enough different products and functions to be useful.
will limit the shape of the future consolidated market? An underlying
architectural pull due to the clumpiness of data gravity within a belly
girdled tightly by cost economics with recognizable internal SSD
organs which are maintained by the heartbeat of software.
kidding. We know exactly what it will look like. We've known for years.
future needs in the enterprise can be satisfied by a small number of distinctly
different SSD products determined by latency and distance from the application
I described the basic product types in my 2012 article -
an introduction to
enterprise SSD silos.
If you haven't read that article - the simple
summary is that there are only about 5 types of SSD box or appliance needed to
satisfy all requirements. (And most organizations can get by using as little as
3 of these.)
- The application server. (That's the server or cluster of SDS servers which
run your app.)
- The SAN box. (That's a storage box which sits outside the apps server).
There are 3 main types of SAN box - which are different by virtue of
latency and the related cost of their raw capacity. The 3 types are:- fast,
fast-enough and archive.
Within every one of these top level box products there is
also the business possibility to differentiate these latency based product
segments by internal SSD controller architecture.
(auto tiering / caching appliance - sometimes integrated in a hybrid and even as
a micro hybrid inside the apps server) is a box or software layer which
intermediates between any 2 vastly different latencies which are in your
There are only
2 types of SSD
controller architecture which are relevant in this market transition
Big controller - in which the base level controller
management assumptions can be sure of working with large numbers of flash
devices (hundreds or thousands). Examples of such products being Violin's VMEM,
IBM's FlashSystem, EMC's Extremio, Skyera's skyHawk, Virident's FlashMax,
Fusion-io's ioMemory etc.
When they can keep pace with memory
advances these controller architectures typically outperform arrays of standard
SSDs. But the difficulties of supporting new memory types (in particular the
has meant that all but one of the big architecture controller makers have given
up the challenge of native support for cheapest memories and instead rely on
more expensive memory - typically some kind of branded eMLC. The market result
has been for most of these controllers to congregate at the faster end of the
speed segments in SSD box space.
Small controller - in which the
base level controller management assumptions can only be sure of working with
small numbers of flash devices - typically in the region from 3 to 10. Examples
of such products include most (but not all) of the SSDs which you would regard
as "standard" commercial off the shelf (COTS) SSDs - having SAS or
The benefits of this approach ffrom the user
perspective is that arrays of such products benefit from "crowd based
intelligence" re advances in flash memory design. This group of SSDs tends
to be among the earliest to be able to deploy cheaper kinds of memory within
The main disadvantage of arrays of small architect
SSDs used to be intrinsically lower utilization
was because the controller is unaware of what is happening in flash chips in
other SSDs which are outside its own reach but within the same array. An
emerging trend in software pioneered by webscale entities has been to switch
off some of the flash management done in the SSD's controller and. by using
custom application specific APIs to manage some key functions from the
applications server. These techniques can reduce the original gaps in
efficiency between small and large controller arrays.
So for each main
box type the core flash array can be either small or large controller type
(standard COTS array or proprietary array if you prefer). No one type appeals
to all users.
And the market has developed with multiple similar
designs competing in each product type (overlaid with many different shapes of
the SSD ASAP intermediation products).
And each of these came with
their own software. These are some of the reasons the market has been
complicated and tolerated (you might say suffered)
many different designs
upto now. And we will still need the same number of distinct box types in the
But even if you allow for the diversity of having 2 or 3
different key suppliers dominating each SSD box type in the future with maybe a
couple of cost differentiated #2, #3 companies in each latency segment - then
by the time we get to 2021 (which is about the time it will take for the new
software to enable consolidation) the total of distinctly different suppliers
in an efficient market will be a lot less companies than we have now.
- boundary analysis meets consolidation
In the discussions I had
with readers to test my ideas about understanding the transitions towards a
consolidated market I talked through some simple thought exercises.
use a simple trick I call
analysis - which helps me understand and predict market behavior in future
markets even when the products being analyzed and the architecture use cases
don't yet exist.
Now imagine this... you're considering a new
alternate supplier of enterprise SSD storage
You're the top level
person who decides what type of SSD boxes your organization buys. Maybe your job
title includes words like "logistics", "systems architect",
and your industry could be banking, cloud infrastructure, broadcast,
manufacturing. Doesn't matter. These conversations go the same way.
talking because you're thinking about switching to a new supplier. Maybe you've
got a strong candidate in mind, or maybe you've got a list of 10 possibles which
you want to refine down. If you've got a strong candidate of one company, I add
some more to make it more like a list.
Why are you looking at
alternatives? You tell me. But there are many possible reasons.
you don't think your current supplier offers the best fit for what you do.
Maybe you like the products you're buying now but anticipate changes in your
supplier which might not be guaranteed to go in the direction which suits you.
