SMART Storage Systems - (upto
2013)
is now SanDisk Enterprise Storage Solutions |
| SMART Storage Systems is a
technology leader in the design, development, and deployment of current and
next-generation enterprise solid-state storage products. SMART delivers
high-quality, high-reliability solutions to a broad customer base, including
tier one storage and server OEMs. |
.... |
 |
see also:-
SMART
- mentions on StorageSearch.com,
the SSD
Bookmarks
which SMART SSD company?
In
2012 - SMART
split into different parts - which confused me even 2 years later in
2014
- despite having known some of the details in advance.
That's because
instead of creating a pure SSD business and a pure memory business - they
subdivided the SSD product lines too.
- for SMART Storage (the enterprise SSD business - which was spun off from
SMART Modular) please stay on this page and read the many articles about what
they did below.
| |
|
Who's
who in SSD? - by
Zsolt Kerekes,
editor - April 2013
SMART Storage Systems is a
Top 20 SSD company -
which designs legacy
oriented enterprise
SSDs - using its own proprietary
small controller
SSD architecture - which leverages
adaptive R/W
and DSP IP - to yield
efficient SSD designs.
SMART also recently anoounced it's collaborating with a company
called Diablo Technologies
to design a new type of
new dynasty
fast server-side SSD which will be compatible with DRAM server sockets but which
will be flash rich - and
pitched as an alternative way to accelerate apps compared to standard
PCIe SSDs - using
flash as a memory tier instead of as fast SSD storage. The first sample ultra
low latency SSD products from this venture are anticipated towards the end of
2013 or early 2014.
SMART has traditionally been a supplier into the
industrial SSD
market. In some of these designs which use industry standard 3rd party
SSD controllers from
LSI - SMART is able
to offer enhanced
endurance
due to using SMART's preconditioned flash memory.
SMART is 1 of
hundreds of companies listed in these directories:-
SAS SSDs,
2.5" SSDs,
3.5" SSDs,
1.8" SSDs,
SATA SSDs,
PATA SSDs,
SCSI SSDs,
enterprise SSDs,
military SSDs and
industrial SSDs.
SMART
doesn't currently market PCIe SSDs. The company did launch one back in
June 2009 but it
isn't a current product line. Recently SMART has indicated its willingness to
re-enter this market when conditions seem right to them with the establishment
of industry standards. However, it looks like that will be after the company's
upcoming memory
channel storage developments.
Here are earlier profiles and articles about SMART's SSDs
my
preconceptions were out of date...
Until February 2012 -
I had thought of SMART as being just one of the many (30+, now 50+)
makers of
SandForce inside SSDs
although having some special
reliability
wraparound twists - for example in the
sudden SSD
power loss and
flash memory
screening departments.
SMART had gotten into the higher performance rugged SATA and SCSI SSD
markets by acquiring Adtron
(a top 10 SSD maker) in
February 2008.
Prior to that - SMART's flash SSD family had been low power PATA SSDs.
But
for years after - it seemed to me that nothing much exciting emerged from
SMART in the SSD field.
Sure - the company had steered more towards the
enterprise market (and away from its rugged / MIL roots) and SMART had gotten
its SandForce driven 2.5" SAS SSDs qualified in some IBM computers in 2010
- but so what?
Reliability? - yes. Dependability? - yes. These were
traits I associated with SMART's SSDs. And they were among the best at writing
white papers which explained the behavior of their SSDs to systems designers.
But SMART wasn't a company which came to mind - when I was thinking about the
subject of leadership in SSD controller and architecture innovation.
conversation
with SMART's President - John Scaramuzzo
My preconceptions were
wrong - as I learned in a conversation I had with John Scaramuzzo,
President at SMART Storage Systems in February 2012.
John
joined
the company in January 2010 and his SSD team have done a lot in the last 2
years - mostly out of sight of people like me - until they were ready.
We
got right down to the subject of SAS SSDs and their new
Optimus
Ultra.
