key SSD ideas which changed in 2014 /
endurance - the
|Skyera's new skyHawk FS
Editor:- October 29, 2014 - It was with great enthusiasm last
week that I looked forward to having my first conversation with Frankie
Roohparvar, CEO of Skyera.
For me the interesting aspect wasn't that it was Skyera - because even though I
think they are really important - I wasn't really expecting any earth
shattering new announcements from them till next year. And in that respect I
guessed about right. As you can see for yourself by reading their
release about the new skyHawk FS - a 1U very efficient rackmount SSD.
(More of which later).
Frankie has helped to create flash history.
So the first thing I said to him was that I'd really enjoyed his
history of flash memory
and storage (video). Which if you haven't seen it - and if your network
admin permits you to watch youtube - is very worthwhile.
small spattering of what we talked about.
- what's different
about the new product?
Skyera's new box is an incremental
improvement on what they did before.
Twice as much capacity in the
same 1U space.
A smaller number of fast 10GbE ports - replaces the old
plethora of slow ports.
Addition of another hot-swap power supply.
said that the 2nd version of any product benefits from learning the reactions
to what came before.
So I see this version of Skyera's box as an
interim marketing course correction - due to better understanding of
customer needs - rather than as being a technology advance.
- the mobile data center
If you see someone carrying a 1U
rack in a pack onto a plane - and find out later this box has 136TB of raw
flash, which can do 400K IOPS on a SAN at under 1 micro-second latency - even
with one power supply down - it may be Frankie - or someone else from Skyera.
the product is from a competitor - however - and you see just one single person
carrying the same storage payload - you may want to reconsider whether it's safe
to get on the plane. (Especially if their shirt looks tight and their skin
color is green.)
Size weight and power (swap) savings are some of
the consequences of Skyera's relentless pursuit of the ultimate
efficiency in SSD
array design - which was something I wrote about last year - in my article
acronym - swap - comes from
- and marketers in various companies adorn it with their own customizations -
depending on what point they're trying to emphasize. You can see how Skyera
stretches this to
- in the skyHawk FS briefing paper (pdf)
Frankie said the mobile
data center - such as is enabled by having hundreds of terabytes of flash in a
Hummer for example - is an interesting
enterprise market segment in which Skyera's product enables entirely new
types of mobile deployments.
- re unified storage and gentlemen's clubs
One of the
dimensions in which Skyera has invested resources recently - is the
population of bullet points associated with software compatibilities.
trick here - if you're a marketer - is to make the list so long - nobody has
the time to read it. Instead if your
attractive then people will ask you if you have what they want - and from
those conversations you can decide which features should be prioritized.
this theme - almost a null bullet point - which I'm seeing from many flash
box vendors is the phrase - "unified SAN / NAS storage".
unsurprisingly it's in Skyera's blurb too.
So I said to Frankie - in
some ways - by doing this (and the other compatibility stuff) - Skyera is
putting a lot of effort into joining something which looks a lot like an old
fashioned English gentleman's club (think- Forsyte Saga or Sherlock Holmes).
But we know that the current members of the club are
so old they will
die soon anyway. So is it worth it?
Frankie laughed at that.
And that's how I raised the topic of unified storage...
I said - When it comes to unified storage I see that Skyera has the usual 2 -
(SAN + NAS) but I expect something better from Skyera than the usual 2 out
of 3 which the old timers do. So... what about
- I said - if a customer has invested in some of your state of the art boxes
(which we know inside have latency characteristics similar to the best server
based accelerators) it would be reasonable for them to expect to be able to do
big memory type software APIs via PCIe too - using next generation software
such as SanDisk's
If they own petabytes of flash in your (Skyera) boxes-
why should they be forced to go and buy more flash from someone else (in a
server box for example) - to get the job done?
Frankie said - I
wasn't the first person to have raised that issue.
Skyera's customers have mentioned those needs too.
thinking about it.
So my guess is that in 2015 we can expect to see
that kind of access model to the flash array - as an API supported, low
latency, big memory type - via interfaces like
InfiniBand - being
added to leading-edge flash storage racks.
That will make "unified
storage" a more meaningful concept.
Who Needs 10Gbps USB?
