|
Intel, the world leader in silicon innovation, develops technologies,
products and initiatives to continually advance how people work and live.
Additional information about Intel is available at www.intel.com/pressroom. |
.... |
 |
See also:-
Intel
- editor mentions on StorageSearch.com and
Intel's
SSD page |
| |
| . |
Who's
who in SSD? - Intel
by
Zsolt Kerekes,
editor January 2012
Intel was ranked #19 in
the
top 20 SSD companies
(Q4 2011), and has never been listed in the
fastest SSDs.
Intel
is 1 of more than 100 companies in the
2.5" SSD market, 1
of more than 50 companies with plans or products in the
PCIe SSD market and
recently acquired the InfiniBand
related product lines, IP and business assets of
QLogic.
In the
big versus small
SSD architecture zone - Intel is in the small camp. In the
skinny versus fat
RAM cache architecture - Intel's SSDs have been mainly at the regular
end. But Intel is agnostic in this area of the
SSD heresies - and
doesn't have a hard positiont.
Intel entered the SSD market in
2007. The
company's products since then have mainly appealed to the consumer and embedded
markets. Intel did manage to get some design wins into enterprise storage arrays
- but was dumped by oem customers due to many design flaws and recalls which
affected performance stability and reliability. Because the enterprise market is
such a huge future
opportunity - it's reasonable to assume that Intel will keep hammering away
at this door - and as it improves its knowledge of SSD fundamentals - it may get
somewhere. Its best SSDs have been those which didn't use its own controllers -
but instead those from 3rd party merchant
SSD IP companies.
Intel has invested in Anobit
- a company which is advancing into the higher performance small form factor
territory staked out by STEC,
SandForce and
SanDisk (Pliant).
Intel has designed some really bad SSDs - which demonstrated the company's lack
of knowledge about the fundamental building blocks needed to design a reliable
storage drive. In the summer of 2011 - after yet another SSD / firmware recall -
this led me to comment - "If Intel's SSD design business was a horse - it
would have been shot a long time ago and put out of its misery..."
Intel's customers have painfully learned that there's a lot more to designing
SSDs than soldering a bunch of memory chips to a controller and host interface.
Intel's designers should have known that too.
In the long term I expect
Intel will buy an
SSD company or license more SSD technology. |
| . |
 |
| . |
|
|
Intel was a pioneer in
SSD market
history. But then took a 20 year sabbatical.
In the
early 1980s - Intel shipped an SSD based on
magnetic bubble
memory technology which emulated a 1Mb floppy drive. (I had one of the
evaluation kits.) But this early foray into solid state storage didn't meet
Intel's need for scalability either as a technology or as a business. So Intel
spun off the magnetic division in 1987 to
Memtech. Memtech
ditched bubble memory but became a pioneer in the rugged and military flash SSD
market (an example product was the
3.5" PATA
compatible Wolverine). In August 2005 Memtech was acquired by
STEC.
Intel's
troubled past with memory products - (many of which it had invented - but
abandoned to Asian competitors) was probably a factor in delaying its decision
to re-enter the SSD market till 2007 - which
was 2 years after Samsung
had publicly declared this to be a strategic market. Within a few years of this
re-entry, however, Intel was shipping
2.5" SSDs with
performance specs superficially better than the
leading products
previously available from Asian companies (Mtron and
Memoright).
Intel's
troubled present with SSDs - In the rush to gate crash the
frothing SSD market
bubble, and lacking vital end-user storage industry experience -
Intel has produced succeeding products (up to 4 SSD recalls, upgrade fixes or
delayed shipments according to industry estimates) with undesirable
halo effects or
flaky operation -
which have seriously dented its reputation among designers of enterprise class
SSD arrays.
In
October 2008 - Intel
started shipping the X-25E - a
fast
2.5" 32GB
SATA SLC
flash SSD. Read
latency is 75 microseconds and a 10 parallel channel architecture enables it to
sustain R/W throughputs of 250 / 170 MB/s. Random IOPS performance is
impressive with a 10 to 1 R/W ratio which is inline with the best
designed enterprise flash SSDs. Using 4kB blocks - random R/W IOPS are 35,000
and 3,300 respectively.
In his October 2008 blog, Linux
creator Linus Torvalds
wrote about his own
experience
with Intel's new SSD. Just as relevant are the many comments which followed
about better (and worse) products.
In December 2008 -
Hitachi and
Intel announced they were
jointly designing a new range of high IOPS flash SSDs with
Fibre Channel and
SAS interfaces for
the server market. The new products, which will be exclusively marketed by
Hitachi GST - are expected to ship in Q1 2010.
In January 2009
-
Kingston Technology
announced it will sell rebranded high speed SSDs supplied by
Intel as Kingston's
SSDNow E Series.
In February 2009 -
Solid Data Systems
published a Test
of Intel's X25 Flash SSD Performance (pdf). The white paper reveals the
degradation in performance in Intel's headlining SSD, due to weak garbage
collection. This is something which had been known about in the industry - but
not in this level of detail (except under NDA).
In April 2009
- a report on TGDaily.com
said that Intel
is EOLing its
Z-P230
SSD module which was aimed at the netbook market. If you look at the
1.0" SSDs directory
here on StorageSearch.com you'll
see that 25 companies now make SSD chips, DOMs or modules designed to
fit into very small footprints.
In July 2009 - Intel announced a
process
shrink for its
X25-M -
SATA 2.5" MLC flash SSD. The new 34nm devices deliver upto 8,800
(4KB) write IOPS and up to 35,000 read IOPS. R/W speeds are 250MB/s and 70MB/s
respectively. R/W latenciy is 65µS and 85µS. The 160GB model is
priced at $440 (1,000 unit price point).