John said he knew from his experience at
Seagate/Maxtor/DEC
that getting a SAS
drive qualified in tier #1 server companies was hard to achieve.
Only
a small number of companies had managed to do this. And some had spent a lot of
money and failed.
This was one of the projects he set in motion when he
joined because he expects that SAS SSDs are going to be a big factor in the
enterprise.
SMART has put a lot of resources into getting its SAS SSDs
accepted in tier #1 - and they are already qualified in a spread of different
servers within IBM - just to name one customer.
Doesn't that take a
lot of effort? - I asked. - And doesn't it mean you need a technology roadmap
to make it worthwhile? Otherwise the effort isn't sustainable.
John
confirmed it did - and said - "You can't get into tier #1 if you are
only a one trick pony."
That brought me onto my next
question.
Did I read this right? - I asked. - Do you really have a
new controller?
When I had first read about the Optimus last
summer I had come away with the impression (based on the performance figures)
that it was just another SF inside. I had disregarded all that other reliability
verbiage as marketing fluff - which was mostly about their award winning power
loss protection etc.
John confirmed - yes - it is a new controller /
management architecture - and it's SMART's own design.
John said that
to develop the data integrity features - he recruited a bunch of ex recording
guys with traditional storage media backgrounds. He told them - take a look at
this flash stuff and see if you can use your skills to create more reliable
media management. That's what they did - and the flash controller in the
Optimus
Ultra is the embodiment of those efforts.
John said it uses a
variety of advanced techniques including DSP and adaptive programming.
SMART
is now a world leader in enterprise flash management
SMART is now
regarded by some of the memory companies they talk to as one of the world
leaders in converting unreliable consumer grade flash into very reliable
enterprise grade SSD.
He said only 2 other companies have anything
close to their DSP technology -
Anobit and
STEC.
John
said he was relieved when he heard that
Apple had acquired
Anobit - because that
effectively removed their technology as a potential competitor in the
enterprise market.
He didn't have to say anything about STEC - because
they are the company which every 2.5" enterprise flash SSD maker has to
contend with. STEC has been taking a battering in the market - from many
competitive directions - but SAS SSDs is one of their core territories.
John
said SMART was already getting a lot of positive outcomes from its tier #1
qualification efforts - and he named a bunch of customers and systems which I
can't reveal to you at this stage. I took this as a sign that STEC now faces
serious competition in its home turf. If STEC's enterprise marketing was as
good as its SSDs... But that's a very old story now.
SLC write life
from consumer MLC
Here's why SMART's tier #1 customers are excited
by the new Optimus Ultra.
You put in 3K write endurance consumer grade
MLC - and get an SSD-wide reliability closer to 40K.
The flash
management is scalable and not tied to a single vendor's product or generation.
Far from it.
The new controller learns from each chip the best
way to handle it - and can even use different parameters in different parts
of the chip and at different points in time too.
Adaptive writes -
not just clever wear-leveling
Take a write pulse for example.
(And this is my inference here rather than what John told me.)
Externally
from a black box viewpoint you do a single write of a prescribed logical width
and that's that. But internally the flash chips initiate a series of high
voltage erase pulses.
The chipmakers choose cell erase constants
which guarantee that enough blocks will work to yield a good device yield over
the life of the product for the worst cells. But that means - early on in the
device life and for most of the cells throughout their life - the erase
pulse energy is much more than it needs to be. This causes undue stress in
the device - and results in sub-optimum life.
What you can do instead
- if you have the trust of the memory makers - is to get access to the internal
programming pulse circuit parameters and tune them yourself. Cutting down the
over-kill write-erase energy reduces the stress in the chip caused by the
high field strength - so you are in fact getting a better life out of the
chip. It's not just about clever wear-leveling techniques.
When the
first flash devices came to market over 30 years ago - this programming
mechanism wasn't hidden - and I remember being surprised when I peekd into the
numbers of my home-brew programming algorithm - that some parts of a flash
chip were so much harder to program than others. Later flash generations hid the
pulse mechanics inside the chip to make the memories easier to sell and to
prevent accidental destruction which could happen if the system designer got the
voltages, pulse shapes and duty cycle wrong.