Maybe your supplier has been acquired. Maybe they're greedy and bundling in
things you don't value or feel you should be paying for. It doesn't matter.
You've bought lots of SSD boxes before. They're essential for your IT
infrastructure. But you don't want to be exposed to the
whims of the
enterprise SSD next product casino. You want to take control of your own future.
we're looking at a list of 10 well known suppliers of flash arrays. For
simplicity let's call them SAN boxes. (IP SAN or FC SAN doesn't matter. The
point is they're not the servers which run your apps.) They are all used by
sane, sensible enterprise customers - so we're not worried about them not
Probably you've spent days already (or weeks) going through
the features, which types of SSDs are inside the arrays etc, etc.
line them all up again in your mind.
Now imagine that every single one
of these products is actually the same hardware. And when I say the same - I
mean the same. Exactly the same hardware box - just a bunch of COTS SAS SSDs,
or SATA SSDs or array of something else. They are all exactly the same. In one
discussion like this we also replaced Violin and IBM proprietary arrays too.
Now what are the differences you see? And why would you buy from one
vendor rather than another?
At its simplest - the main differences
But I haven't finished yet.
those SSD boxes from all those different vendors?
Let's put them back
to their original state. So the boxes from Pure, or IBM or HP or EMC or Nimbus
or Violin or (the lists I've discussed with real users include a longer cast of
well known characters)... these boxes are all different again.
relief from some of you out there in SSD vendor land.
Now line them up
in your head.
Now imagine that they all come with exactly the same
Now what are the factors which would lead you to
preferentially choose one supplier over another?
It's good to start
thinking about this. But I don't want you to spend too long on it because I
forgot to mention that you also have to pretend - they all use the same hardware
too. Oh and you can get the same services for all of them too - either branded
by the company you buy from - or from a well known third party.
does that leave us- when nearly all the hardware boxes you buy are actually the
same box inside, and all the software and services which you find useful (as
opposed to being thrown into the feature set) are the same too?
don't need hundreds of different products. You never did - it's just that
they looked different because every vendor was doing something very similar but
keeping you captive with their different software, services and marketing.
still need the small double digit number of different products which are
differentiated by latency and distance from the apps server. But you don't need
10 different versions of each one. You only need:- a standard array, a
proprietary array, and some cost differentiated #2 and #3 companies in each
segment. That's it.
The hardest to source missing piece to get us to
that state of the market are (as described above) is software.
say - what about the hardware?
Oh yeah. Well here's the interesting
thing. When I started having these discussions which imagined replacing all the
different SSD boxes with the same box - it was months before
SanDisk launched its
system. BTW The no frills, standard array of COTS SSDs isn't a new idea. But
the market is different now. And the big difference is user education and the
confidence of experienced users to abstract their own needs away from the
plethora of junk which gets attached to many product lines in the deluded idea
that more technical features make a better product.
When I having these
conversations after the InfiniFlash launch it gave me a convenient reference
point - but it made no real difference to how users thought about their future
plans - compared with a "hypothetical" array replacing all the branded
Users I spoke to liked the idea and were already starting to
make plans about a hypothetical white box style of enterprise SSD being in their
futures. That didn't mean they would run out and buy an InfiniFlash box to
replace an EMC or Pure box today - but it did mean that they now regarded all
branded boxes as white boxes with proprietary bits added. And they were actively
engaged at looking how they could use minimal branded feature sets in future
deployments so as to ease their own transitions towards infrastructure in which
the software, and services would be distinct products sources elsewhere - and
not just a mash of bundled things which accidentally came with that product from
And that means, incidentally, that they're not willing
to pay the same high prices to their existing suppliers - because they are
viewing them in a different way.
The enterprise SSD market is now big
enough as an ecosystem to enable software and services vendors to think viable
and sustainable thoughts about creating products which disentangle the bundles
which have been appearing in SSD storage boxes in the past 6 years.
marketing and an inefficient market
Going back to the "imagine
this" scenarios above - the main reason we have so many products in the
market today is that the enterprise flash market has been a very inefficient
market up to now. Most vendors do things badly from the user interest point of
view (and from the viewpoint of their own investors).
Despite that many have done OK as SSD businesses despite being bad
at marketing for the simple reason - they didn't need to be any better to get
where they are now.
Even a badly marketed and suboptimal fit
enterprise SSD deployment could deliver outstanding
propositions for users whose infrastructures mostly comprised
traditional HDD array and servers.
