| leading the way to the
new storage frontier|
This page includes the archived SSD news from
StorageSearch.com for April 2013.|
mentioned in SSD news this month included:-
Hybrid Memory Cube Consortium,
SMART Storage Systems,
Texas Memory Systems,
|NetApp validates FlashSoft caching|
April 30, 2013 - SanDisk
that its FlashSoft (auto caching) software
has been validated for use with NetApp's
enterprise storage products.
Editor's comments:- NetApp are
pleased that their old fashioned storage arrays can be made to look more
sprightly through the correcting lens of server side SSD cache. They've got
a video and faqs page about the FlashSoft technology
SMART samples 2TB $3,999 SAS SSD
Editor:- April 30,
2013 - SMART
Storage Systems today
it is sampling a new 2.5" SAS SSD with 2TB capacity (oem price under
Using 19nm MLC - the 100K/45K R/W IOPS -
Eco - is rated at 10 drive writes per day
Diablo names SMART Storage as exclusive flash partner to pioneer
memory channel SSDs
Editor:- April 25, 2013 - You may remember
reading here before about a company called Diablo Technologies -
which while in stealth mode - hinted it was working on a new technology which
would enable SSDs to run on server motherboards with latency and throughput even
better than PCIe SSDs....
Diablo has been creating the interface side of things - but I learned
recently that implementing the flash side of this - in a manner which is both
effective and affordable - requires a world leading mastery of enterprise flash
IP - which Diablo wisely recognized it doesn't have.
has publicly announced an exclusive partnership agreement with SMART Storage Systems
which will leverage its flash IP and controller assets to co-design a new family
of ultra-low latency SSDs and system accelerators which connect via Diablo's
memory channel storage architecture and which will be sold exclusively by
SMART but jointly supported by both companies.
comments:- there are a lot of
the future direction of SSD server acceleration if this collaboration
succeeds in delivering competitively attractive new types of SSDs. But there
are also very difficult technical problems and ecosystems development problems
to solve too in order to make it viable.
I discussed these topics in
a conversation earlier this week with John Scaramuzzo,
President and Esther Spanjer,
Director SSD Marketing at SMART.
Among the many questions inspired
by that conversation:-
- how is the new technology different to what has been done before? -
particularly with PCIe SSDs and with DIMM class flash?
- if successful - what impact would memory channel SSDs have on the PCIe SSD
market? - and application server architecture?
- how will the new types of SSDs stretch the demands of flash endurance and
- how will competitors respond to this new technology? And how much of what
they say should you take take on board or disregard?
- who are going to be the among the first wave of customers to adopt these
I'll be writing about these
matters and more in
home page blog on StorageSearch.com which will be published Monday 4pm ET.
See you then.
- when will the first products be ready?
Fusion-io enters the iSCSI array market
April 24, 2013 - Fusion-io
made 2 significant announcements today.
The 1st of these was
financial results for the quarter ended March 31 - revenue of $88 million
(down 27% from the preceding quarter and down 7% from the year
The 2nd of these was the real news -
FIO has acquired another company -
NexGen Storage (for $119
are SSD ASAPs
(hybrid caching systems with integrated real-time
dedupe and QoS
controls for VDI apps) which use Fusion's PCIe SSDs in standard servers with
conventional hard drives to deliver
iSCSI hybrid storage for
SME and departmental needs in a 3U rack which delivers upto 150K IOPS and
16TB to 192TB raw capacity.
that on a per-U basis their systems deliver 10x more IOPS than HDD arrays,
3x more IOPS / U than conventional hybrid arrays and 3x more GB / U for VDI apps
than pure SSD arrays.
These kinds of comparisons always depend on
which competitor you're comparing with and when the comparison was done.
However - the company has enough customer case studies and independent
analysis papers on its site to show that real customers liked the products.
up the 2 stories today?
FIO had already indicated that its revenue
from its known biggest customers would decline for a few quarters - so the
financial results are not a great surprise. But the NexGen announcement has
opened the door to an entirely new type of customer for Fusion-io at the other
end of the SSD adoption scale - compared to the well known big customers which
have until now dominated FIO's business.
Will it work?
used to being the leader in the
PCIe SSD market which
it largely helped to create as a significant new part of the server ecosystem.