And that's just one of
the many of the ingredients
in the flash SSD cola wars.
back to SMART
I asked
John what this meant about the positioning of SMART's SF based enterprise SSDs.
He
said that there was enough of a gap between the two controllers to ensure a
market for both types of SSD in the product line.
I also picked up on
the idea that SMART's new controller means their enterprise SSDs can be more
profitable while also being more reliable - compared to competition which used
eMLC. In theory that's also a strength of
STEC's CellCare too
BTW - but you only get the profit after you get the
sales...
Funally
I asked - what about seeing a PCIe SSD family based on the new controllers?
Surely that would be even easier for SMART to design than the SAS SSD had been.
Was that in their plans too?
John said he didn't want to make a
definite statement about that.
As you can imagine we also talked
about some past storage history - the disruptive effects of the SSD market etc.
And I nearly forgot to mention here that SMART Storage Systems has been
officially been spun out as a separate business entity - although that was
covered in a recent news story.
enterprise SSD doesn't stand
still
What's exciting from my point of view is whenever I think I
understand who are the leading players in the SSD market - and start making
mental lists of who's best at what and why - another conversation shows that
the pace of innovation in SSD hasn't stopped. The surprises keep coming.
For
more info about SMART take a look at the links above and
SMART
- editor mentions on StorageSearch.com
I currently talk to more
than 300 makers of SSDs and another 100 or so companies which are
closely enmeshed around the SSD ecosphere - which are all profiled here on
the mouse site.
I learn about new SSD companies every day, including
many in stealth mode. If you're interested in the growing
big picture of
the SSD market canvass - StorageSearch will help you along the way. Many
SSD company CEOs read our site too - and say they value our thought leading SSD
content - even when we say something that's not always comfortable to hear. I
hope you'll find it it useful too. |
|
editor's earlier comments:-
January
2012 - SMART Storage Systems (which changed its name recently from SMART
Modular Technologies) was acquired in August 2011 by private equity
company
Silver Lake.
SMART traditionally offered high performance industrial / military grade SSDs
in standard disk form factors. Like many other SSD companies SMART has been
focusing more attention to the
enterprise market
and recently launched some fast high capacity 2.5"
SAS SSDs.
SMART
uses a variety of merchant market
SSD controllers inside
its current range of SSDs - and it's tempting to think that
all SSDs with the
same controller are the same. But they're not - for a variety of reasons -
including the different ways that designers have of dealing with critical
elements like data
protection in the event of sudden SSD power loss. As a company with a long
history
in the industrial and military markets SMART's SSD design culture and process
experience means it's in a different reliability class to newbie SSD companies
like Intel. |
| . |
|
|
In October 2008 -
SMART started
shipping the Xcel-10 SSD - a 2.5" SLC flash SSD with upto 128GB capacity.
Sustained read speed is 115MB/s, and write speed is 125MB/s. (It really is
faster than the read speed). It delivers 5,580 IOPS at 100% read or 980 IOPS at
67% read, 33% write, for random I/O using 4K block size.
In February
2009 - SMART
announced new 3.5"
parallel SCSI SSDs with
upto 128GB and faster secure erase for industrial, defense, and other embedded
applications that require extremely rugged storage devices and legacy
interfaces.
In June 2009 -
SMART disclosed
it had used
Marvell's SSD
controller in SMART's new
XceedIOPS
PCIe SSD which offers
upto 400GB capacity and 140,000 random IOPS performance.
In August
2009 - SMART
announced a new range of
rugged
2.5" 256GB SSDs for
defense applications.
Data
declassification compliance is implemented by the company's EraSure
technology. The models comply with MIL-STD-810F environmental specifications for
operating shock, vibration, humidity and altitude, and each drive passes a
demanding 8 hour, full-temperature range
burn-in test prior to
shipment.
