|
|
| .. |
|
|
| . |
1/3 of Micron's nand flash trade sales go
into SSDs
Editor:- December 20, 2012 - Micron today
announced
that revenues from sales of its NAND flash products were 4% lower in the
quarter ended November 29, 2012 than they had been a year ago.
Sales
volume of the company's nand flash decreased by 9% - but average selling
prices increased 5%. Overall Micrion reported a net loss in the quarter of $275
million on sales of $1.8 billion.
In a
conference
call Micron said that SSD shipments had grown 20% compared to the previous
quarter. SSDs are 17% of Micron's nand business and the company estimates that
35% of the nand flash it supplies to trade customers end up in SSDs. MLC
was about 80% to 85% of nand flash wafer production with SLC and TLC making up
the rest.
See later:-
an SSD
guide to semiconductor memory boom-bust cycles
OCZ starts beta testing SSD cached Linux accelerator
Editor:-
December 17, 2012 - OCZ
today
has begun a
beta test program for its Linux Acceleration (LXL) cache software which the
company says can intelligently reduce external traffic by upto 90%.
LXL
- which supports the company's own enterprise SSDs across a range of
interfaces - splits the caching policy engine between the core mechanism
running in the Linux kernel and an advanced statistical out-of-band analysis
module that optimally determines which data needs to be placed in the local SSD
cache. General availability is anticipated to be Q2 next year. See
also:- SSD caching /
tiering
Proton's DSP flash IP now available on eASIC for enterprise SSD
controller designs
Editor:- December 17, 2012 - Proton Digital Systems
today
announced
the immediate availability of an
LDPC
( Low Density Parity Check) NAND FLASH read channel for enterprise storage
applications compatible with implementation using
eASIC which
enables enterprise storage vendors to double the throughput performance at
approximately half the power that can be achieved using state-of–the art
FPGAs.
The Proton Digital Systems LDPC read channel enables
enterprise FLASH storage system companies to leverage low cost MLC flash devices
and increase its
longevity to 45,000 program/erase cycles, compared to only 5,000
program/erase cycles with
traditional
BCH algorithms.
"We were keen to work with eASIC as we are
increasingly seeing eASIC devices being selected as platforms for enterprise
grade customized flash controllers," said Dr. Andrei Vityaev, CEO
at Proton Digital. "In enterprise storage systems, production volumes are
often not high enough and the market changes are too dynamic to justify
cell-based ASICs but performance and low power requirements are beyond the
capability of FPGAs. This makes an eASIC
flash controller
solution ideal for this space."
Editor's comments:- this
adaptive
flash DSP technology enables oems to do the kind of things which
SMART and
STEC already do in
SAS SSDs, and which
Skyera does in its
rackmounts. This type of technology will become essential for fast-enough SSD
makers to remain efficient
and competitive in the next few years. The only other game in town for
licensing something similar right now - is
DensBits.
SSD revenue data from - Forward Insights
Editor:-
December 17, 2012 - I've lost track of how many new
SSD reports and
updates have been announced recently by Forward Insights
- but one of them - SSD Insights Q4/12: Client Down, Enterprise Up - includes
data and revenue forecasts for the enterprise SSD market.
Author Gregory Wong told me that
his estimate for enterprise SSD revenue in 2012 - which includes enterprise
drives and modules (SAS, SATA and PCIe) but excludes
rackmount systems
and therefore also excludes proprietary SSDs built for use within racks from
companies like Violin
and Texas Memory Systems
- is $2.9 billion.
Samsung acquires an SSD software company
Editor:-
December 17, 2012 - SSD
software sells more SSDs. And that idea has been behind many of the
SSD company
acquisitions in the past
year.
On
Friday Samsung
announced
that it has acquired
NVELO whose
SSD ASAPs (caching)
technology is designed for the
notebook SSD
market.
are you happy with how enterprise SSD benchmarks relate to
real-life? - user survey results
Editor:- December 14, 2012 -
Virident Systems
recently
published
the results of a survey of 145 enterprise IT professionals regarding their
requirements for flash-storage purchases. ...click
to see graphical summary (pdf)
Among other things:-50% said
they have difficulty finding benchmarking tools to accurately evaluate
flash-storage performance.
