- editor mentions on StorageSearch.com and
|Editor's comments:- September 17, 2013 - PLX
is the secret sauce behind many of the advances we're seeing in high
availability PCIe SSD designs today.|
Who's who in SSD?
editor - February 9, 2012
PLX is 1 of many companies listed in
storage interface and glue chips
And especially in this context - PLX is the leading
supplier of PCIe
interface chips (extenders, routers, switches).
got to with SSDs?
These chips enable SSD designers to create simple
and complex fabrics right on the SSD card.
Just as in earlier years -
the SSD market learned to appreciate the importance of the
SSD controller and
in predefining the performance, reliability, cost, strengths, weaknesses,
capabilities, and limitations of SSDs - so too - it's important to realize that
some chipmakers make it easier to support specific subsets and supersets of all
the possible PCIe topologies than others.
Among the many oem
customers who PLX's PCIe switch technologies inside their
PCIe SSDs are
industry leaders you already know for example -
Due to the
emerging size of the PCIe SSD market - I expect that many more SSD companies
will be evaluating what's involved in PCIe for the first time - and trying to
get a sense of the direction this technology is going.
- educational resources
PLX has created a
big bunch of
white papers and videos which may be helpful for product marketers (to get a
feel for the performance numbers) and designers (to get a feel for what glue
technology is available to get them started quickly).
Even if you're
not in an SSD company - but just a user or analyst - these educational materials
will help you understand what are the possibilities in the PCIe SSD market today
- and what new types of SSD will these chip technologies support in the future?
rocking the PCIe SSD cruise boat
The PCIe material
from PLX which most impacted my thinking was not their stuff about
enterprise SSD designs (video). That's just a prettification of what we've
already been seeing in the market - interesting if you want to see a coherent
narrative - and well presented - but nothing to rock the architecture boat.
when I saw this video -
as a fabric for data center clustering (video) - my center of gravity moved.
an overview level it shows how PLX's PCIe extenders and switch chips support
lossless, software light, high performance CPU to CPU memory transfers -
and shared I/O - for upto about 200 nodes - as a mini-cluster. At the
next level of scalability you can interconnect these clusters
that so important?
If you take a
architecture view - it suggests that network physical interfaces are things
like Ethernet and
Fibre-channel - whereas
PCIe SSDs are usually
viewed as DAS (direct attached storage).
Right from the start of the
PCIe SSD market some pioneers challenged that view and pushed the boundaries
of what you might expect - with both marketing rhetoric and actual products.
Fusion-io referred to their original ioDrive as a "SAN" capable SSD.
It wasn't true for the original model - although as we now know it is true for
the architecture - but 5 years ago it was a way of positioning PCIe SSDs as "enterprise
Dolphin launched SSD
racks which were stuffed with their own design of PCIe SSDs - and could be
located 1,000 feet away from the host server using a fiber extended PCIe
interface. - Now you might think that's not a great distance - but I'd like to
remind you that Ethernet started as a departmental / campus standard too.
is a storage network?
To my way of thinking - if you can access
hundreds of servers and thousands of PCIe SSDs in fault tolerant modes over the
distance of a few cabinets using copper cable - and around your datacenter using
fiber - that starts to look distinctly like a network typology to me. And if the
physical layer is low latency / high bandwidth PCIe - then that could start
resetting some long held assumptions about how things get done - at the core
of big enterprise.
The PLX video I mentioned above - suggests that
systems designers could easily have Ethernet protocols as the management and
software layers above the raw chips.
and that brings us to
software for PCIe SSDs
Now it's easy to see how some PCIe SSD oems
- can leverage the extended topology features enabled by the new generation of
PCIe chips to provide fault tolerant arrays and much higher capacity and
performance then you see in current products in the market today.
how will these new "PCIe as a SAN" topologies interact with 3rd party
software from SSD ISVs?
Earlier this week I contacted Serge Shats -
who's CTO at FlashSoft
- and I asked
- How much visibility do you as an SSD software ISV have of the PCIe switch
- How important is it for you to know about the brand of PCIe chip?
and do you support vendor specific features?
Serge gave me this very
useful answer - "FlashSoft cluster SW (for persistent, distributed
read-write cache) replicates data between pairs of cluster nodes. We call them "replication
pairs". Any fast interconnect between nodes is useful for our solution.
Replication pairs design is the basis for our N-node cluster implementation.
- Or do you leave this to the SSD oem?
"In order to use that PCI switches our SW must fully control
initiator and target. It means that PCI switches should have an open API for
kernel modules. Number of nodes in the cluster based on the switch is limited.