In September 2009 -
SMART announced
it has selected the
SandForce
SF-1500
SSD processor for use in its next-generation enterprise-class
SATA SSDs sampling
later this year.
In October 2009 -
SMART announced
that it has been
selected
by Harris Corp to provide SSDs for use in its Mass Storage Unit
program. The new MSU, which is part of a larger F/A-18 program, is the first of
a new family of avionics file servers. Harris selected SMART's
XceedSecure
2.5" SATA SLC flash SSD for the in-flight file server application.
XceedSecure SSDs include EraSure technology, which provides
secure erase
features.
In December 2009 -
SMART started
sampling the
XceedIOPS
SATA - SLC and "enterprise grade" MLC flash SSDs in
1.8" and
2.5" form factors
- based on the SF-1500 processor from
SandForce.
Performance is upto 30K
IOPS random
read/write. SMART uses a combination of
write attenuation
technologies to attain a 5-year projected lifetime for its 400GB MLC
XceedIOPS SATA model ($2,900 oem qty price) in an environment that demands
250MB/s sustained write and a 40% duty cycle.
In August 2010 -
SMART entered the
SAS SSD market with the
announcement that it is sampling the
XceedIOPS
SAS SSD - a 400GB eMLC SSD with 26,000 / 20,000 R/W
IOPS and
250/230 MB/s sustained throughput.
In October 2010 -
SMART
announced
it is sampling a new 2.5" 200GB SATA SLC flash SSD for mission-critical
defense and industrial
applications. The
Xcel-100
solid-state drive achieves up to 30,000 IOPS random read/write and 250MB/s
sustained read/write. Validated to MIL-STD-810F it's designed to operate in a
temperature range that extends from -40°C to +85°C, with the ability
to sustain 50g operating shock and 16.4g operating vibration. The Xcel-100
offers high reliability
and data integrity
(< 1 in 1017 bits read) that is supported by extensive
error-correction and detection capabilities, multi-level data-path and code
protection, data-fail recovery, and data-integrity monitoring. The Xcel-100 also
supports the ATA-7 Security Erase.
In November 2010 -
SMART announced
that
its
XceedIOPS SAS 2.5" solid state drive (SSD) will be used in new models
of IBM POWER7
supercomputers instead of hard drives.
In April 2011 -
SMART
announced the release of its
XceedIOPS2
6Gb/s SATA SSDs in 1.8"
and 2.5" form
factors which use with the latest
SandForce SF-2000
series SSD processors.
In
May 2011 - SMART
began sampling a new 64GB 1.8"
Slim (39x54x4mm ) SATA SLC SSD for embedded applications. The XceedIOPS
iSATA Slim has a typical power consumption of 0.5W at +5VDC and supports
automatic 5V/3.3V input power selection. R/W throughput is upto 120/115MB/s
and 14K/1.4K R/W IOPS.
In August 2011 -
SMART
launched
a new range of 2.5" SAS
SSDs which provide upto 1.6TB usable capacity, 100K/50K random IOPS and
500MB/s sustained R/W transfer rates.
In September 2011 -
SMART announced
imminent sampling of a new 2.5" SATA
MLC SSD aimed at
cost sensitive / general purpose enterprise storage apps. The
XceedStor
500S (60GB to 480GB capacity, 600MB/s burst throughput) combines a
SandForce
controller with
SMART's proven power fail technology. UBER is less than 1 in 10-18 bits read.
In
October 2011 - SMART
announced
imminent sampling of a SATA
3 version of its MIL-STD-810
compliant 2.5" SSD family - which includes encryption and
fast erase. The new
Xcel-200 provides from 60GB to 240GB
SLC capacity,
500MB/s sequential R/W speeds and 60K/40K random R/W IOPS. It operates at
standard industrial
temperature ranges and is certified for operation at altitudes up to 80,000
ft.