Of these respondents, 41% find it
difficult to recreate close representations of real-world application
environments, 39% indicate existing benchmark tools do not accurately represent
real-world application workloads, and 20% do not believe benchmark tools exist.
Editor's
comments:- 4 years ago I published an article -
Can you trust flash
SSD specs & benchmarks? - and I've updated it from time to time over the
years - so it's become a diary of market responses to this issue of benchmark
authority.
I know from my conversations with Virident that their (not
so) hidden agenda with performance benchmarks is that their product is benchmark
agnostic and maintains good
performance and age
symmetry across a wide range of workloads you can throw at it. That's why
they would like the industry to adopt a set of more revealing benchmarks which
they have developed. But you can understand the resistance to that idea. I
wrote about the problem that different SSDs look best in different benchmarks in
- the 3 fastest
PCIe SSDs list or is it really lists?
I think this problem is going
to stay with us for many more years. That's because - even if every SSD
vendor agreed that a given benchmark was the most valid today - there are many
new technology and architectural features creeping into SSDs which could
introduce new weaknesses and vulnerabilities which don't show up on current
tests.
And that's before you even introduce apps and
software related
optimizations. Customers may say - I don't care if this SSD's performance drops
off in an application which I don't have - because it's the best in the app I
use on most.
It always comes back to better
SSD user education
- and the ability to pick out the points which are salient to your situation and
ignore stuff which may be more important to someone else. Even if it's good for
you - not everyone likes broccoli.
SanDisk invests in WhipTail
Editor:- December 13,
2012 - WhipTail
today
announced
it has secured $31 million series C funding from a group of investors which
includes:- SanDisk
(as part of an ongoing strategic priorities
investment
initiative), an unnamed "Silicon Valley industry titan" and some
other (named)
private equity companies.
CRN's cool SSDs list
Editor:- December 13, 2012 -
The
10 Coolest SSD Products of 2012
is a new pictorial article in CRN.
Listed
in alphabetic order there are a couple of random / odd selections here.
LSI pins hopes for SSD controller boom on ultrabooks
Editor:-
December 13, 2012 - An
article today in
DigiTimes relays LSI's views which
anticipate better
SSD adoption in
ultrabooks next year and - if that really happens - more business for the
consumer facing controllers within its
SandForce processor
range.
Editor's comments:- within the enterprise SSD market -
the choice of viable competitive options of
controllers and IP is
much greater now than it was 3 years ago when
SandForce rocked and
seemed to take over this part of the market.
As we approach 2013
anyone looking for a fast
or fast-enough enterprise SSD component - whether it's
SAS,
PCIe or
SATA - now has 4 or 5
strong contenders to choose from - which don't have SF inside. So even
if the enterprise SSD market doubles - the cake is being sliced in more ways.
This analysis isn't new. I said something very similar on these pages
last year - when the number of competitors in each category at that time was
closer to 2-3.
On the other hand if you consider any of the
projections for continuing enterprise SSD growth - fighting for control of
the cake slicer is really worthwhile.
who are the ideal customers for BiTMICRO 's new (but late to
market) Talino based maxIO PCIe SSDs?
Editor:- December 12, 2012
- BiTMICRO
today
said
it's begun manufacturing its pre-announced
maxIO
(PCIe SSDs) which use
the company's Talino (means "talented",
big architecture,
SSD controllers).
The first product in this range will be a full height, half-length
400K random
IOPS (4KB), 4.5TB (eMLC) PCIe SSD.
Editor's
comments:- A few weeks ago I asked BiTMICRO's new VP of Marketing Zophar Sante
(who among other things founded
SANRAD) what the
company's plans were for leveraging the Talino controller - because it had
seemed to me that BiTMICRO didn't have the resources to exploit this technology
in all the markets where it might be applicable.
Zoph said BiTMICRO
plans to offer the chip set to other companies in what would be in effect -
a "BiTMICRO onBOARD" kind of program.
The company's
target market for the maxIO product line includes what I call
dark matter enterprise
end users in addition to the usual enterprise channels such as oems.
Architecturally
the closest product to the maxIO is the
RamSan from
Texas Memory Systems -
which is now part of IBM. Closest insofar as it's got one or more ultra fast
central controllers which expand via data movers / intelligent I/O peripherals
which fan out to more flash.