Editor:- October 29, 2014 -
SSDs are at the forefront of the thinking in a new article today -
Really Needs USB 3.1? by Eric
Eric's blog sketches out a 5 years into the future application
picture for this new (10Gbps) iteration of the USB connected story.
also:- storage interface
chips, market research
McObject expands reach of in memory database for serious embedded
Editor:- October 28, 2014 - First 2, then 3 and
finally - 4 interesting things caught my eye in
news about version 6.0 of
eXtremeDB - an in-memory database system from McObject
- Data compression. This release adds data compression for both in-memory
and on-disk databases. Once upon a time compression was a value add feature in
some products - but now in the SSD age when compression is almost latency free
- it has become a must-have on the feature list - especially for embedded
- Avionics platform support. This upgrade adds compatibility with
River Systems' VxWorks 653 COTS platform for delivering safety-critical,
integrated modular avionics applications.
- More flexible transaction scheduling. Applications using eXtremeDB's
multiple-user, single-writer transaction manager can override the default FIFO
scheduling policy within priority levels to favor either read-only or read-write
""Demand for distributed
query processing cuts across market segments, but is especially relevant to the
automation and control field, where eXtremeDB is historically strong"
said McObject CEO and co-founder Steve Graves.
- Distributed query processing support. eXtremeDB partitions a database and
distributes query processing across multiple servers, CPUs and/or CPU cores -
which can accelerate performance.
SSDs, military SSDs,
VIMMs, controllers and firmware's place in the box
October 27, 2014 - As Violin's
flash SSD controllers -
- are such an important part of their system offerings - and as they have
established themselves as a de-facto proprietary standard in the enterprise
(in the same way that if you ship enough quantities of anything over several
memory product generations and over 6 years expanse of time it is an
industry standard - despite having no
ORG as custodian of the
That was my pretext (6 months ago) for suggesting
Herzog, Violin's CMO - that readers like you - might be interested to
know how many VIMMs the company has shipped.
I said this would enable
prospective customers to compare the installed weight of this controller
architecture to other things which they can get numbers for - such as units of
PCIe SSDs etc.
don't have those VIMM shipment numbers - but I found it refreshing to see a
recent blog from Violin which talks again about the hardware inside its arrays
- coming as it does after a stream of announcements from Violin which have
focused mostly on software compatibilities - in which almost the exact same
words could have been said of other competing rackmount SSD despite them
having completely different innards.
does firmware come from? - by Erik Ottem,
Director of Product Marketing - Violin - answers this question more from a
business perspective than a technical software writing angle.
point is that Violin's biggest systems competitors have made up their deficits
in raw flash array architecture by
buying companies or
licensing 3rd party software - whereas Violin controls everything about the
way that the thousands of flash memory chips in each
architecture system are connected and managed. ...read
Editor's comments:- While nothing in Erik's
blog is really new - for anyone who has known the company for a long time - I
think it's a healthy restatement of what makes this company different from most
of its competitors.
Having said that - owning your own controller
architecture - also carries its own consequences in terms of the
cost of the raw
memory you have to buy too.
flash backed DIMMs - new directory on StorageSearch.com
October 21, 2014 - Although StorageSearch.com
has been writing about flash
backed DRAM DIMMs since the first products appeared in the market - I didn't
think that subject was important enough before to rate a specific article or
market timeline page.
That's unlike -
SSDs - which has become 1 of the top 10
viewed by readers after having had its own directory page here since
Despite my lack of initial enthusiasm for bybrid DIMMs (or
for that matter too) I realize that sometimes a
market is defined as
much by what it isn't as by what it is. Which is why I have relented.
so - to help clarify the differences between these 2 types of similar looking
storage devices (one of which I think is much more significant than the other
- but both of which are important for their respective customers) I have today
created a directory page for
hybrid DIMMs etc -
which will act as the future launch pad for related articles.
tests Fusion-io's Atomic PCIe SSD
Editor:- October 21, 2014 -
Earlier this year - in
June 2014 - and
just weeks before SanDisk
announced it was acquiring the company - Fusion-io launched a
new generation of PCIe
SSDs - the Atomic Series - which was an overdue catching up exercise to
use cheaper 20nm flash.