In February 2012 -
SMART Storage Systems
launched the
Optimus
Ultra (a 1.2TB 2.5" 100K/60K IOPS, 500MB/s R/W
SAS SSD) which uses
the company's new, in-house developed, high reliability enterprise SSD
controller IP - which includes DSP and adaptive programming techniques to
deliver industry leading SSD data integrity and upto 25x / day full disk
writes for 5 years endurance - while using low cost consumer grade MLC.
In
April 2012 - SMART announced that
it has figured out a way to get 5x more endurance from consumer grade
flash when using unmodified industry standard controllers from LSI/SandForce. The
technique - which is used in a new
fast-enough (enterprise
component) 2.5" SATA SSD product family sampling this quarter -
involves preconditioning parameters in the flash memory when the SSDs are
assembled (or at first boot) using intelligence gained from experience with
the company's population of closed-loop adaptive DSP controllers which are used
in its Optimus family. SMART applies a safety margin to compensate for the
open-loop aspect of these factors when managed by a SandForce controller to
which these underlying dynamics are invisible.
In June 2012 -
SMART Storage Systems
announced
it's sampling yet another new variant of
SAS SSD which uses
adaptive
flash DSP technology - the
Optimus
Ultra+ is a 2.5"
SSD with 100K/60K R/W
IOPS - which
can endure
50 full random drive writes per day for a period of 5 years using commercial MLC
NAND flash technology.
In April 2013 -
Diablo Technologies
named SMART Storage
as its exclusive flash partner to pioneer a new type of (faster than
PCIe SSD) memory
channel SSDs
In July 2013 - SanDisk
announced
a definitive agreement to acquire SMART Storage Systems
for approximately $307 million.
In August 2013 -
SMART Storage
Systems
announced
it had begun sampling the first memory channel SSDs compatible with the
interface and reference architecture created by Diablo Technologies. |
| . |
|
|
. |
| Surviving SSD
sudden power loss |
Why should you care
what happens in an SSD when the power goes down?
This important design
feature - which barely rates a mention in most SSD datasheets and press releases
- has a strong impact on
SSD data integrity
and operational
reliability.
This article will help you understand why some
SSDs which (work perfectly well in one type of application) might fail in
others... even when the changes in the operational environment appear to be
negligible. |
|
| | |
| . |
| SMART is new competitor in
STEC-class enterprise SSDs
|
Editor:- February 22, 2012 -
when you've got a memory business which also designs SSDs that creates hard to
reconcile business tensions.
Success in the memory market comes from
caution and long term planning to enable survival in the inevitable feast and
famine memory business cycles.
Success in the SSD market comes
from...
Well it's too early to say definitively what it comes from
- but investing in your own IP and understanding a small set of focused
customer application cases better than anyone else - is a good starting point
(judging by those in the
top 10 SSD companies list)
and being prepared to do things which are different to the way that others are
already doing them may be a good plan too - (as long as you are eventually
proved right).
Several leading companies in the past, including
STEC and
OCZ, have found that the
best thing to do if you're the SSD part of a memory business is to forget about
those memories and do your own SSD thing.
The latest company to
follow this route is SMART Storage Systems
which has officially been spun out as a separate entity - it was
announced
today.
The SSD bit is the only bit of the company I've been
interested in - and probably the same goes for most of you too. So you could say
- what's changed? - apart from a few legal formalities.