But it will require a different type of marketing and business development
approach to convert the potential of NexGen's technology into an equivalent
leading role in the more conservative and crowded iSCSI market.
the other hand if you add NexGen's hybrid iSCSI IP to the marketing magic of
Fusion-io - it's safe to predict that the iSCSI market will soon be getting a
wake up call the likes of which it has never seen before.
...Next on the SSD world domination agenda - create better value
in the cost sensitive iSCSI market
Editor:- April 23, 2013 - The iSCSI market hasn't been a
fertile business development ground for SSD sales - a factor which I ascribe
to the mood prevailing at its birth. At the start of 2001 - when the idea of
iSCSI first attracted interest on the web - the
storage market was still in
a recession which would continue for another 2 years. Users could buy new
or little used servers and storage recycled from the spending spree of failed
dotcom companies for next to nothing. There was already a proven fast way of
doing fast network storage - fibre-channel
which had been around since
1994 (but it was complex to set up). Those various factors meant that iSCSI
evolved - by necessity - into a cheap, simple to set up and maintain storage
ecosystem for frugal applications which needed data.
was nothing hard wired into the technology which prevented it from being scaled
up - most of the early attempts by vendors to nudge iSCSI into the fast lane
with dedicated hardware accelerators failed. There was no real customer
appetite in the iSCSI base to encourage vendors to push for fast random IOPS
or low latency. iSCSI was the frugal way of doing complicated network
That's another reason why - prior to 2013 - none of the top
10 enterprise pure SSD array companies started in iSCSI. There wasn't enough
market demand for the kind of low latency and fast IOPS which could open enough
doors for SSDs in storage cabinets to make it worthwhile. Instead, most of the
iSCSI arrays which have been in the market until recently were originally
developed around technology optimized for FC SAN or were simply iSCSI HDD
arrays with some SSDs thrown into some of the bays. When you saw "iSCSI"
on the datasheet of a fast SSD you knew it had most likely been added
to a model which had already been optimized for another market.
could say that iSCSI has been a safe haven for enterprise
hard drives - because
whenever there has been a tension in the feature set between the cost of
incremental capacity versus the value of incremental performance - it was cost
- and getting the cost down as low as possible - which usually won.
explained in my Petabyte
SSD roadmap article a few years ago why one day - even the mantle of low
cost per raw terabyte wouldn't be enough to protect delinquently slow and
ineffcient hard drives from being evicted from enterprise network storage
racks. And this culture shock will be knocking at the door of the iSCSI market
from various different vendor directions in the coming year - with increasing
I was pondering these factors last week when I was waiting to
Rosenthal, Senior VP Marketing Astute Networks who
wanted to talk about the
of new models in their ViSX family of fast-enough iSCSI rackmount SSDs -
which have upto 45TB of raw SSD storage in a 2U rack which with
can deliver $2,000 / TB and even with dedupe switched off - comes in at about
$5,000 / TB while being able to offer more than double the IOPS of much higher
priced competing SSD systems.
The first thing I asked about was the
company's iSCSI accelerator chip - which is one of the two technology factors
which give them an edge in iSCSI. I had heard about it many years ago - but the
company doesn't say much about it now. Len told me they were now on the 3rd
generation of their iSCSI accelerator chip. The 1st generation had been
designed for a US Navy project to enable fast access to embedded storage
located around a ship while using COTS (Commercial Off The Shelf) servers and
In Astute's current ViSX systems I think you can view the
iSCSI accelerator as being the technology which buys the time (in latency cost)
which can then be spent on dependable real-time dedupe.
me that although Astute have always known this gives them a theoretical
performance advantage compared to competitors who use similar types of flash -
it's only when he engaged Demartek
to do some comparative testing recently and gave them a free hand to explore
the differences - that they realized just how good their systems were. (I've
seen summaries of these benchmarks - and they do confirm the advantages of the
Astute's new systems do now seem to offer a hard to
beat SSD package for users in the mainstream iSCSI market. Len described
this as "making flash affordable for the mid market."
earlier generations of iSCSI flash were too expensive for most users. But the
current generation - not only offers attractive pricing - but comes with proven
technologies - and cost effective replication - by what the company calls
availability groups" (pdf)- which enables users to choose which
systems provide failover clustering - and whether that's local or remote. In
addition to providing data continuity when things fail - this scheme can also
provide load balancing and imporved performance in the normal (unfailed)
One of the things which came across clearly from talking to Len
is that Astute Networks is totally focused on the iSCSI SSD market. They
know the market, they know the apps - and they aim to be one of the leading
suppliers in this niche. For them iSCSI isn't something on the tick list - it's
the whole list.