At the time the headline product specifications seemed to me to be
adequate - rather than spectacular.
It's probably safe to assume they
were intended to restore the competitiveness of Fusion-io's products rather
than push performance boundaries.
But what are they like? And how
do they compare?
A new article -
ioMemory PX600 Review - published in StorageReview.com provides the
answers to how it performs and how it compares on popular OS platforms and
benchmarks. ...read the
how fast can your
SSD run backwards?, SSD
testing & analyzer news
HP plus EMC? - the SSD DNA mix from Jurassic Park
October 20, 2014 - I'm still here. I've been working on a new article (see
PS - I didn't think it was worth commenting here earlier about
that EMC / HP merger story - which is covered in many other places such as
reasoning was this.
No matter how you mix the DNA from 2 dinosaurs -
the result is unlikely to be a mammal.
and metaphors in the storage market and
animal brands in the
SSDs are made of this
Editor:- October 14, 2014 -
Without memory - there would be no SSDs.
And while naturally the
emphasis in SSD thinking is mostly on - how can we do useful and affordable
things with SSDs? - despite how terribly flawed the
raw material is which we
have to work with (which leads you to
and software) - it can
nevertheless be strategically useful for SSD specifiers to sometimes brace
themselves for a deep dive down into the cold details of how much better (or
worse) those raw memory characteristics are going to get - so you can
anticipate future developments.
This week the best place to look is
HA SSD arrays - are now mainstream
13, 2014 - I've long had an abiding interest in the architecture of fault
tolerant / high availability electronic systems - ever since learning that such
concepts existed - when (in about 1976) our digital systems design
R G 'Ben' Bennetts at Southampton
University suggested we should read a paper about how
NASA's Jet Propulsion Labs used triple
(I can't remember the details of that paper - but
the JPL people and their collaborators and descendants have never stopped
inspiring and writing a rich literature about the design aspects of computer
systems which operate a long way from a service engineer.)
In the early
part of my career - such ideas were good to know about - but far too exotic and
expensive to incorporate into most products. But I was reminded about them in
the 1990s - when in the publication
which preceded StorageSearch.com - some of my customers were advertising
their FT/ HA SPARC servers for the telco market.
The more you
investigate the architecture of FT/ HA computer systems the more you realize
it's a philosophy rather than a technology which you can implement as a plug
and play inconsequentially within the cost goals of mere mortals.
results are always compromises - which balance reliability (aka functionable
survivability) against other tradeoffs - such as performance. (And
performance itself has many internal
fault tolerance too.)
Violin's 6000 SSD and HA
years ago (in September
2011) when I was talking to Violin's CEO (at that
Basile about the launch of Violin's first 6000 series (the first no
single point of failure, competitively priced, fast flash rackmount SSD) he
expressed some concern about how I would tell you (my readers) what was unique
about this product and signal whether it was relevant to you or not - as it was
competing for attention with thousands of other SSD stories for applications
ranging from phones to drones.
I didn't see that as a problem - because
my readers are smart - and I had been publishing a directory page dedicated to
But just to make sure that the systems embodiments of
FT/HA/SSD architecture from a growing base of competitors didn't get washed
away by other stories - I launched
a dedicated ft/HA
enterprise SSD directory in
January 2012 - to
serve an emerging base of reliability focused readers - which in those days
measured around 10,000 readers / year in that niche topic. (Until recently
HA SSDs have rarely entered the
top 30 SSD
articles viewed by my readers.)
But something in the market has
I noticed this week that the topic of
has risen to be 1 of the top 10 topics that you've been looking at this month.
Which means it's mainstream.
Looking back at other past niche
10 years ago I didn't think that more than a few hundred
people would be interested in the intricacies of
And to begin with - SSD vendors were nervous about even acknowledging that
there was such a thing as SSD wear out. Now you can't shut them up. They all
want to show you how clever they are at handling it
types of flash memory and
different generations of arcane
flash care schemes spawned a huge industry literature of understanding and
misunderstanding - so I wouldn't be surprised if the enterprise FT/HA flash
array market now started to do something similar.
PS - After a
communications gap of 37 years - I exchanged some emails with my old
university lecturer - Ben Bennetts while writing this - to see if I had
remembered things correctly.
