overview of Seagate's world leading enterprise SSDsby
editor - StorageSearch.com
offers 3 main SSD product families each of which in its own right can be be
considered to be either the most significant or certainly one of the
most significant in the market space it occupies - due to its history of
customer adoptions, technical lineage and market share.
The roots of these 3 product groups
originally evolved in 3 different companies SandForce, LSI and Seagate -
although even before they were all brought together by
a unified business group at Seagate it was clear that there had already been
elements of collaboration and co-operation by these companies with the intention
to optimize SSDs for particular applications.|
controllers - SandForce range
Since the first generation SandForce
SSD controllers came to market in
April 2009 this
product line has become the best known and most widely adopted flash SSD
controller used in enterprise, industrial and consumer markets.
the 3rd generation of products the SandForce design has a proven technology
roadmap which has spanned multiple flash generations. I've written more
about this in many articles -
for example here. You
can read more about the product family on
web site here.
2012 as the
SSD market was embarking on the
towards various types of
ECC I wondered how these techniques would be implemented (as indeed they
were) in the next generation (3rd generation) of SandForce controllers which
was unveiled in November
2013 . That was the product I called the "SSD market on a chip".
can get an idea of how that was done in the paper -
Error Correction Technology which describes the principle of how the
strength of ECC code is dynamically determined depending on the noisiness of
each segment of flash and also how LDPC code filtering can be used to
reinforce the data integrity on an "as needed" basis.
overview of all the main techniques used in the SandForce controllers can be
seen in the paper
Seagate's PCIe SSD - Nytro range
I've learned one useful rule of thumb from having been involved with the
PCIe SSD market since
it began in 2007 - it's this:- no single PCIe SSD design is optimal for all
And that's one reason why the Nytro range
appears so confusing at first glance.
Because it includes different
models which have been value engineered for the RAS, density, scalability and
performance needs of widely different use cases within the enterprise.
The Nytro product which is cost effective when it's part of an installation of
thousands in the same datacenter is different to the product you might choose
to accelerate legacy apps in a small population of traditionally architected
servers and storage.
range of PCIe SSDs includes 3 main product types - but I wouldn't be
surprised to see more in future as the enterprise PCIe SSD market spreads out
into more emerging application types.
This has been a very successful
product family. In January
2014 we learned that over 100,000 units had been shipped in the first 20
months of the product line - making it the #2 enterprise PCIe SSD product by
volume at that time.
Considering the number of competitors in the
market that was impressive.
My guess is that one factor for that past
success and for the Nytro's continuing success may be the lower perceived risk
of a PCIe SSD product line which uses the widely deployed SandForce controller
architecture (and which comes from the same source). That provides a reassurance
that improvements in the flash capabilities of one (due to its volume of use in
more markets) will ripple through into quicker improvements in the other.
That's in comparison to the weaknesses of some competing pioneer
enterprise PCIe SSD products I observed in 2014 in which support for newer
memory types was tardy - leading to higher than anticipated costs for users of
those products compared to the projections they must have expected.
It's 14 years since I began writing about the SAS (Serial Attached SCSI)
storage market and I recalled that Seagate and LSI which Seagate recently
acquired were both prominent in those early years - so I checked the timeline
from my history pages.
Sure enough - the first functioning silicon for
SAS was demonstrated in January 2004 by LSI, and Seagate was among the first
hard drive companies to complete interoperability testing with SAS hard drives
Seagate launched its first enterprise SAS SSDs (Pulsar brand)
in March 2011 - and its current high end SAS SSD product - the
SSD - is a lineal descendant of those years of experience in the market.
1200 SSD was the first 12Gb/s SAS SSD to ship in volume and can sustain a data
transfer rate of 750MB/s and random R/W IOPS of 110K / 40K respectively (4KB)
and an endurance rating of 10 DWPD
for 5 years.
For outline info
here and for more specifications see
|Seagate previews 60TB 3.5"
|Editor:- August 9, 2016 - Seagate today
it 's demonstrating prototypes of a 60TB 3.5" SAS SSD which will be
available next year.|
Micron collaborate on enterprise |
February 12, 2015 - Micron
a strategic multi year agreement which among other things will secure for
Seagate a supply of nand flash for the
SAS SSD market while
also providing for Micron a framework of SSD controller IP and designs with
which it can populate gaps in its own enterprise SSD range.|
announces strategic technology agreements with Baidu|
Editor:- September 17, 2014 - Seagate today
it has signed a strategic cooperation agreement with Baidu,
China's largest web services firm.
Under the agreement, Baidu will give priority to Seagate products when
considering components for all Baidu servers and storage facilities. In return
Seagate will give priority to Baidu when providing enterprise storage products
and relevant support, as well as maintain a dedicated engineering team for
Editor's comments:- This is a very significant
business announcement for Seagate. But it shouldn't come as any surprise - as
the destinies of the companies were already set on a natural convergence of
interests course which only needed the missing part of the IP jigsaw (SSDs) to
complete the required harmony.
Here below is a verbatim quote from
my coverage of Seagate's acquisition of LSI's SSD business in
think that even if Seagate disregarded any new markets - and focused only on the
high volume potential of existing cloud infrastructure customers and big web
entities (like Google and Baidu) - who need value based enterprise SSDs - but
who are perfectly capable of designing their own software and APIs and firmware
tweaks - then Seagate could leverage the LSI SandForce SSD roadmaps for the next
several years as a business tool to establish it as one of (several) leaders in
the utility SSD segment of the