Most enterprise SSD revenue until
recent years has come from SSDs replacing something else. It hasn't been
necessary for SSD box makers to be good at SSD vs SSD marketing wars.
the marketing right from the user perspective is difficult for a disruptive
market like enterprise flash. I've discussed some aspects of this in
hidden segments in the enterprise for rackmount SSDs but that article
describes symptoms rather than causes.
why bad marketing?
marketing is the absence of good marketing. And bad enterprise SSD marketing is
the norm seen nearly everywhere in the market today. It's diiifuclt for me to
pick our the good examples becaue they're intertwined with bad even in the same
The root cause is it takes years to construct a
marketing organization which works well at large scale.
pipeline of activities which need to be done to create the right products,
educate and motivate the people inside your own organization and outside who
have an influence, and factor in all the business impacts on your current
products and competitors and partners is a very complex science. I call it a
science because a lot of research and analysis has been done on what works and
It costs a lot of money to invest in training, and it takes a lot
of time and commitment to make good adaptive marketing work. Hardly
surprising you don't see much evidence of this in our industry.
big vendors in the market today have strategic SSD product lines which were
sourced from outside by
acquisition. The marketers they acquire from small companies may be good at
technical marketing or good enough to manage startups but lack the product
management skills to activate the mammoths which acquired them.
the current start ups in the business lack the marketing skills and the
financial success they need to scale up their ambitions.
know it takes many years before a new enterprise software product becomes safe
and useful to use in big scale rollouts. Now consider that the creation of that
software and the idea of where it would fit in with the ecosystem 5 years after
the version 1.0 launch are all actually a subset of a product management
portfolio which someone should have been thinking about years ago. But of course
that didn't happen.
It never could have happened in our industry.
The next thing in a disruptive market is not just a predictable follow-on from
what happened before. But as the market gets more predictable - it will become
more sensible to make such investments.
how much is better SSD
I often write about the contributions to
efficiency that can be obtained from different types of design tricks.
The same goes for marketing.
In an inefficient market like we've had
until now it didn't matter if a vendor sold a few percent more or less of a
new system. It was easy to get sales and the total market size was small. But
we're moving towards a market in 2020/21 where it will harder to get sales
without being good at marketing and where each 1% difference in outcomes could
be worth around $1 billion of revenue. That will be the reward for those who are
still in the game at that time.
I'm not going to make any prediction
about where the "good marketing" in enterprise SSD will come from. But
one day when the technology differences don't matter anymore and when it's as
easy for users to change their enterprise SSD box infrastructure as it is to
change their PCs the SSD market will be very different.
interesting times ahead though as we try to spot where the enterprise SSD
software decoupling platforms are coming from.
enterprise SSD systems market is entering a phase where easy revenue growth
opportunities which vendors experienced before due in large part to displacing
older technologies will disappear. Future competition will be more like SSD on
The user base will change in the way it views SSD vendors as the
market moves towards an ecosystem in which most user organizations have
experience and confidence in deploying flash and managing it.
result of those experiences the traditional enterprise will adopt some of the
strategies which we have already seen in the earliest big SSD adopters - a
willingness to abstract their own needs and best interests away from dependence
on solution types which they will increasingly view as being variations
around a small group of possible white boxes which vary simply according to
latency and distance from the application - as bundled architecture samples
rather than as architecture types.
Meanwhile the size of the enterprise
SSD ecosystem will encourage the emergence of new types of suppliers who offer
software products which can support the long term vision of users and help
migrate them away from the dependencies they don't feel are worthwhile.
- re InfiniFlash
In the article above I mentioned that I used
SanDisk's InfiniFlash as a convenient example of a white box in some of my
prepublication conversations about this article. And in the months before the
InfiniFlash launch I was just talking about a white box SSD with similar
It didn't make any difference to the willingness of
the users I spoke to when it came to their engagement with the ideas. What
was interesting though was how willing some users were in thinking about
an InfiniFlash style of box in their plans as replacements a branded storage
system. If not now then soon.
The transition strategy would involve
reducing their future dependence on bundled software features in branded systems
- even if all the ingredients aren't available now to move everything to white
box. So they're already thinking of the branded systems as supersets of white
I'm not saying that the InfiniFlash product is a good or bad
It's just a well known example in a trend which will grow.
And if you ask me about how good or bad the marketing of this product
was? I have spoken to various marketers at SanDisk about this aspect.
enterprise marketing has improved a lot in recent years but it's still closer
to the norm which is bad rather than anywhere near approaching good. This is
due to their newness of being in the market and the other common factors I
mentioned above. I could give you a list of other companies which have been in
the enterprise storage market for a very long time and which and are also bad
at marketiing enterprise SSD systems too.
vendors may be unhappy with the ideas expressed in the above article -
especially the dangerous notion that almost everything they do from a technology
point of view and all the treasured things they stuff into their boxes don't
equate to flash array heirlooms which their customers will continue to
Before you send me such emails and complaints let me just
say that adapting to the changing needs of your customers and anticipating what
they need next, and next and after next, is part of what good marketing is
about - and if you were doing it well - you'd be the ones who wrote this
article - not me.