He said - "Yes, that was me. I
lectured on fault-tolerant systems and JPLs Self-Test And Repair, STAR,
computer, based on triple modular redundancy, used to feature in my
So that enables me to pin point the original
source of that inspirational IEEE Transactions paper about fault tolerant
computing - which I remember having read in 1976 (although I haven't read it
since) to Prof.
Algirdas Antanas Aviienis - whose visionary work on - what is
today called - "Dependable Computing and Fault-Tolerant Systems" -
Samsung versus Samsung
3D MLC v 2D TLC (wrapped in a
Editor:- October 10, 2014 - What are the differences
seen at the SSD level when you compare Samsung's
nand (32 cells stacked vertically in the chip
MLC - with 2 bits
per cell discrimination) with 2D (planar) TLC (1 cell high fabrication but 3
bits per cell discrimination)?
versus TLC are different memory manufacturing techniques with different
technology roadmaps - but both are out in the wild of the market in
SATA SSDs at the
Handy - the SSD Guy - has published
a new blog -
Samsung's V-NAND and Planar SSDs which compares the possible differences
in performance and power consumption as viewed in the context of 2 SATA SSDs -
respectively - which use the same controller.
Re IOPS - Jim says "the
850 performs consistently better than the 840 at all block sizes for constant
streams of writes and for a 65:35 read/write mix. At the 100% read workload the
performance of the two SSDs is roughly equivalent."
comments:- Although small superficial differences emerge in the benchmarks -
I don't think they're significant.
Because even if you introduced 2
SSDs using the same memory technology at different dates for the
consumer market -
you might - as a product marketing decision - choose to tweak the controller
biases a different way.
Instead what I find more interesting is
how closely Samsung was able to match 2 SSD characteristics using different
memories. This is a shrewd strategic signal to their oem customers that they
know enough about the internals to deliver a consumer SSD with a given set of
characteristics regardless of what is the most convenient memory technology to
take from their fab.
In other markets - such as
industrial SSDs - being able to retain and deliver precisely the same SSD
power and performance envelope - despite using different types and
generations of raw flash
over a 7 year or so market period - has developed into a more refined
Netlist asks court to shut down SanDisk's ULLtraDIMM production
and recall units already sold
October 10, 2014 - Netlist
it has filed a motion for
injunction which seems to be intended to restrain Diablo and its
flash-side SSD integration partner SanDisk from any
further manufacture or sale of the ULLtraDIMM (memory channel SSDs).
Court has set a tentative date of December 2, 2014, for a hearing on Netlist's
motion. If granted this would immediately shut down any further manufacture and
sale of the ULLtraDIMM. Netlist has further asked the Court to order the recall
of any ULLtraDIMM products previously sold.
This series of legal disputes has been going on since last year. If you haven't
read those stories - the essential story is something like this.
some time in the past Netlist and Diablo had a technology supplier agreement -
as a result of which - Diablo had access to Netlist's IP related to
minimizing the capacitive load of complex circuits susch as controllers when
they sit on a DRAM style of bus. Netlist doesn't have any products which are
similar to SanDisk's ULLtraDIMM - but Netlist claims that the ULLtraDIMM
design has used its patented interface technology without a license. Countering
that - Diablo says it did get the rights to use some of the DRAM load
interface technology - and that in any case - this aspect of the design is
not the essential defining characteristic of their flash SSD architecture.
Diablo's CEO - Riccardo
Badalone retorted to Netlist's latest legal move saying this - "After
a year of court proceedings and months of discovery, Netlist still cannot
decipher how Memory Channel Storage works, much less substantiate that it
infringes on any of their IP."
Efficiency is important for web scale users - says Coho
October 9, 2014 -
as a file system - a web scale case study - a new blog by Andy Warfield , cofounder
and CTO - Coho Data
- made very interesting reading for me - as much for revealing the
authoritative approach taken in Andy's systematic analysis - as for the object
of his discussion (Facebook's storage architecture).
It reveals useful
insights into the architectural thinking and value judgments of Coho's
technology - and is not simply another retelling of the Facebook infrastructure
it you may get different things out of it - because it's rich in raw
enterprise ideas related to
dark matter users.