In September 2009
- Pillar dumped
Intel SSDs due to
flaky operation and
switched
to STEC. Maybe they
should have spent a bit more time qualifying the Intel product beforehand - or
done a better job at it?
Also in September 2009 -
Kevin T Crow, Strategy Specialist, NAND Solutions Group, -Intel shared his
SSD Bookmarks
with readers of
StorageSearch.com.
In
October 2009 - Intel
joined the growing roster of SSD
companies who have
announced
support for Trim functions. These benefit flash SSDs which don't have
internal fast active garbage collection. The company recommends users install
the firmware update and toolbox, and run the Trim function daily to ensure best
performance.
In February 2010 -
Intel and Micron announced they
are sampling the
world's
1st 25nm NAND flash memory. This gives 8GB MLC (classic 2 bit)
flash memory in a
stackable TSOP.
In July 2010 - the terrible tale of one
enterprise customer hitting
endurance limits
with Intel's SSDs
was mentioned in an
interview with
Fusion-io's CEO.
In September 2010 -
Intel's SSD
Bookmarks on StorageSearch.com
were updated with new links to help prospective enterprise SSD users.
In
November 2010 - Intel Capital led a $32 million funding round into
Anobit (an
SSD controller
company).
In
March 2011 - Intel
published version 1.0 of a new proprietary standard for designers of
PCI SSDs in systems
which use Intel processors - the
NVM Express Optimized PCI
Express SSD Interface.
Intel launched a new
2.5" SSD aimed at
legacy notebook
designs which have 3Gbps SATA
ports. The
Intel
SSD 320 (which includes 128 bit encryption) is available with MLC
capacities from 40GB ($89 1k price) to 600GB ($1,069 1k). R/W speeds are
270MB/s and 220MB/s respectively. R/W
IOPS are
39,500 and 23,000. In this new design , Intel has added redundancies that
will help keep user data protected, even in the
event of a
power loss.
In October 2011 - an article in
VR-Zone
discusses a "leaked" Intel SSD roadmap which indicates the company may
enter the PCIe SSD
market in 2012. It's hardly a revelation - because Intel is member of
technical groups which are
co-ordinating standards in this segment of the SSD market - and until standards
for the Hybrid Memory Cube get
established (which could take another 3 years) - the PCIe SSD market is the
closest attachment that an SSD can make to an Intel server host processor bus.
And it has the additional attraction of not needing 3rd party
storage interface glue -
unlike SATA,
SAS,
FC and
IB - thereby giving
more control to any chip company which does it right. Over 30 companies have
already shipped PCIe SSDs in the past 3 years. This will be a multi-billion
dollar market segment according to StorageSearch.com's long range enterprise
SSD market model.
In January 2012 -
Intel
announced
an agreement to acquire the
InfiniBand related
product lines, IP and business assets of
QLogic. |
|
|
| |
"Intel
would like to be
where Fusion-io's SSDs are now...
snuggling
up close to the host CPU." |
| ...October 2011 editor's comment
in
PCIe SSD news | | |
| . |
|
|
| . |
| SSD power down
management |
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. |
|
| | | |
| . |
| sugaring MLC for the
enterprise |
When flash SSDs started to be used as
enterprise server accelerators in 2004 - competing
RAM SSD makers said
flash wasn't reliable
enough.
RAM SSDs had been used for server speedups
since 1976
- and in 2004 they owned the enterprise market. (Before 2004 - flash SSDs
weren't fast enough and had mostly been used as rugged storage in the
military and
industrial
markets - and in space
constrained civilian products such as smartphones.)
By 2007 it was
clear that the endurance
of SLC flash was more than good enough to survive in high
IOPS server
caches. And in the ensuing years the debate about enterprise flash SSDs shifted
to MLC - because when systems integrators put early cheap consumer grade SSDs
into arrays - guess what happened? They burned out within a few months - exactly
as predicted.
Since 2009 new
controller
technologies and the combined market experience of enterprise MLC pioneers
like Fusion-io and
SandForce have
demonstrated that with the right management - MLC can survive in most (but
still not all) fast SSDs.
Now as we head into 1X nanometer flash
generations new technical challenges are arising and MLC SSD makers disagree
about which is the best way to implement enterprise MLC SSDs.
Which
type of so called "enterprise MLC" is best? Can you believe the
contradictory marketing claims? Can you even understand the arguments? (Probably
not.)
And that's why marketing is going to play a bigger part in the
next round of enterprise SSD wars as SSD companies wave their wands and reveal
more about the magic inside their SSD engines to audiences who don't really
understand half of what they're being told. |
|
| | |
| . |
| 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.
After
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
PCIe SSDs.
In
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
ideal SSD. |
|
| | |
| . |
| this way to the Petabyte
SSD |
In 2016 there will be
just 3 types of
SSD in the datacenter.
One
of them doesn't exist yet - the bulk storage SSD.
It will start to
replace the last remaining strongholds of
hard drives in the
datacenter due to its unique combination of characteristics, huge storage
density, low running costs and operational advantages.
Bulk
storage SSDs will displace the last remaining hard drives in the enterprise
server market by 2020 - even if the price of a new hard disk drops to zero
and enterprise HDDs are given away free! |
 |
... |
The new business and
architectural models of the datacenter - how we get from here to there - and
the technical and problems which will need to be solved - are just some
of the ideas explored in this
visionary article. | | | |
| . |
| |
