click to visit home page
leading the way to the new storage frontier .....
top 10 SSD oems
top SSD companies ..
click to see the collection of  SSD reliability articles here on
SSD reliability ..
SSD SoCs controllers
SSD controllers ..
click to read the article - Big versus Small SSD  architectures
Size in SSD architecture ..
click image to read the article - principles of bad block management in flash SSDs
bad blocks in flash SSDs ..

DensBits - circa 2014

Founded in 2007 and headquartered in Haifa, Israel - DensBits has designed the world's first Memory Modem, a revolutionary technology created to address the ever-increasing need for low-cost, high-performance NAND Flash-based storage systems.
..... DensBits logo - click for more info

see also:- DensBits - editor mentions on

DensBits was acquired by Broadcom for about $15 million in December 2015.

articles related to DensBits' place in SSD market

SSD controllers
SSD data integrity
acquired SSD companies
the Top 20 SSD Companies
SSD endurance - the forever war
adaptive R/W, DSP ECC flash IP in SSDs
principles of bad block management in flash SSDs
Why does size matter in SSD controller architecture?

Who's who in SSD? - by Zsolt Kerekes, editor - April 2013

DensBits is a leading company in the adaptive R/W DSP ECC flash management IP market.

The company profile, below, which I wrote exactly a year ago - is still the current picture today. The only thing which has moved on since that was written is the state of the SSD market's willingness and readiness to engage with this type of technology - something which I discussed in another article - Strategic Transitions in SSD.

DensBits is an SSD controller / IP company which has been working on the problem of how to get better data integrity and reliability at the SSD level when using new and future generations of high density x3 MLC (TLC) flash. The effects of scaling in these new memories presents a set of characteristics which traditional SSD controller management techniques fail to deal with effectively and economically.

At first glance DensBits's technology looks like an impressive and clever way of viewing the problem and efficiently dealing with the vagaries of the flash bit data stream in a systematic way – rather than bundling up a collection of ad-hoc schemes which might have worked for SSDs in earlier memory generations – but which are independent and conflict for the same CPU and capacity resources.

DensBits says that the combination of their techniques results in significant gains in endurance, cost reduction and performance - when viewed at the system (whole SSD array level).

DensBits's technology - which is available in controller and IP form offer utility in both the consumer and enterprise SSD markets.

In a superficial way DensBits's technology package looks similar to Anobit, STEC's CellCare and Smart's Optimus.

But just as there are many different ways of doing wear leveling - there are also many different twists on how to apply adaptive R/W, DSP ECC and information theory to the patterns of good and bad cells in a flash array.

DensBits has created a complete IP suite - which includes their own collaborative DSP, ECC and flash management instead of bolting DSP onto a traditional suite of flash management tricks as an afterthought.

At the SSD system level the benefits are additive and multi-dimensional.

Amir Tirosh (EVP Marketing) gave me an interesting example.

Suppose you want to optimize the SSD for speed - as in an enterprise product. He said that DensBits enables you to get 60% faster write speeds. That's not (as you might suppose) because of a parallel architecture or memory cycling trick. They can do faster writes because their ECC is so strong - they can get away with using shorter write pulses than traditional designs - and still reconstruct the data reliably.

The DSP and flash manager learns from the ECC system how good the flash pages are - and can deal with them adaptively. The bag of memory modem tricks includes for example - using stronger and weaker codes in different regions of the flash space. That gives more efficiency when it comes to usable versus raw flash bits. And it's dynamically determined. Going back to the faster writes thing... Shorter write pules place less stress on the memory - so yielding better intrinsic reliability.

At the other end of the market - the consumer end - the same IP suite can be set up to favor high capacity (and lower performance) which optimizes power consumption and battery life.

DensBits isn't the only company which says it can get good results from cheap flash. There are others who also claim very high gains in endurance (x10, x20 - and I've even seen claims of x100) from particular SSD companies in the consumer, industrial and enterprise markets. But DensBits is the only company I know of - which appears to have credible ambitions across such a wide range of flash markets while using a single coherent IP set. Specifically - DensBits says it would like to become the "ARM" of the flash industry, providing reference designs for a wide range of applications.