For alternative and competing companies in this
market segment search for
new SSD module for mobile military systems
April 22, 2013 -Curtiss-Wright
the availability of conduction cooled secure 1TB SATA SLC SSD modules
for use in its rugged 4 port NAS module which is designed to fit on an ARINC
SSD - designed for applications such as helicopters, UAVs and mobile
radar systems - is certified to FIPS 140-2 and provides 4 modes of key
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.
Our PCIe SSD business is negligible today - but we plan to
change that - says SanDisk's CEO
Editor:- April 18, 2013 - Nearly
revenue still comes from SAS
SSDs - derived from their
March 2011 - and
the company's PCIe SSD revenue today is "negligibly small" but they
see PCIe SSDs as a
large market opportunity which they want to get into with products they will
launch in the 2nd half the year.
That was the gist of the message
Mehrotra, cofounder and CEO SanDisk - in the company's earnings
conference call yesterday.
Other things which emerged:- SSDs are 20%
of SanDisk's sales this year, and like other
flash memory makers
SanDisk is reluctant to invest in new wafer fabs while there's still
uncertainty about the exact direction and proven viability of flash technology
beyond the current 2-3 years window. ...read
transcript on SeekingAlpha.com
OCZ will exit SandForce driven consumer SSD market
April 17, 2013 - OCZ
estimated revenue for the quarter ended February 28, 2013 in the range
from $65 million to $70 million.
Editor's comments:- That's
in comparison to
revenue of $110 million in the year ago quarter - which for most companies
would indicate that business has been getting worse.
However, due to
auditing problems which placed the company outside NASDAQ compliance limits
last year and led to the departure of its founder - it may be that a better
quality of new revenue - due to getting more of the right kind of business -
may lead to a more positive place than getting more of the old wrong kind.
The company also announced today that it will move the majority of
its consumer SSDs to its own fast in-house SSD controller technology in the
next few quarters.
Effectively exiting the very competitive
controller driven consumer
SSD market should make it easier for OCZ to differentiate its products
and get better profit margins.
the Top SSD Companies in 2013 Q1
Editor:- April 17,
2013 - StorageSearch.com
today published a new edition of the Top SSD Companies.
new WebFeet report on 2012 non volatile memory market
April 16, 2013 - the
flash memory market was
worth just under $28 billion in 2012 - down 3% from the year before -
according to WebFeet
Research - who have
a new report CS700MS
($2.5K) which analyzes nvm market share.
WebFeet have got a new website design too. It's worth a look and is
significantly better than what they had before. Having said that - the design
uses reversed text (white text on a dark background) - which is OK for sites
where you aren't going to read much - but not so good on the eye for longer
IBM aims to be multi-billion dollar flash systems supplier
April 12, 2013 - 3 years ago I wrote a
blog about the
confusing nature of the "RamSan" brand of SSDs from Texas Memory Systems
given that all the recent models in the family were in fact
flash memory rather than
RAM based - and
furthermore some of the models didn't connect via an
FC SAN but used
it wasn't a surprise to see in yesterday's
by IBM (who
last year) that the RamSan designation has been dropped in favor of the more
accurate sounding "FlashSystem" in those models which migrated
intact to IBM's
enterprise flash product line.
So - for example in the category of
availability rackmount SSDs - the old RamSan-720 (SLC) and RamSan-820
(MLC) have become the new
FlashSystem 720 and 820. If you're not familiar with these fast HA SSDs -
the thinking behind their design came out in an
interview I has with
Holly Frost, CEO of TMS when they were launched in
I missed them - then it doesn't look to me as though TMS's PCIe SSD models
have been so fortunate. I couldn't see them in IBM's range of PCIe SSDs (High
IOPS Modular Adapters) which are based on products and technologies from
LSI. That no-show
may be due to the fact that - unlike TMS's rackmount systems which were
software agnostic - a lot more work is required to efficiently integrate server
based SSDs into a wide range of server products. But I anticipate that
architecture SSD controller technology will resurface in future IBM SSD
Much more significant was the news that IBM is investing
$1 billion in research and development to design, create and integrate
new flash solutions into its portfolio of servers, storage systems and
middleware. IBM also announced plans to open 12 centers of flash competency
around the globe. That demonstrates confidence in the
future scale of the
SSD market and a clear sense of
SSD's place in computer history.