One of the problems I see for BiTMICRO
though is that - because it's so many years late coming into the enterprise
PCIe SSD market - it will have to focus its efforts on a very restricted set
of technical customers who want to better the performance they can get from
arrays of small architecture PCIe SSDs (like those in established PCIe SSD
products from OCZ and
LSI) and who prefer to
use their own software
and who don't want to get tied into
Fusion-io's seductive
APIs - and who also like the idea of raw brute peak performance rather than
the predictable rounded performance of
Virident and
customers who maybe would have been satisfied with a RamSan type of product -
but who either see themselves as competitors to IBM or want to retain the
option of buying their critical supplier (BiTMICRO) or having large share of
mind of that supplier if the maxIO product becomes strategic in the customer's
own business.
SSDs are the hottest topic in enterprise IT
Editor:-
December 12, 2012 - IT Brand Pulse
has published results of its
Q4
mini survey of hot topics in enterprise IT (pdf) - which includes - among
other things:-
- the hottest IT company in 2012?
#1 - Amazon
#2 - Fusion-io
- the most game changing IT technology of 2012?
#1 - SSD
#2 -
CloudStack For the other lists - such as the individuals who have
contributed most to enterprise IT in the past 20 years - click on the link to
the free report.
Editor's comments:- one of the signs of the
SSD market growing into a seriously big business has been the growing number
and diversity of market data services which analyze different aspects of
this market. SSDs are getting into a lot of conversations - even when the
original topic - as in this case - starts out being something else.
Although IT Brand Pulse hasn't been focused on SSDs for very long - Frank Berry and
his team have done a lot of
SSD related
surveys and business reports in the past few quarters. Their style of
report can help its users have more confidence when making difficult decisions
in a chaotic market where perfect information doesn't exist or would be
obsolete by the time it's collected. See also:-
market research,
vintage SSD analysts
PS
- Some of you are going to think to yourselves - "I already knew that SSDs
were a hot topic in the enterprise many moons ago."
And I would
say - Amen to that- and (some of you) please take note - as we approach the
festive season of new marketing budgets - that next month will mark the 14th
year of StorageSearch.com's
publisher selling enterprise
SSD advertising and the
22nd year of publishing enterprise IT guides.
But going back to the
headline - isn't it nice to see that other people see the talent in our
village's children
too?
Violin names new VP of Sales
Editor:- December 11,
2012 - ""The
datacenter of tomorrow
will be much leaner and (more)
efficient than what
you see today" - said Richard Dyke,
Violin's new
senior VP of North American sales in an
announcement
yesterday about his appointment.
SSD HDD ratios in big data? - blog by Panasas
Editor:-
December 11, 2012 - "An important piece of our research involved
determining how much SSD capacity customers would need...
" and
whether it would make a big enough difference in system performance to be worth
the incremental cost of
including SSD storage in the system. To do this, we extracted key data from
production file systems in the field..." - says
Geoffrey Noer,
Sr. Director of Product Marketing at Panasas in his recent
blog SSDs
and parallel storage (part 3) - in which he also says that in HPC
workloads users can get good enough results by using as little as 1.5%
ratio of SSD to HDD - compared to (read his blog if you want to see the exact
ratio) needed for financial and other markets.
Editor's
comments:- Although I was told about the earlier episodes in this blog
series - which includes an
introduction
to SSDs - I didn't mention them before - because I assume if you
don't already have a good idea of
what's an SSD?
before you get to these pages - then you won't linger here very long.
Another
reason for my earlier reticence about these Panasas blogs - is I think that
a description of SSDs written from the viewpoint of an
SSD ASAP company
whose boxes average over 90% hard drives - may be subtly tilted to a
perspective which I consider to be a sideline along the
enterprise SSD adoption
road rather than the straightest way to the final destination.
Nevertheless
- to give credit where it's due - some of the other past blogs I've seen from
Panasas have included valuable insights which are storage media agnostic.
I've said to a couple of readers recently (including a Panasas
customer) - that if I had to make a list of the last 5 companies on this planet
which I thought would still be shipping hard drives in arrays into the
enterprise - then Panasas would be one of them. That's because a small group of
companies have invested enough talent into new ways of managing large HDD
populations in a more effective and efficient way than the small controller
architecture of classical RAID
systems.