I asked - how does DensBits aim to sell its technology?

I learned that the company is willing to license parts of its SSD IP set to companies as diverse as flash memory makers, consumer SSD oems and enterprise SSD oems.

In some markets the IP will be used to reduce power consumption, in other it will increase capacity, or it can be used to get better performance and reliability.

The interesting thing is that DensBits has designed its licensing model around charging a percentage of the ASP of the whole SSD (that's the sum of the controller and memory parts).

Finally - on the naming of SSD product names thing... The more I think about about the more clever I think is the name that DensBits chose to describe what it offers - as the Memory Modem. That's because one way to view the resources in a flash memory is as a stream of bits which interact with the SSD controller.

I guess that not many people will understand the nuances of that reference - because unless you know about information theory and communications and noise - you don't get the full message:-) But most people who will look at DensBits with a view to using their technology (or figuring out where they stand in the competitive landscape) will probably have enough of an electronics background to appreciate the perspective they get on DensBits from this particular analogical angle.

related articles

adaptive flash care management IP (including DSP) for SSDs - what is it? and who else does it?

MLC flash lives longer in my SSD care program - the new fad in selling flash SSDs... rival claims about life assurance and health care

how fast can your SSD run backwards? - Mysterious behavior in SSDs? - 11 core SSD asymmetries are often the root cause.
selected DensBits milestones - from SSD Market History.

In October 2010 - DensBits closed Series B funding, led by a new investor, Bessemer Venture Partners and existing investor Sequoia Capital.

In April 2012 - DensBits announced it is sampling a new SSD controller - the DB3610 - which supports the latest 2Xnm and 1Xnm TLC (3 bits/cell ) MLC flash with an extreme endurance figure of more than 10K P/E cycles and R/W performance of up to 95MB/s / 65MB/s and 4,000 / 1,100 R/W IOPS (4KB), for sequential and random operations, respectively.

InJune 2012 - Seagate announced it will use DensBits's flash care technology in the design of forthcoming consumer and enterprise SSDs. Seagate has also made an equity investment in DensBits.

In August 2012 - DensBits announced that its 3 bits/cell adaptive DSP IP controller technology (which the company brands as its Memory Modem technology) was a Best of Show award winner at the Flash Memory Summit in the category of most innovative flash memory technology.

In October 2013 - DensBits announced that it has licensed its advanced Memory Modem technology to Toshiba for use in new designs of SSDs.
sugaring nand flash 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.

Since then flash has dominated the installed base of enterprise SSD starting with SLC, followed by MLC, then a correction to eMLC and now some SSD makers are saying TLC (x3) may be good enough.

But it's not just the raw memory type which determines the suitability of which flash can work reliably in what type of enterprise SSD. The controller IP and cache architecture can make a difference to the endurance of x5, x10, x20 - and I've even heard claims of x100...

That means TLC (aka x3) - with the right SSD IP - may be as good as SLC in some types of applications. And it costs a lot less and has higher capacity.

What do you need to know?

Who are you going to believe?
click to read article Unlike the Cola Wars - you can't take the risk of a bad enterprise nand flash SSD taste test. the article

storage search banner

The interesting thing about NVMdurance's IP is that it delivers endurance amplifying results using a lightweight runtime controller. But this doesn't stop you getting even better results by adding DSP correction as an additive process.
varrious conversation with NVMdurance
SSD ad - click for more info
Even more reasons to need adaptive R/W DSP in 3D
Editor:- January 2015 - Even if you already thought that adaptive R/W and DSP was an essential way for getting usable SSDs out of smaller 2D nand flash - then there are even more reasons for using this technology on the journey in 3D.

That's the conclusion you'll come away with after seeing DensBits's paper (presented at the 2014 Flash Memory Summit) called the Necessity for a Memory Modem in 3D Memories (pdf)

Among other things in this paper:- DensBits says that the scope for inter-cell interference grows from 8 identifiable routes in 2D to 26 for each cell in 3D.

But memory modem technology (DensBits's branding for their collection of adaptive R/W DSP IPs) will (over and above everything it already does for 2D) intelligently decouple read operations according to the severity of read operations expected in the new 3D architectures - and even supports the notion of TLC (x3) within 3D. (Which "needs state of art decoder and signal processing".)