Last week I
spoke to SMART's president John Scaramuzzo and
learned more about the company's new enterprise SSD controller - which is used
in their new
Optimus
Ultra (a SAS SSD)
launched today. The new controller has reliability characteristics above and
beyond the industry standard products - from
SandForce - which
SMART also uses. SMART's new SSD design - like those from
STEC - can guarantee an
industry leading level of SSD write longevity - while using low cost consumer
grade flash. |
 |
Some tier 1 storage
customers have been sufficiently impressed to qualify the new SSD family in
their systems. | | | |
| . |
|
| |
.. |
| The first thing I asked
NVMdurance's CEO - Pearse Coyle was - how does your work relate to a news
story 2 years ago when SMART said they were getting 5x better endurance from
standard SandForce driven SSDs using magic numbers learned from their adaptive
R/W controller pool? |
| the context of
NVMdurance's machine learning (July 2016) | | |
| .. |
 |
shown above - an
ad from SMART Storage
which ran on StorageSearch.com
during the first half of 2013. |
| ... |
|
|
| .. |
 |
Above - a banner ad from
SSD market
history (June 2013) |
| .. |
| "SAS SSDs stood out as
the stellar performer in the enterprise segment with 69% revenue growth..." |
| SSD news - March
20, 2014 | | |
| ..... |
 |
| .. |
| SanDisk, SMART and
enterprise SSDs |
Editor:- August 22, 2013 - SanDisk today
announced
it has completed its acquisition of SMART Storage Systems
whose president John
Scaramuzzo will now assume the new role of senior VP enterprise
storage solutions at SanDisk - reporting to Sanjay Mehrotra,
president and CEO.
Editor's comments:- I've been reporting on SanDisk's metamorphosis on
the way to becoming a serious enterprise SSD company since long before they
acquired Pliant
Technology (an SAS SSD maker) in
May 2011.
Lessons were learned from that - and a transformation seemed to take place in
SanDisk's thinking triggered by their acquisition of
FlashSoft in
February 2012 -
which enabled SanDisk to get much better clarity on what was really happening
in a wide range of enterprise SSD users' sites - as a result of feedback coming
from customers using its SSD software.
The 4 key assets which SMART
brings to SanDisk in the enterprise SSD context are:-
- enterprise grade (world class)
adaptive R/W
flash SSD technology.
SMART has already demonstrated - in
launched products - that this technology is
scalable across
a wide span of cost, power consumption and IOPS use cases.
And at the
fab level adaptive R/W IP increases
SSD efficiency
(usable SSDs per wafer).
- an attractive (tier 1 validated)
SAS SSD and enterprise
SATA SSD product mix
which has already displaced competitors in many leading server and storage
oem design wins.
- an embryonic new type of SSD -
memory channel
storage - which aims at the market space of fast
PCIe SSDs - and which
- if successful - could change the future mix of motherboard memory used in
servers.
- a team of technical, sales and marketing people with a long track record of
successful product innovations in the mission critical flash SSD market.
Unlike the other
Top 10 SSD company
acquisition which is currently in the pipeline (Stec by
WD which may need a
lot of reworking to make it fly) - the SMART product lines within SanDisk
have already been expanding their reach of new customer destinations.
If
you look at what LSI
did with SandForce -
I think that provides a better idea of the future scale and speed of ramp up
which you can expect to see with the new enterprise SSD business in SanDisk.
| | |
| .. |
|
|
| .. |
| SMART samples first
ULLtraDIMM SSDs |
Editor:- August 8, 2013 - SMART Storage Systems
today
announced
it has begun sampling the first memory channel SSDs compatible with the
interface and reference architecture created by Diablo Technologies.
SMART's
first generation enterprise
ULLtraDIMM SSD
(ULL = ultra-low latency) can be deployed via any existing DIMM slot and
provides 200GB or 400GB of enterprise class flash SSD memory with upto 1GB/s and
760MB/s of sustained read/write performance, with 5 microseconds write latency.
Throughput, IOPS and memory capacity all scale with the number of ULLtraDIMM
deployed in each server. |
 |
Editor's comments:- With the current
design -only one DIMM slot in each server has to be reserved for conventional
DRAM. Apart from that constraint any DIMM slot can be used for either flash or
DRAM as deemed necessary for the application.
For more about the
potential of this technology, the thinking behind it and the competitive
landscape relative to PCIe
SSDs etc see my earlier articles on the
Memory Channel
SSDs page. | | |
| .. |
| SGI's past and present SSD
partners |
Editor:- June 12, 2013 - When SMART told me
they would today be
announcing
that their
CloudSpeed
500 (a SATA SSD
which uses the company's own brew of
adaptive R/W
IP and is rated at 1.2 DWPD) has been certified by SGI for use in its storage
appliance platforms - they asked if I had any questions?