All of which makes it hard to pick out any single quote. But here are 2.
- re -
match between enterprise products and user needs
says - "In the past, enterprise hardware has had a pretty hands-off
relationship with the vendor that sells it and the development team that builds
it once it's been sold. The result is that systems evolve slowly, and must be
built for the general case, with little understanding of the actual workloads
that run on them."
There are many more I
could have chosen. ...
read the article
- re efficiency
Warfield says - "Efficiency is important. As a rough approximation, a
server in your datacenter costs as much to power and cool over 3 years as it
does to buy up front. It is important to get every ounce of utility that you
can out of it while it is in production."
You don't need to worry about the endurance of our FlashSystems -
Editor:- October 7, 2014 - Worried about
"None of the thousands of
products (fast rackmount SSDs) which IBM has shipped has ever
worn out yet! - says Erik
Eyberg, Flash Strategy & Business Development at IBM - in his new
storage reliability: Aligning technology and marketing. "And our
metrics suggest that will remain true in almost all cases for many, many years
(certainly well beyond any normal and expected data center life cycle)"
goes on to explain that's the reason IBM can now officially cover flash
storage media wear-out as part of its standard IBM FlashSystem warranty and
maintenance policies - without changing the prices for these services.
his blog has a
to a white paper about the reliability architecture underlying this product
(although it's behind a sign-up wall - which seems counter productive to me.)
comments:- Don't expect all other flash array vendors to follow suit (with
no cost endurance guarantees) - because this product range from IBM is based on
design rules and memory reliability architectures experience in FC SAN
compatible enterprise SSD racks which have evolved since the 1st generation
RamSan from TMS (in
2000). And for more than a decade
using other popular enterprise storage interfaces.
Holly Frost - who founded
Texas Memory Systems - and who was the CEO when TMS was acquired - told me a
revealing story about TMS's policies concerning the reliability of their SSD
systems and customer care procedures.
This conversation took place
in December 2011
- when the company was launching its first high availability SSD - which
became the basis of IBM's FlashSystem.
It still makes interesting
reading today. You can see it in
this article -
in the right hand column - scroll down to the box titled - "no single point
of failure - except..."
Toshiba orders 1 million SSD controllers this quarter from Phison
October 7, 2014 - A
report on Digitimes
says that Toshiba
has ordered "about one million"
SSD controllers from
delivery in the current quarter.
Editor's comments:- You can
get an idea of who else uses Phison's
Kingston etc) - and for what purposes - on
SolidFire gets another $82 million funding
October 7, 2014 - SolidFire
it has closed an $82 million Series D round of funding, bringing its total
funding to $150 million.
New investor Greenspring Associates led the
round along with a major sovereign wealth fund, with participation from current
investors NEA, Novak Biddle, Samsung Ventures and Valhalla Partners. SolidFire
will use the additional funds to extend its global reach.
building blocks of SolidFire's SSD systems are 1U
iSCSI rackmount SSDs
which include 10x 2.5"
SSDs. At that level it's the same as 100 or so other competing systems.
If you want fibre
channel access - you add a special 1U adapter rack to the native IP array.
So it's expensive - but keeps the unit costs of the most common building blocks
down - compared to including native unified storage in each rack. So in the case
of a big installation - it's a reasonable cost optimization tradeoff.
key difference is
software architecture and the fact they use a
architecture type of RAID
- which they call "no-RAID".
In SolidFire's no-RAID (which
is really big RAID) - the data is more widely dispersed across the drive
population than in classical (small architecture) RAID.
is much less disruption to data access and
performance when a drive
because SolidFire's software can manage upto about 100 racks as a raw storage
resource (1,000 SSD drives) - so the impact of a single drive down is small.
Users also have a high degree of flexibility as to how they micro manage
different virtualized segments of storage to meet their different QoS goals.
also:- VCs and SSDs
We need new software abstractions to efficiently handle all the
different emerging flavors of persistent enterprise memory - says SanDisk
October 3, 2014 - New enterprise software abstractions are needed in order
to efficiently utilize
all those unruly developments in
And laying the
framework for those ideas - along with some practical suggestions for where
applicable solutions might be coming from - is the theme of a recent blog -
Emergence of Software-Defined Memory - written by Nisha Talagala,
Fellow at SanDisk
- who (among other things) says:-
"We're seeing excitement build
for a new class of memory:- persistent memory - which has the persistence
capabilities of storage and access performance similar to memory.