That's why replication pairs design is useful here.
"The PCIe chip is invisible for us now but we can use it in the
future for implementation of low latency flash access model. However, the API
to PCIe chip should be commoditized. BTW,
NVMe to provide low latency,
page-based access to the flash."
back to PLX and what I call - the PCIe FT SAN
- the interface chipmakers can do it,
- the PCIe SSD designers are already using these chip families, and
That means the PCIe
SSD revolution which you've already seen flashing the enterprise core in the
past 5 years is far from over. It's only just begun.
- the ISVs have various ways handling new PCIe SAN and FT (fault tolerant)
functionality - when they have access to the APIs.
Or - as the
Borg might say - to
HDD array makers - "Resistance is futile. You will be assimilated."
more info about PLX take a look at the links above and
- 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.
- PLX Technology announced a definitive agreement to acquire all of the
outstanding shares of capital stock of
That gave PLX a range of SoCs for
NAS and DAS (USB,
interfaces and RAID
Iin March 2011 -
working with system partners worldwide to accelerate adoption of PCIe SSDs. PLX is a founding
member of the (Intel led)
Memory Host Controller Interface (NVMHCI) Work Group whose goal is to enable
the broad adoption of SSDs
In October 2011
OCZ agreed to acquire
the UK Design
Team (approximately 40 engineers located in Abingdon) and certain assets
from plx which will enable OCZ to accelerate the development of its next
generation of fast SSDs - while also reducing development costs. I learned later
that this core technology group was based on PLX's acquisition of
- "Strong (design) activity around
PCIe SSD platforms"
was mentioned by David Raun,
President and CEO - PLX -
as a growth opportunity in the company's
FY2013 Q2 financial
press release (pdf).
|Overview of PCIe
topologies for enterprise SSDs|
|Editor:- July 17, 2013 - Taking PCIe Out of the
Box - is just one of the many different topologies discussed in a new white
paper about PCIe and enterprise storage from PLX Technology called -
and PCI Express|
|so you want
to learn more about PCIe inside SSDs?|
|Editor:- PLX has created a
bunch of white papers and videos which may be helpful for product marketers (to
get a feel for the performance numbers) and designers (to get a feel for what
glue technology is available to get them started quickly). |
you're not in an SSD company - but just a user or
analyst - these
educational materials will help you understand what are the possibilities in the
PCIe SSD market today - and what new types of SSD will these chip technologies
support in the future? Here's a selection.
in enterprise SSD designs (video) - This includes an introductory tutorial
into PCIe and its performance and architectural capabilities for SSDs including
automatic failover and multi-host capabilities. PLX's switch chips also supports
failover if the fault occurs in the PCIe switch fabric chips themselves.
as a fabric for data center clustering (video) - PLX's PCIe extenders and
switch chips support lossless, software light, high performance CPU to CPU
memory transfers - and shared I/O - for upto about 200 nodes - as a
mini-cluster. At the next level of scalability you can interconnect these
PCIe SSD performance is the title of an article (co-written by Larry Chisvin, VP
strategic initiatives, PLX,
and Shawn Kung,
director of product marketing at
Marvell - which was
published in EETimes August 2012) - but it goes much further than that title
suggests. The article goes on to discuss scalability and configuration options
which enable PCIe based storage to provide benefits far beyond simply speeding
up legacy storage interfaces.
|SSD sudden power
|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
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
|don't all PCIe SSDs
look pretty much the same?|
|When you look at the
photos and headline specs for high speed PCIe SSDs - it's easy to come away with
the impression that they all look the same and have about the same performance.|
all - how different can they be?
But don't let the experience of the
2.5" SSD market -
in which clusters of consumer SSD vendors use the
same or similar
controllers and hover
close together inpopular
(consumer) performance rankings - give you the wrong idea about
this market the performance limits and capabilities of the SSD aren't set by an
old hard disk interface
and package limitations.
In the PCIe market the products you get are
limited only by the imagination of the designers - tempered by the guesses of
marketers who are trying to predict the optimum (most salable) features for an
|How big was the
thinking in this SSD's design?|
|Does size really does matter in SSD
By that I mean how big was the mental map? - not how many
inches wide is the SSD.
The novel and the short story both have their
place in literature and the pages look exactly the same. But you know from
experience which works best in different situations and why.
it comes to SSDs - Big versus Small SSD architecture - is something which was
in the designer's mind. Even if they didn't think about it that way at the time.
||For designers, integrators,
end users and investors alike - understanding what follows from these simple
choices predicts a lot of important consequences. ...read the article|