let's hear it again for Samsung's 10nm TLC
April 10, 2013 - the difference between "production" and "mass
production" wouldn't normally be enough to rate a 2nd mention on this news
page - even after an interval of 5 months by which time most editors will have
forgotten the earlier news instance.
But a worthy exception to this
little editorial rule of
mine is Samsung's
10nm, x3 MLC, 128Gb nand
flash which the company
Which way round did the transition go?
That would be
a valid question if you knew nothing at all about the
dynamics of the SSD
And at some time in the distant future the flash
fab taps may indeed be turning down the flow.
But just to reassure
you - in case you have any doubt - it was the transition from "production"
to "mass production" which the company noted today. No need to
change the memory investment portfolio just yet.
Addonics launches SSD duplicators
9, 2013 - Addonics
a family of mSATA SSD duplicators for copying 5, 9 or 11 drives at a time.
Prices start at $849. They can also clone CF and CFast cards.
Crocus gets funding for x8 multibit magnetic semiconductor memory
April 8, 2013 - Crocus
announced it has
been awarded a contract from IARPA
to develop an 8-bit per cell memory based on its Magnetic Logic Unit
This will greatly reduce the energy consumed per
written-bit compared to any other memory technology, including DRAM, Flash,
SRAM and MRAM.
Lee, VP, product development at Crocus compared the 8 bits per cell
which the company thinks it can get from its MLU technology with the
state-of-the-art in nand flash - which is 3-4 bits per cell and also compared
to alternative magnetic semiconductor technologies like MRAM - which is
still only 1 bit per cell storage (SLC).
here's some context.
If it were possible to do x8 MLC flash - then
840 SSD would have 16TB capacity instead of the 512GB which it has using
x3 (TLC) - which is the state of the art bits per cell shipping in a
regular 2.5" SSD.
But don't get too excited by this comparison as
currently exists only in the realm of science fiction.
multibit capability in a magnetic semiconductor cell will undoubtedly be a
breakthrough for that type of non volatile technology. But the density of such
x8 MLU memories would still be 100x smaller than today's flash. The good
news is that unlike flash - MLU will operate at very hot ambient temperatures -
past 200 degrees C.
Intel oems LSI's RAID caching SSD technology
April 8, 2013 - Intel
- which already uses LSI's
SandForce controllers in some SSDs - will oem LSI's dual-core RAID-on-Chip
flash caching technology it was announced today.
LSI says their
caching technology can double the number of VDI sessions supported in the same
sever and flash environment.
"Intel's selection of
Nytro MegaRAID technology is another significant validation of our strategic
focus and investments in flash-based server acceleration technology," said
senior VP and GM, Accelerated Solutions, LSI.
one of HP's most famous former employees
April 7, 2013 - HP -
which began shipping Fusion-io's
PCIe SSDs in its servers
4 years ago - is
now integrating FIO's ioFX
SSDs into some of its workstations aimed at the movie and video editing
market - it was
Editor's comments:- in a
blog about this -
HP's head of (related) product management Jeff Wood used
the phrase - "one of HP's most famous former employees" - to
Wozniak - who before founding Apple - and long before becoming
Scientist at Fusion-io -
worked at HP designing calculator chips.
Those were very sophisticated calculators - I recall - because my parents had
been selling and repairing electronic calculators
in their shop since 1964 - and later
became HP's first calculator and PC dealers in the local area.
chipsets - and solving the problem of how to produce a wide range of useful
products at low cost using a common set of silicon chips -
the genesis of the microprocessor market.