"Given this richness of media technologies, we now have the
ability to create systems and data center solutions which combine a variety of
memory types to accelerate applications, reduce power, improve server
consolidation, and more.
"We believe these trends will drive a
new set of software abstractions for these systems which will emerge as
software-defined memory a software driven approach to optimizing memory
of all types in the data center." ...read
See also:- are you ready to
rethink enterprise DRAM architecture?
NetApp says - the time for taking risks with enterprise flash
startups is over
Editor:- October 1, 2014 -
of the flash startups is the provocative title of a recent blog by Craig Alger at NetApp - who
asserts that the "brief window of time where fast and agile (enterprise
SSD) startups can get the jump on large, slow manufacturers" has now
Craig questions how startups like
Tegile can expect to
compete now that "titans of the industry" such as NetApp, and EMC
(and by implication IBM,
HP too) have gotten their
flash toys acquired,
oemed, licensed and integrated neatly within vast product catalogs?
comments:- If you agree with Craig's premise - that all the disruptive
innovation is now over - then you'd probably also agree that it's not worth
taking risks with new enterprise SSD startups. Just stick with the big safe
vendors and you'll be OK.
You won't be surprised to learn I disagree.
This is what I said to Craig (by email) yesterday...
I saw your blog - Demise of the Flash startups - and might comment /
post about it. Liked the middle but disagree about the conclusion.
I would agree - if the pace of disruptive change in enterprise SSD
architecture had slowed down and if it already did solve most problems.
with a lot of very big
changes in utilization still to come - the potential market size for
genuinely innovative enterprise SSD startups (drives, systems and software) is
bigger than it was before.
"So there will more startup
companies to acquire, license from and compete with. No one's got a whole
stable solution architecture and credible roadmap yet. At best current flash
systems are stepping stones to somewhere else."
Oh - and if you're
wondering - which is the part of Craig's blog I liked the most (apart from the
cleverly provocative title) it was where he says "...those SSDs aren't as
expensive as they
used to be." ...read
|SSD news today|
the Top SSD Companies
trust SSD market data?
sudden power loss
hidden segments in the enterprise
pressures and projections in enterprise flash
guide to semiconductor memory boom bust cycles
why did Memblaze launch its new 14µS latency PCIe SSDs in Europe
|Editor:- October 31, 2014 - Something I saw in
a press release this week by Memblaze Technology
made me think I'd better pause to check the significance behind the words
before mentioning it here.|
The story - from their
PR agency -
A3 Communications - was about
the launch of a new PCIe
SSD - the EBlaze3 - in Europe - which contained the phrase.
will be the first time the technology is available outside China."
wondered if - behind that statement - there might have been an implied message
that the new product will be sold preferentially in Europe - rather than the
US - as we have seen evolving in stories about one of the world's biggest
enterprise companies - to have emerged from China -
this wording - and the showing of Memblaze's EBlaze3 at a trade show this
week in Germany - simply due to the timing of having booked booths at trade
shows and the availability of the new product - rather than an expression of
preference about which market - the US or EU - is more important for their new
enterprise SSD to be seen in?
So I asked Ron MacLeod - who is VP
of marketing for Memblaze in Europe - to clarify.
Ron said - "The
product will be available in the US in the same way it is available in Europe.
And with regard to timing indeed it was just coincidence that the Powering the
Cloud event was running and at which was the first time Memblaze was public in
Europe. There is no favouritism over markets except the US is potentially larger
- so no hidden messaging. We value the European business and other potential
global partners equally."
Editor's comments:- That's good to know. So - my apologies to readers
for having sat on this story for a few days longer than was necessary.
- The Memblaze EBlaze3 (upto 2.4TB in HHHL and 4.8TB in FHHL single slot) is
a new version of the
card - which is aimed at, Hyperscale and OCP (Open Compute Project
It offers an ultra-low latency of 14µs and
enables customers to define and alter key settings to ensure optimum
performance in their specific applications and platforms.
|As little as 2 days
retention after 0.4 hot DWPD for 5 years in modern MLC nand and some other
things worth knowing about emdedded SSDs|
|Editor:- October 23, 2014 - When choosing an
SSD form factor and interface for a new embedded / industrial project - in most
cases that determination will be obvious - and driven by considerations such
- is the new project like something you did before?