Their conclusion? - Memory Modem technology is required for 3D NAND scaling the article
3d interference effects in nand cells
SSD ad - click for more info
Toshiba will design new SSDs using DensBits' adaptive flash controller IP
Editor:- October 21, 2013 - DensBits today announced that it has licensed its advanced Memory Modem technology (a variety of adaptive R/W and DSP flash controller IP) to Toshiba for use in new designs of SSDs.
"A layered approach is a useful abstraction for handling various data failure mechanisms in 1xnm MLC and 2xnm TLC flash."
Inside the Memory Modem Architecture (pdf) - by Hanan Weingarten, CTO DensBits - (August 2012) - presentation at the Flash Memory Summit
"When I spoke to Amir Tirosh at DensBits in April, 2012 I knew within a few minutes that DensBits would be a company which would go straight from stealth mode into the Top SSD Companies List - within the same quarter..."
...from:- the Top SSD Companies in Q2 2012.
Seagate chooses DensBits for TLC and 1Xnm
Editor:- June 25, 2012 - Seagate announced today it will use DensBits's flash care technology in the design of forthcoming consumer and enterprise SSDs.

Seagate has also made an equity investment in DensBits.

Editor's comments:- I've already written more than enough about about this technology trend recently on the home page.

As DensBits told us a few months ago - their business plan is work with companies which sell complete SSDs - instead of licensing their technology to other controller makers. DensBits's technology spans the widest spectrum of adaptive DSP flash SSD applications from consumer and industrial to fast-enough enterprise SSDs.

If Seagate leverages DensBits's flash technology successfully - the result will be tougher competition for companies like SanDisk in the consumer SSD market and tougher competition for STEC and SMART in the fast-enough enterprise SSD market.
SSD ad - click for more info
"....nobody really knows how long NAND can keep scaling. So we have to keep trying and we have to be innovative. But we are aggressively working on the future NAND, future technologies beyond NAND...."
Ritu Shrivastava
V.P., Technology Development SanDisk
SanDisk technology roadmap presentation
DensBits samples new TLC flash controller
Editor:- April 30, 2012 - DensBits today released a new SSD controller - the DB3610 - which supports the latest 2Xnm and 1Xnm TLC (3 bits/cell ) MLC flash with an extreme endurance figure of more than 10K P/E cycles and R/W performance of up to 95MB/s / 65MB/s and 4,000 / 1,100 R/W IOPS (4KB), for sequential and random operations, respectively.

DB3610 employs DensBits' Memory Modem technology (adaptive DSP in SSD IP) which enables a native TLC solution with more than double the endurance of 2 bits/cell (MLC), and near-MLC R/W performance.

Editor's comments:- It's easy to miss the significance of new SSD products and technologies. And you might think from looking at the text and numbers above - this is a consumer style SSD controller - and it's not for me.

But I think DensBits may become one of the top 20 SSD companies real soon - unless it gets acquired before that happens.
SSD symmetries article ... Its flash technology has very high roadmap symmetry and the potential to impact competitiveness in the consumer, embedded and fast-enough enterprise SSD markets with a splash that's as big as SandForce made when it emerged on the scene 3 years ago.
...Later:- guess what? - in the first 10 days of May 2012 - search volume for readers following up DensBits propelled them into the top 5 SSD companies - in line with my prediction above.

Is that just a short term spike or an interest that will endure?

We'll have to wait another few months to find out.
flash SSD capacity - the iceberg syndrome
Have you ever wondered how the amount of flash inside a flash SSD compares to the capacity shown on the invoice?

What you see isn't always what you get.
nothing surprised the penguins - click to read  the article There can be huge variations in different designs as vendors leverage invisible internal capacity to tweak key performance and reliability parameters. the article
What's the best way to design a flash SSD?

and other questions which divide SSD opinion
More than 10 key areas of fundamental disagreement within the SSD industry are discussed in an article here on called the the SSD Heresies.
click to read the article - the SSD Heresies ... Why can't SSD's true believers agree upon a single coherent vision for the future of solid state storage? the article
. is published by ACSL