Hybrid Memory Cube spec ready for chip designers
April 3, 2013 - back in
October 2011 - I
reported on this page the formation of a new industry
ORG - the Hybrid Memory Cube Consortium
- which could have an impact on future SSD packaging densities.
takes a while to get these things going - but according to
press release this week by one of the founding companies - Micron - the 100 plus
companies which are collaborating in this enterprise have agreed on an
A key feature of the new multiplane memory
architecture is that distributed memory controllers in an HMC module will
handle the data I/O packet requests for the bunch of stacked memory chips in its
own vault. This is similar to the distributed intelligent data mover concept
which is already used in all proprietary
SSD controller designs - because it's the only way you can get good
aggregated global system performance while also dealing with low level
local memory management issues at low latency.
As with earlier
generations of remote distributed memory interfaces - such as
InfiniBand - HMC is
designed to optimize the request of small packets - which in the case of HMC is
16 to 128 bytes of data.
With today's semiconductor speeds -
accessing the data in those distributed memory chips within the same HMC module
presents similar technical problems to distributed memory cards in traditional
computer designs - because traversing inches of physical space at high speed is
as difficult as moving data across tens of feet at slower speeds.
has been born as a DRAM
technology - but don't ignore it - just for that reason. (Or because the data
packet sizes are small compared to the block sizes in
nand flash.) If and when
these HMC packaging ideas result in viable products - the ideas and
methodologies will spill into SSDs too -regardless of what the underlying
memories used in SSDs may be at that time.
It's all about speed and
scalability. According to the HMC
faqs page - A single (1st generation) HMC unit can provide more than 15x
the bandwidth of a DDR3 module. See also:-
SSD interface glue chips.
"We've shipped more SSDs to the enterprise than any other
supplier" - says STEC
Editor:- April 2, 2013 - In a
release today STEC's
CMO, Ali Zadeh
makes a contentious statement which goes something like this...
the company that has shipped more SSDs to the enterprise than any other supplier,
we have the unique capability to provide critical application knowledge and
experience to solve our customers' most critical issues..."
comments:- OK you could say there's more than one issue here on which SSD
commentators might disagree - so just to clarify - what I'm referring to - it's
the first half of the sentence which I've highlightened in bold text.
first glance - this positioning for
enterprise SSD market
leadership seems hard to reconcile with the state of the enterprise SSD
market - because based on other publicly available data - there are other
companies which have shipped significantly more
petabytes of SSD
storage into the enterprise than STEC.
read the words carefully - and with respect to "units shipped" and a
very narrow interpretation of what is meant by an "enterprise SSD"
it may be possible to reconcile STEC's statement in the context of what I
assume could be - lower capacity SSDs - defined in a very particular way to
exclude other claimants to this title.
The 2nd half of STEC's
assertion - re their "unique" knowledge of the enterprise SSD market
is easier to defend.
STEC's loss of market share in enterprise SSD
revenue in recent years - speaks clearly enough for itself. It's the kind of
unique understanding of a market which competitors might be happy to do without.
- STEC now sees itself as sTEC. I thought it was a typo when I saw it in my
email this morning. But the lower case "s" has been carried forward
into a compatible new image for STEC's logo. So I guess it was done
deliberately. Should we be reading this as a clue to where the business
priorities are for this SSD company?
PS - this seems like a good point
to bring again to your attention the first bullet point in my article -
Survivor's Guide to Enterprise SSDs - "Don't believe everything SSD
companies tell you about the past, present or future of the SSD market" -
which, among other things, includes examples of very contentious
statements made in the past by various leading SSD companies.
...Later:- April 4, 2013 - a blog by
Taneja Group -
and the Enterprise - Yes, choice matters... argues that STEC's technical
SSDs remains a strong argument for looking look at the company's products
- and the article's author poses the question - "If youre looking for a
highly reputable device these days, it frankly might come down to sTec, Intel,
or one of these other vendors who has been
swallowed up by
a bigger critter. Which would you rather have?"
x4, hot flashes, paramagnetic semico and SSD
April 2, 2013 - I decided not to run any "April fools" SSD stories
yesterday - because the last time I did so -
3 years ago - in
a spoof which linked strong magnetic fields to better
- when used with adaptive read flash technology at the x4 level - both
technologies all too soon had started to blur with real emerging technologies.
And the SSD market is confusing enough without these misdirections as
demonstrated by these real examples.
- new flash chip functions - such as the superheaters described by
Macronix in December
2012 - which stretch endurance past 100 million cycles.
- all the various convoluted magnetic semiconductor strands of product
development still going on in the "alternative to flash" segment of
the nvm market.