|comments related to the
"Always thoughtful Zsolt. Few have the
credibility to write this" - Woody Hutsell,
this is an insightful article that shows your vast experience and unique
position in the industry" - Morgan Littlewood, Founder
at Kodiak Data
|Hmm... it looks like you're seriously
interested in SSDs. So please bookmark this page and come back again soon. |
|The simplest way for
vendors to signal to the world that they are masters and commanders of the
enterprise flash array high seas - rather than merely floating barges of chips
which can be swept along in any direction by the latest technology gust of wind
- is to hoist new colors of SSD pricing. |
Astrological Age of Enterprise SSD Pricing|
| hold up
capacitors in 2.5" MIL SSDs?|
looking at different aspects of
data integrity schemes in mission critical non volatile memory systems
for over 30 years. |
But every time I revisit this vast topic and
compare fresh examples from the market - I learn something a little bit new.
My new blog -
to three seconds - demonstrates the extreme range of hold up times
now in the market inside leading edge 2.5" military flash SSDs.
|"What is/should be
comforting for your newer readership is your capability to make seemingly
outrageous claims years ahead of them possibly becoming reality...
...these and other predictions that were followed by pipedreams
eventually converting to tangible reality, continually drive the ubiquity of
Flash-enabled systems across Data Centers."
thanking the editor of StorageSearch.com for 12 years of thought leadership
on the subject of enterprise flash. |
|"For reasons which
were obvious to systems architects 10 years ago - and haven't changed today -
you will always get better control of performance and cost by designing a hybrid
storage array with distinctly separate HDDs and SSDs compared to combining both
these functions in a single type of drive."|
|Toshiba's new 3.5"
GbE hybrid drive? (May 18, 2015)|
|"The winners in SSD
software could be as important for data infrastructure as Microsoft was for
PCs, or Oracle was for databases, or Google was for search."|
|all enterprise data
will touch an SSD |
| $1 billion may not be
but $5 billion may be too much.
|VCs & SSDs|
|For many enterprise
Users - the question - can you guess what will my next SSD box will look like?
- bears a striking resemblance to the unfairness of Gollum's question - when he
says - what has it got in its pocketses - Precious?|
|playing the enterprise
SSD box riddle game|
|Even in its infancy -
endurance management was a complicated technical subject - but if we look back
from the perspective from the ultra-complexity of today - it was much easier to
manage and understand. |
|SSD endurance - the
forever war - now in 3D|
|You don't have to
understand the internal details of how these individual techniques work. And
with hundreds of patents already pending in this topic there's a high
probability that the SSD vendor won't give you the details anyway (not even
under NDA). It's enough to get the general idea.|
care management & DSP ECC IP in SSDs|
design of memory controllers, and in particular memory scheduling algorithms,
leads to uncontrolled interference of applications in the memory system"|
|Are you ready to
|Smaller nuances of user
behavior (which are easier to discern as patterns in a stable market) easily get
lost under the noise created by headline technology changes and the market's
apparent willingness to slaughter and discard once loved past industry leaders.
hidden SSD segments in the enterprise|
|"A critical test of
whether you really understand the dynamics of a complex market like enterprise
SSDs - is whether you can predict what rational buyers might do when offered new
product options at the extreme limits of - for example - price."|
Analysis in SSD Market Forecasting|
|Like cosmological dark
matter - the SSD dark matter will be bigger in mass than anything which we can
currently see or foresee.
I often say to enterprise SSD marketers - it's easy to create a list
of the top 10 oems or user sites which already use SSDs - but no one's got more
than a small fraction of the list of future SSD user heavyweights - because they
don't exist yet - or if they do - they're in stealth mode. But they can see us.
|The big market impact of
SSD dark matter|
|Here's the real problem for
the big companies. They can lose tens of billions of dollars of revenue by
not participating in the SSD market. |
|Hostage to the
fortunes of SSD|
|SSD makers are different in
the efficiency of their system designs - even when they have access to exactly
the same pool of chips. |
What's important at the systems level is that
some companies can build the same usable capacity, performance and reliability
for the user's app - even when using 20, 30, 40% and even 50% less chips which
come from the same memory generation as their SSD competitors who have less
efficient architecture and don't have the same technical or market knowledge
| Efficiency as internecine
SSD competitive advantage|
|Skyera is putting a lot of
effort into joining something which looks a lot like an old fashioned English
gentleman's club |
(think- Forsyte Saga or Sherlock Holmes).
we know that the current members of the club are so old they will die soon
So is it worth it?
|unified storage and
gentlemen's clubs (October 2014)|
|True archive SSDs using
switched power management may be able to pay for themselves by saving on
electrical costs, disk replacement and datacenter space - even if the competing
hard drives are free. |
More data is better - not worse. Data volumes
wll expand due to new intelligence driven apps.
The data archive will
be seen as a profit center - instead of a cost overhead. (Provided the data can
be repurposed at competitive latency and cost.)
|this way to the petabyte
SSD (March 2010)|