- and what would you like to change based on what you learned from that?
Today I read a tour guide of the
of sizes and interfaces available in small embedded SSDs (pdf) - from Virtium - which says
among other things...
- or the availability of new smaller sizes or lower power SSDs or faster
SSDs or denser (more efficient) SSDs which in themselves can make new
application directions feasible.
"the 10-pin embedded USB module is not
officially regulated by any industry standards body, but thanks to industry R&D
in this case, "rip-off and duplicate" - OEMs can source mechanically
equivalent modules from multiple sources."
The above paper is
several years old old - but still contains many relevant ideas. And the reason
I noticed it now was because it was easier to find than it had been before -
as it's one of a
of papers on the theme of selecting embedded SSDs according to design and
environmental considerations which Virtium has collected together in a new
resource page this week.
how does DWPD and retention come into
Within this set - a more recent paper -
considerations in SSDs (pdf) includes some stark graphs and observations
about data retention - which you should be aware of - even if you're not in the
"This shows the dramatic effects that temperature has on data
retention for given workloads. For the same 750 full drive writes (0.4
drive writes per day for 5
years), SSDs operated and stored at 85C will only have 2 days of data
retention, whereas those drives at 40C will have 1 year and those at room
temperature 25C will exhibit characteristics of nearly 8 years of data
the article (pdf)
| DRAM SSD
interfaces and PCIe fabrics are hotting up the top storage searches|
|Editor:- October 1, 2014 - Total
SSD article views on StorageSearch.com
grew 5% year on year in September 2014- despite all the changes in
Google algorithms and increased competition in the SSD market reporting space.
But what have readers actually been looking at?
reporting on the 30th quarterly
Top SSD Companies
later this month. That involves a lot of work, cross checking and writing
associated articles. All of which takes time. But what I can reveal today are
these observations - based on reader metrics.
- In Q3 2014 -
SSDs became the 2nd most popular SSD form factor which readers followed up
in articles and news stories. (#1 - in case you can't remember - has been -
since 2009 - PCIe SSDs)
What does that signify?
- In September 2014 - the 3 SSD related companies which our readers were
reading about most were:-
#1 - Diablo (DDR3/4 flash
#2 - A3CUBE
(PCIe memory fabric) and
#3 - Fusion-io (no
It won't come
any surprise to long term readers that there are still significant changes
coming in SSD enabled data server architecture.
accustomed to the idea that low latency flash inside servers has become an
essential part of the job description of any multi-user enterprise server - and
being offered a rich variety of competing alternative ways to bind CPUs and
storage with SSDs by the SSD
software market - the next natural questions for users and applications
developers to ask are these:-
- why do we have such low limits being set in directly addressible low
latency memory capacity?
- why should the performance in a single server box still dictate the
perfomance limits for critical data integrity and synchonization house keeping
tasks in strategic applications - when we have access to thousands of servers?
The roadmap vision I'm seeing emerge from
enterprise SSD developments in 2014 - is that while oems and users are being
offered more choices in form factors and flash memory types - each of which
adds to the raw confusion of which one is best to use - the mission statement
for the software developers and fabric enablers - or those who want to please
their investors - will
be to create SSDcentric platforms which enable these disparate pieces to be seen
as interoperable subsets of a bigger
- in which users can move freely across wide cost/performance boundaries without
hitting walls which restrict their freedom to expand in any direction they want
- why are our most expensive and fastest SSD enabled servers and storage
systems being forced to use different software to the cheaper ones we use in
But it will get more complicated than that.
early modern SSDs had to interoperate with legacy software and data storage in
order to justify their costs - future SSD software developers will have to look
at the messy patchwork of SSD accelerated servers and SSD SAN storage which
are being installed today as part of their future "legacy problem".
Seen from that angle - some solutions in the enterprise SSD jigsaw
puzzle box today - already seem to have better longevity prospects and
opportunities for future upcycling than others.