It will be lonely for the HDD array sales guys when that
happens though - just as it has become quiet lately in the
tape library industry
conferences.
Happy 7th birthday Fusion-io
Editor:- December 10,
2012 - 7 years ago this month - Fusion-io was
founded. (Although it was called something else at the time.) So much
disruption already - and only 7! What will it be like as a teenager? I
wrote about one of Fusion-io's several historic disruptive impacts today on
our home page and then later moved
it to a new permalink in December 2005 news. ...read the aticle
LSI moves to NASDAQ
Editor:- December 7, 2012 -
LSI today
announced
that it will move its stock listing from
NYSE to
NASDAQ later this month.
A Better Mousetrap or Better Mice?
Editor:- December
7, 2012 - Sounds a bit scary to me - but that's
one of the many sessions
on the agenda at Storage Visions
next month.
See also:-
Animal brands in
the SSD market
An early shareholder valentine for STEC's board? - we love
your IP assets but not how you manage them
Editor:- December 6,
2012 - A company called Balch Hill Capital - which apparently owns about 9% of
STEC's common
stock today sent an open letter to STEC's board of directors urging them to
conduct a "hard-nosed review of its spending patterns" and to
consider other business options (to wit - reduce costs, sell IP or sell
the company) which the letter's authors imply might yield better outcomes
for all shareholders.
Editor's comments:- the
full
text of the letter contains analysis - very similar in tone to many of
the peer comparison based criticisims which I've made about STEC in past
quarters.
There are 2 months still to go till next Valentine's
Day... - but if you ignored the dates and and some of the data in this
letter - it would be hard to judge - merely from the sentiments expressed
within - whether it had been posted too early - or had already been
delayed 10 months in the mail.
Nimbus blog seems fixated on EMC's XtremIO
Editor:-
December 6, 2012 - when it comes to blogs written by SSD vendors - there's a
huge variation in the frequency and quality of publishing content - even
among leading SSD companies.
I had noticed during the 2nd half of
this year that the blog
written by Nimbus's
CEO, Thomas
Isakovich - never changed.
For over 6 months it was stuck
in a time warp of May
2012 - commenting on EMC's
acquisition of XtremIO.
Then
I got an email this morning saying Tom had written
something new.
Imagine
my surprise when I looked and found it was fixed on exactly the same core
topic - but from a different angle - in which he critiques the scalability
and efficiency of XtremIO's architecture. (His timing was apparently prompted
by earlier blogs this week by Robin
Harris who interpreted a recently published
interview
(elements of which it now seems may have been
misunderstood)
with EMC's Chuck Hollis -
none of which I had read before today - because I don't expect to see SSD
thought leadership to come from that direction (EMC).
Which isn't
news to regular readers. But if you've missed the last
N years
of my SSD ramblings and want me to clarify this stance - it's simply because
EMC's response to the SSD business model - in the modern SSD era - has been
reactive and therefore has been pragmatically weighted towards
acquisition, badge
engineering and software integration - which if you know the company's
history - looks a lot like what they were doing since the
1990s too and which was a
very successful business strategy for them. When all the hardware components
are the same for everyone (as they were for 20 years in the
HDD array market) then
it's the software and services which make the enterprise difference. We haven't
got to that level of stasis yet in SSD hardware (and are nowhere close to
fossilization) - which is
one of the
reasons that leadership in the
SSD software market is
still up for grabs too.
Going back to
Nimbus's new blog -
Thomas Isakovich says - "There
are numerous limitations with the XtremIO design."
If the
past is any guide to the future in this respect - then that link - which is to
the home page of Nimbus's blog - rather than a permalink for that specific post
- may still be stuck in that groove in another 6 months time too.
To
my mind - when SSD vendors compare their products to those from EMC - it's
almost as regressive as
comparing SSDs
to HDDs.
...Later:- on December 13, 2012 - Jean-Jacques Maleval,
the editor of StorageNewsletter.com
- who had conducted the original interview with Chuck Hollis (EMC) - published
a
new
article in which he utterly refutes the implication that he didn't
understand what was being said. And to emphasize his point Jean-Jacques also
posted a an MP3 recording of the said interview (which I don't have the time
or inclination to listen to) but which some of you might be curious enough to
sample.