"Just
one" - I said. "How many of these CloudSpeeds do they think SGI
will buy in the next year?"
Which - as you may guess - SMART
declined to disclose to all the competitors among you - who will also be
reading this. But it was worth a shot. As it is - I'm left without much of a
story.
So the only interesting thing I can say is that
2
years ago SGI was offering Virident's PCIe
SSDs in its servers.
And - judging from the list in
SGI's SSD page I
think SGI have also used Stec's
3.5"
RAM SSDs in the past too.
Which all goes to demonstrate what
I've said before in my enterprise
SSD article - that no single enterprise SSD (or supplier) can economically
satisfy all the needs of server customers. | | |
.. |
| 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.
So today
Diablo
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.
Editor's
comments:- there are a lot of implications for 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
latency?
- 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
new SSD?
- when will the first products be ready?
I've about these matters
and more in a new blog on StorageSearch.com -
Memory Channel
Storage SSDs - will the new ultra low latency SSD concept fly? - should you book
a seat yet? | | |
| .. |
|
|
| .. |
| SMART proliferates adaptive
DSP IP in SAS SSDs |
Editor:- July 5, 2012 - SMART Storage Systems
recently
announced
it's sampling yet another new variant of
SAS SSD which uses
adaptive
flash DSP technology - the
Optimus
Ultra+ is a 2.5"
SSD with 100K/60K R/W
IOPS - which
can endure
50 full random drive writes per day for a period of 5 years using commercial MLC
NAND flash technology.
Editor's comments:- This is SMART's 3rd
turn of the screw this
year already in leveraging its new controller IP in enterprise SSDs - and
the company has published various pretty graphs for editors, analysts and users
which identify the different sub markets and use cases within enterprise SSDs
where differing amounts of SSD write life can be economically and sustainably
deployed. (It's an SSD component maker's view of sites inside
enterprise SSD silos.)
More
guaranteed SSD writes generally mean more
cost - because - in
the limit - when all the write
attenuation tricks have been done - the extra endurance in classic flash
SSDs either comes from
overprovisioning
or RAM cache
which add to the component count. But for the pioneers in DSP enhanced flash IP
- a new degree of design freedom comes from applying new
flash care
tonic to change all
the rules.
Unlike human cures for baldness which are applied when the
symptoms start to appear - SSD
baldness cures have to be applied from birth - and the strength of the
tonic has to increase with age to get the best results.
I discussed
the business development opportunities of this new technology trend with Mike Lakowicz,
SMART's - VP of Marketing.
Mike said that their ideal customers for
these enterprise SSDs are big oems where it can cost millions of dollars to get
products qualified.
SMART's
new controller technologies enable them to customize an SSD which matches
the performance, endurance and operating power needs of enterprise oems -
at any point in a wide spectrum of use cases - in a manner which is efficient
(lowest cost memory bill of materials) from the same core set of parts.
This sounds similar to what
STEC was telling me
about their CellCare
last year. But SMART's documents make what Mike called the "rational"
design rule choices along the SSD envelope much more explicit and easier
to focus in on.
To me - this seemed like a pragmatic and well
engineered set of marketing propositions for the type of customers that SMART is
aiming at.
So I asked again (I had asked SMART's President - John
Scaramuzzo earlier this year) what are their plans for designing a fast-enough
PCIe SSD using the DSP
technology?
Mike Lakowicz explained that the taxonomy of PCIe SSDs in
the market today wasn't clear. (I agreed - there are many different
types of
functionality in the PCIe SSD space - so it's hard for a new entrant to
decide which segment within PCIe SSD will work best as a sustainable business.)
Mike
explained that as a supplier of legacy compatible storage SMART would prefer to
see more stability and the emergence of clear standards within the PCIe SSD
market before they launch any such product.