Lite-On made 1.5 million SSDs in 2012
Editor:-
December 5, 2012 - Lite-On
IT says it could make in the region of 4 million SSDs next year -
according to a
report in DIGITIMES.
IBM paper on SSD's data challenges
Editor:- December
4, 2012 - I've been browsing the
online
papers from last week's Server
Design Summit.
If you can only read one - it should be this:-
How
do we handle all the data? (pdf) - by Andy Walls, Distinguished
Engineer, Storage Hardware Chief Architect, IBM.
I won't give
away the plot - because I think you'll enjoy it more if you read it yourself.
But here are a few bullet points which resonated with me.
Among other
things - Andy Walls says...
- "90% of all the data in the world today was created in the last 2
years."
- "Flash is starting to free up the IO bottleneck - but the bottleneck
is more complicated..."
...read
the article (pdf)
See also:-
future SSD capacity
ratios , SSD
bottlenecks, where
are we now with SSD software?
SSDs are eBay's site acceleration helpers
Editor:-
December 4, 2012 - ""SSDs have been around for a long time... and are
no longer being seen as an exotic technology, but as an enabling technology that
allows companies to achieve a competitive advantage, as is the case with eBay."
That sounds like it could be a headline from today doesn't it?
But
it was said on this very news page 12 years ago by Robert David who was -
at that time - VP of sales and marketing at an enterprise SSD pioneer company
called Imperial
Technology - which was announcing that eBay was using its SSDs to speed
up transaction processing in the busy online shopping ramp to Christmas.
For
the full archived story see
Imperial's profile
page. For the context - and
other storage news in
that week - click here. For hundreds of similar stories - see
SSD market
history.
$1 / year salary for STEC's CEO
Editor:- December 3,
2012 - STEC
today
announced
that the company's interim CEO, and former CEO and founder Mark and
Manouch Moshayedi have reduced their salaries to $1 and they
have also recently purchased approximately 275,000 shares of STEC common
stock (each) on the open market to show their continuing confidence in the
future of the company and to help the company reduce its operational costs.
Editor's
comments:- Re the $1 salaries - I don't think we're going to see this hair
shirt idea become the new fashion for SSD CEOs.
Re - the stock
purchases -I don't think that's relevant in the places where it can really
make a difference. The real voters in STEC's future gameshow are its potential
customers and not investors.
new benchmarks from Violin and Fujitsu
Editor:-
December 3, 2012 - Violin
today
announced
that its SSD racks were used as the storage in recent
VMware benchmarks posted by
Fujitsu.
experimental technique eliminates flash endurance limit
Editor:-
December 2, 2012 - An article in IEEE
Spectrum -
Flash
Memory Survives 100 Million Cycles - summarizes a recent research paper by
Macronix -
which described an experimental technique to redesign flash cells to improve
endurance.
The
technique - which StorageSearch.com does not think is feasible to scale for
commercially competitive memory densities - involves designing addressable
heaters in the memory array which can pulse upto 800 degrees C for a few
milliseconds. This thermal "refreshing cycle" anneals the chip
material and heals common wear-out defects while also enabling the cells to be
run faster.
"Afterward, we realized that there was no new physics
principle invented here, and we could have done this 10 years ago" said
Hang-Ting Lue,
the project director at Macronix
| |
| . |
 |
|
| |
.. |
|
|
| .. |
SSD market
history
the Top
SSD Companies
what's
the state of DWPD? |
| . |
 |
| . |
| IOPS / $ as a value of
SSD goodness is bad |
Editor:- December 5, 2012 - The
cost of SSDs is one
of the arguments most often cited by antis to explain why (in their view)
the transition to a pure SSD storage market can't happen.
I guess
the designers of the first ships made from iron (which unlike wood doesn't
float) and the first airplanes (which were heavier than air) must've got used to
hearing similar objections.
In the past 10 years in various articles
I've described what I thought were the real
user value
propositions for SSD adoption in different markets. And I've got so used to
filtering out the inappropriate arguments about the cost per gigabyte of SSDs
which come from anti-SSD vendors and from those many
storage market researchers
who still don't understand what SSDs are about. These nonsensical blog posts
and white papers about SSD cost justifications don't upset me as much as
they used to - because there will always be crazy stuff on the web.
But
the broad sweep of principles which govern SSD adoption in the enterprise still
leaves much room for misleading analysis from those who should know better,
And
when that mistaken analysis comes from leading SSD vendors - then I feel it my
duty to show that it's
plain wrong.
So be wary of arguments for enterprise SSD
adoption which cite IOPS per dollar (or the other way round) as a
justification for filling a gap in some cleverly drawn curve.
I've
seen this recently from leading SSD companies who should know better.
As
human beings we feel comforted when we think we see new patterns. But they
don't always reflect reality. SSDs fitting gaps in IOPS vs dollars charts
isn't a sufficiently good reason to buy SSDs. Just as buying stocks based on
past performance charts isn't a good idea either. The future isn't a tidied up
remake of the past.
If you think about it for more than a microsecond
you'll see that the IOPS per dollar argument - which sounds plausibly
eco-technical when you first see it written down - doesn't lead to any
safe conclusions at all.
the zero cost but slow SSD fails this
test
Apply some
boundary
analysis to the situation and imagine an SSD which costs zero dollars.
According to the IOPS/$ advocates - this is your perfect enterprise solution
and at this cost it should even replace hard drives. But if the IOPS of the
SSD is considerably less than that of the hard drive - you'd be nuts to use
it - because you wouldn't get useful work done on your apps. (The model has
broken down in this direction.)
ultrafast SSDs with low capacity
fail this test too
At the other end of the scale - it's easy to
picture ultrafast SSDs which could score very well on the IOPS / $ scale - but
whose capacity or electrical power consumption or or physical size or
reliability doesn't make them scalable or attractive solutions.
I said
recently on these pages that if you're going to try and come up with a single,
plausible-sounding eco-technical marketing-jargon metric for SSD adoption
in virtual server environments - then a better suggestion is "cost per
happy virtual user". That's the total system cost BTW (TCO) - which
includes all the servers and software licenses and service costs etc etc.
IOPS
per dollar is a useless metric
Once you start looking at complex
real-world systems - which run more than a single type of app - you'll see that
the only way in which users can optimize their total economics is by
mixing and matching different speeds of SSDs in what I call different
silos in their apps
server architecture.
Even if you haven't read that read that
article it's easy to grasp why I say you can only get the truly lowest
system cost by using different types of SSD with different speeds,
different capacities and different operating characteristics (and certainly
different $/IOPS) - within ratios which make sense for that type of app mix and
user load.
$/IOPS is an illusion which doesn't take you anywhere far.
Now
- everything above - up to this point here - is something I've already said or
implied before - so what can I leave you with on the SSD news page today which
is new?
Well here's something I've been thinking about.
Faster
SSDs may reduce the amount of SSD capacity you need in a well designed VM
environment - compared to using slower SSDs - for the same number of users. The
reason is that while there is some overhead in capacity which you can't avoid -
on a per user basis - such as their unique data - there's also a lot of
transient system storage capacity which gets allocated and deallocated
dynamically to get their work done.
As we know in the hard drive VM
world - at the threshold of usable performance you need more capacity (really
- more drives) simply to get enough IOPS performance. The HDD experience is
the best sales person for introducing SSDs.
But when you're comparing
different types of SSDs in a VM situation - then the software makes a big
difference too. If an SSD from vendor X is twice as fast as that from vendor Y
- it means that SSD-X frees up some of its capacity faster than SSD-Y and can
therefore handle more users for the same purchased capacity within the same
elapsed time.
How big is that percentage difference?
I'm sure
we'll see more competing claims coming out from the
SSD software industry
in the months ahead.
...Later:- several years later - due to
difficulties of advocating enterprise SSD adoption to users in who often had
high uncertainty about how their installations worked, uncertainty about their
future workloads and business directions coupled with doubts about vendor claims
- a new phenomenon in SSD pricing arose which bypassed ration justifications
altogether. I discsussed this in
Exiting
the Astrological Age of Enterprise SSD Pricing | | |
. |
 |
. |
"Thanks
for the offer, but...
we don't want to deploy any new hard drive arrays.
Not even if you're giving them to us free!" |
| This classic article described the pivotal
future storage market climate in which enterprise users will cease to regard
hard drive arrays attractive or usable - even if the cost of buying a new hard
drive array drops away to ZERO! -
this way to the
petabyte SSD | | |
. |
 |
.
. |
 |
|
| |