I said to Mike that I
thought their VC owners should be happy with what SMART was doing - because
even discounting their SSD products - it was clear that within the
SSD controller market
- the new adaptive DSP technologies were going to attract a premium value once
enough analysts understood how this technology could permeate the SSD market.
But I also warned that in 2-3 years every flash SSD company would be using this
type of technology.
To which Mike replied - that companies like SMART
which has already been using this technology for several years - would still
have an advantage over those who were coming in with their first products at the
20x endurance level - for example. | | |
| .... |
| SandForce driven SSDs get
5x SMARTer |
Editor:- April 26, 2012 - SMART is close to
sampling a new
fast-enough (enterprise
component) 2.5"
SATA SSD - called
the CloudSpeed 500 - which uses a controller from LSI/SandForce - and
I can already see a lot of you starting to yawn or click away or look at
something different on this page or out of the window - which is a good
argument for putting
duct tape over that camera lens on your browsing device BTW.
"Let
me cut you off right here!" - I said to SMART's President - John
Scaramuzzo last week - right after we had done with the conference bridge
pings and greetings. "Because if all you want to talk to me about is
just another
SandForce
driven SSD then this is going to be a very short conversation. Can we talk
about something else instead? - because this is not going to get onto my news
page - which has a big yellow note on it explaining why."
Just to
drive home my point I followed up by saying - "There has to be something
different in a SandForce inside announcement to make it news for me and my
readers. So for example - if STEC
or SanDisk launched a
SandForce based SSD that would be news."
John said - "I saw
your note - and we have a new twist on this which no one has done before."
So
I listened. He was right. Here goes...
SMART have learned a lot -
about better ways to interact with cheap consumer flash - from the experience
they accumulated designing the
adaptive DSP
based SSD controller which goes inside their Optimus SSD (which I wrote
about in February).
That gives them raw stats for factors like
write pulse length,
endurance
and data integrity
- when they are in this closed loop adaptive DSP managed flash environment -
with brand X flash memory and generation Y process. In a new wave DSP SSD
controller those figures are going to be different in different parts of the
memory and different over time - but it gives you an idea of what the memory can
do. (With a view which even the original memory manufacturer doesn't have.)
What
SMART have done is ask - if we know this stuff about this type of memory - and
precondition it during manufacture or first boot (with different parameters
to the default setting done in the fab by the chip maker) then can we get better
reliability even when using an SSD controller which doesn't see the memory the
same way and runs open-loop?
They found the answer is - yes! They can't
get exactly the same reliability that they would get with their Optimus
controller - but they do get a 5x reliability boost by using the new
setups with unmodified and lower cost SandForce chips.
SMART has
to back off the preset numbers - to allow a safety margin - because they're
optimizing the numbers for the whole population of memories they will use -
rather than optimizing each part of each memory.
To use an analogy
here - your average safe running speed is different for you as an active kid
compared to you with a broken ankle or you at 3 score years and 10. And the
average safe running speed for an olympic athlete is different to the
population as a whole. Traditional SSD controllers can't tell if this piece of
flash is an olympic champion or not - and unlike new wave controllers - they
can't give them juice.
And here's another analogy. I changed my
behavior in this interview with SMART from what I thought I would to - to do
something different instead as a result of what I learned. (And I hope it was
worth it for you.)
SMART's trick with the SandForce controllers is like
using Dolby correction with a 1980s cassette tape. Whereas SMART's trick with
its Optimus controller is like having a dynamic sound equalizer built into an
iPod.
What's this got to do with SSDs? There's a new wave of flash SSD
management - based around combining adaptive flash interaction with DSP. In one
way it's not that new - because some companies have been working on it for
years. What's new is that more vendors are doing it - and in different ways.
|
 |
Now coming back to the "cloud" thing.
Attaching the word "cloud" to an SSD name isn't new. I recall that
OCZ have done it before -
and there are many other examples. Maybe there should should be a new editor's
filtering rule about SSDs and cloud names. Don't look now but here comes
another one... | | | |
| . |
| |
